Housing First or Treatment First Interventions for Mental Illness and Substance Use Disorders among the Chronically Homeless?
George Galster (Wayne State University)
After a decade of steady declines, since 2016 the number of chronically homeless individuals—both sheltered and unsheltered—has been inexorably and dramatically increasing. Perhaps most worrisome is the fact that unsheltered chronic homelessness is rising most rapidly, now registering almost 100 thousand individuals, almost twice the number as those in shelters. A nontrivial number of our chronically homeless population manifest substance use disorders and/or unresolved mental illnesses. This population is likely most disruptive of the civic spaces and neighborhoods that they frequent, challenging to house stably over the long term, and costly to the nation through high use of publicly funded health and criminal justice systems resources.
The longstanding US approach for treating chronically homeless people with mental illness and/or substance use disorders has been Housing First (HF): providing permanent supportive housing with wrap-around services, with no requirements on participants to engage in behavioral health support. The Trump Administration has recently reversed course, adopting an alternative strategy: Treatment First (TF). This research note addresses the question of which approach is more effective for treating mental illness and substance use disorders, based on the best available scientific evidence. Specifically, I review results from all random control trials (RCTs) conducted in the US for HF or TF interventions for homeless populations. RCTs for both HF and TF strategies have, to my knowledge, never been comprehensively assembled and compared to ascertain which type of intervention is more efficacious for America’s chronically homeless populations’ mental and behavioral health.
The results show that, compared to “treatment as usual,” HF: (1) does not provide superior outcomes on general mental health or symptom severity indices; (2) inconsistently may provide superior outcomes regarding use of professional mental health care; (3) does not consistently provide superior substance use outcomes. Of the 21 outcome indicators tallied in these domains across multiple RCTs, only seven showed a statistically significant favorable outcome for HF; another statistically significant outcome was unfavorable for HF. The results for TF also indicate that only in a minority of the RCTs were there statistically significant improvements in homeless adults’ or youths’ depression, anxiety, mental health status or PTSD symptoms, compared to “treatment as usual.” A higher share of RCTs showed the efficacy of TF in reducing homeless youths’ use of alcohol and hard drugs, however.
Unfortunately, it is challenging to make precise cross-study comparisons because extant studies differ in: homeless populations investigated, the outcome measures employed and when they are assessed, the characteristics of and potential services provided to the “treatment as usual” control groups, and the details of the interventions themselves. To address this challenge I propose a major, federally and foundation-funded study that would directly compare the mental health and behavioral outcomes produced by the “best” HF and TF approaches administered to common chronically homeless populations. Specifically, an expert panel would identify the most promising HF and TF approaches based on extant research. One or more multi-city, long-term RCTs would then be designed to test these approaches as alternative treatment streams compared to similar controls, thereby reducing (if not eliminating) most of the key sources of cross-study heterogeneity currently bedeviling the field. I envision the famous Moving To Opportunity RCT of the 1990s as a useful prototype here, providing what was broadly seen as definitive evidence regarding the economic, health and behavioral effects of giving families living in public housing amid concentrated poverty neighborhoods one of two alternative treatments: housing rental vouchers with “no strings attached” vs. vouchers that only could be used in low-poverty neighborhoods but came with case-managed relocation assistance. Admittedly, acquiring the requisite high-quality evidence base poses immense challenges inasmuch as there are so many different “flavors” of HF and TF interventions.
In the absence of such a definitive study(ies), however, what can be said to policymakers? It is clear that relatively few US-based RCTs suggest that either HF or TF demonstrate significant efficacy in altering mental illness and substance use disorder outcomes for the chronically homeless. This evidentiary ambiguity renders affirmative policy implications difficult; warmings about what not to do thus seem more appropriate. In this vein, the evidence here implies that neither HF nor TF strategies should been seen as a panacea for homelessness and its associated mental and behavioral health issues and become the sole focus of federal policymakers. Instead, what seems more prudent is a more balanced (and expanded) policy that encourages and financially supports the flexible pursuit of bothHF and TF strategies, so that local officials and case managers can apply the treatment deemed most appropriate for the particular client at hand.
Relieving the burdens of severe mental illness and substance use disorders among the chronically homelessness will require more than building more supportive housing, providing more rental vouchers, and enhancing tenant protections, as invaluable as these initiatives are. We will need to treat this epidemic as a public health problem. As such, it will take coordinated efforts among governmental and non-profit entities that often have been isolated in their own housing, mental health, and addiction recovery program silos. Even such an integrated approach will fail, however, if the requisite physical and human resources in the public health domain are not in place.
In sum, my review of results from randomized control trials conducted in the US demonstrates that neither HF nor TF approaches represent panaceas in helping the chronically homeless improve their mental health or free themselves from addiction. These sobering findings imply that we should adopt a more heterodox strategy that can better tailor treatments to individual circumstances—and build the requisite human, housing and institutional resources in the public health domain to support such—if we are to fight the symbiotic challenges of chronic homelessness, mental illness, and substance use disorders successfully.
George Galster earned his Ph.D. in Economics from M.I.T. and is the Clarence Hilberry Professor of Urban Affairs (emeritus) at Wayne State University in Detroit. He has published 185 peer-reviewed articles, 10 books and 49 book chapters on a wide range of urban topics. The Urban Affairs Association placed him on their “Service Honor Roll” in 2014 and awarded him the prestigious “Contributions to the Field of Urban Affairs” prize in 2016. He currently resides in Portland, OR’s Pearl District.