The traditional approach to chronic homelessness, often called the treatment-first or staircase model, required people experiencing homelessness to demonstrate sobriety, treatment compliance, or employment readiness before accessing stable housing. The logic was intuitive: address the conditions that led to homelessness before providing the stability of housing. The evidence, accumulated across randomized controlled trials and large-scale real-world implementations in multiple countries, consistently shows the opposite. Housing stability is the precondition for addressing those other challenges, not the reward for having addressed them. People who are not stably housed cannot reliably manage mental health treatment, substance use recovery, or employment programs.
The Evidence from Randomized Trials
The landmark At Home/Chez Soi study in Canada randomized over 1,000 individuals experiencing homelessness and mental illness to Housing First or treatment as usual across five Canadian cities. After two years, Housing First participants spent significantly more time stably housed, had better quality of life and community functioning, and were more engaged in mental health treatment than control participants, not less. Contrary to the premise of the staircase model, providing housing without requiring treatment readiness did not reduce treatment engagement. It increased it, because stable housing provided the foundation that makes treatment engagement possible.
Similar results have been replicated in the United States, Finland, Denmark, France, and Australia. The effect sizes vary by population and context, but the directional finding is consistent across cultural and policy contexts: Housing First produces better housing stability and at least equivalent mental health and substance use outcomes compared to treatment-first approaches, almost always at equivalent or lower cost per person over time because it prevents the costly emergency service cycling that treatment-first approaches fail to interrupt.
The Housing Supply Problem
The evidence for Housing First as a model is strong. The barrier to implementation is not clinical or programmatic. It is a housing supply problem. Housing First requires affordable, available, permanently subsidized housing units, which in most major American cities are in severe and growing shortage. The affordable housing shortage is not primarily a failure of Housing First programs. It is a product of decades of federal disinvestment in public housing, inadequate production of subsidized units, exclusionary zoning that prevents dense affordable development, and land values that make affordable housing economically unviable in the neighborhoods where services and employment are concentrated.
Addressing homelessness at scale requires addressing both the programmatic model and the housing supply conditions that determine whether the model can be implemented. Programs that successfully place individuals in housing only to lose those placements because of lack of available units are not program failures. They are policy failures operating upstream of the program level.
Beyond Chronic Homelessness
Housing First was developed for chronically homeless individuals with serious mental illness, the population for whom the traditional staircase model had most clearly failed. Its principles have since been extended to families experiencing homelessness, youth experiencing homelessness, and veterans. The evidence for these populations is less mature than for the chronic adult population but consistently supportive of rapid rehousing approaches over shelter-based transitional housing. The underlying logic translates: stability enables recovery, and the sooner housing stability is achieved, the better the long-term outcomes across every domain that matters.
