The “Housing First” Approach Has Failed: Time to Reform Federal Policy and Make it Work for Homeless Americans

Report Housing

The “Housing First” Approach Has Failed: Time to Reform Federal Policy and Make it Work for Homeless Americans

August 4, 2020 23 min read Download Report
Christopher Rufo
Christopher Rufo
Former Visiting Fellow for Domestic Policy Studies
Christopher was a Visiting Fellow for Domestic Policy Studies at The Heritage Foundation.


Homelessness has gained national attention with the growth of public encampments and street disorder, particularly in West Coast cities. Over the past decade, the federal government has spent billions on “Housing First” programs, which provide permanent housing for the homeless without requiring sobriety or participation in treatment. Although Housing First programs demonstrate strong rates of short-term housing retention, they do not improve symptoms related to drug addiction, mental illness, and general well-being—and have not reduced overall rates of homelessness. Moving forward, policymakers must re-orient federal homelessness policy toward better outcomes, prioritizing programs that require treatment, accountability, and a path to self-sufficiency for the homeless.

Key Takeaways

Policymakers must rethink the federal government’s Housing First policy, which has failed to reduce overall homelessness and does not improve human well-being.

Treatment First programs address the substance abuse and mental illness behind much homelessness, and provide the most effective pathway to self-sufficiency.

Policymakers should redirect housing funds to programs with a proven record of helping the homeless to overcome addiction, find employment, and achieve independence.

Homelessness in America has gained national attention with the growth of public encampments, particularly in West Coast cities. According to the most recent point-in-time count, there are 567,715 homeless individuals in the United States, including 356,422 in emergency shelters and 211,293 who are living unsheltered in tents, cars, and on the streets.REF In some city neighborhoods, such as in Seattle, Portland, San Francisco, and Los Angeles, thousands of men and women are sleeping outdoors, and the authorities have become overwhelmed by widespread drug abuse in public, crime, and social disorder.

Federal policymakers have responded by devoting additional resources to the U.S. Department of Housing and Urban Development’s (HUD’s) programs for emergency shelters and permanent supportive housing. However, despite billions in funding for these programs over the past five years, the number of people living unsheltered has increased by 20 percentREF—with no signs of abating in the near future. The increase has been particularly steep in the three West Coast states of California, Oregon, and Washington, which account for 61 percent of the nation’s unsheltered homeless.REF

This issue has come to greater public consciousness because of the visible, persistent, and tangible failure of some local governments to address public camping and disorder in their streets. Local policy responses must tackle such problems.REF Progressive political leaders have insisted that homelessness is caused by lack of affordable housing, but in Seattle, San Francisco, and Los Angeles, the number of people on the streets has increased year over year despite large-scale investments in subsidized and permanent supportive housing.REF

In order to reduce homelessness, policymakers at all levels must understand that chronic and long-term homelessness is not primarily a housing problem—it is a human problem. According to a 2019 report from the University of California Los Angeles analyzing data from 64,000 surveys, 75 percent of the unsheltered homeless have substance-abuse disorders, 78 percent have mental health disorders, and 84 percent have physical health conditions.REF In other words, these are not simply people who lack shelter; the majority are suffering from profound human pathologies.

As homelessness threatens to become an entrenched problem in American cities, it is more important than ever for policymakers to have a clear understanding of the failures of current policy, as well as the potential for reform. In re-orienting the public response toward better outcomes, policymakers must begin with a simple premise: Any effort to reduce homelessness must address addiction, mental illness, and social pathologies—not just physical housing, lack of which is frequently a reflection of deeper problems.

Currently, the federal government plans to devote record-high resources to homelessness programs. President Donald Trump has proposed $2.8 billion in funding for HUD’s homelessness programs in his fiscal year (FY) 2021 budget, and the recently passed Corona, Aid, Relief and Economic Security (CARES) Act provides an additional one-time $9 billion expenditure for HUD’s Emergency Shelter and Community Development Block Grant programs.REF HUD has the responsibility to spend this funding in a way that brings people off the streets and works toward recovery, rehabilitation, and self-sufficiency. This Backgrounder outlines the limitations—and outright failure—of current federal policies, and proposes an agenda for reform.

Federal Homelessness Policy: Background

Federal policymakers have been engaged with the issue of homelessness since 1983, when President Ronald Reagan established the first Federal Interagency Task Force on Food and Shelter for the Homeless. Four years later, based on the task force recommendations, President Reagan signed into law what is now known as the McKinney–Vento Homeless Assistance Act, establishing the U.S. Interagency Council on Homelessness (USICH), the Emergency Shelter Grants Program, and HUD’s Homeless Assistance Grants—which remain the cornerstone of the federal response to homelessness.REF

Next, through a series of administrative and legislative changes beginning in the 1990s, HUD introduced the Continuum of Care (CoC) Program, which created local quasi-governmental entities in metropolitan areas that can apply for HUD funding and administer local homelessness programs.REF This CoC model is the primary mechanism for distributing federal funding for homelessness programs. It bypasses state and local governments and goes directly to CoC organizations, which are responsible for administering contracts to local shelters, housing programs, and service providers.

In recent years, the federal policy response to homelessness has centered on a single concept: Housing First, which holds that the public should provide permanent housing for the homeless without requiring abstinence from drugs and alcohol, or even participation in substance abuse treatment or mental health services.

Under the George W. Bush Administration, USICH presented the Housing First approach as the key breakthrough to ending homelessness. Over the course of the following decade, USICH and HUD redirected billions of dollars in funding away from transitional housing programs and toward Housing First programs. Despite concerns from service providers that this policy would reduce resources for emergency shelters, transitional housing, and treatment programs, the Housing First coalition plowed ahead. By 2009, Housing First had become the de facto policy at the local, state, and federal level, as local service providers oriented their programs toward federal funding priorities in order to maximize their ability to receive HUD grants. During this period, 234 cities officially adopted the philosophy of Housing First and submitted “10-year plans to end homelessness.”REF

Today, the vast majority of federal homelessness funding is spent on Housing First programs. According to a recent analysis, HUD allocates approximately 74 percent of all competitive grants to permanent supportive housing projects—even though these projects were originally intended to only serve the small fraction of the general homeless population that is “chronically homeless.”REF As a result of this dramatic shift in funding, between 2007 and 2018, the number of Housing First beds increased from roughly 189,000 to 361,000, and the number of transitional housing beds fell from roughly 211,000 to 101,000.REF

Following the federal government’s lead, many states and municipalities have adopted the Housing First philosophy at the local level. The City and County of Los Angeles, which are “home” to more than 59,000 homeless men and women, recently passed a $1.2 billion bond for the construction of permanent supportive housing.REF Other major West Coast cities, including San Francisco, Portland, and Seattle, have adopted similar policy objectives and significantly increased spending on Housing First programs.

The Failure of Housing First Programs

Unfortunately, the promise of Housing First—solving homelessness and reducing public expenditure simultaneously—turned out to be little more than wishful thinking. More than a decade after hundreds of American cities submitted their Housing First-inspired “10-year plans to end homelessness,” none of these plans has survived contact with reality. In many West Coast cities, homelessness is more acute than ever before, and, as of January 2020, is the top concern among voters in California,REF Oregon,REF and Washington State.REF

At the federal level, there is a growing consensus that Housing First must be reconsidered. The Trump Administration’s Council of Economic Advisers recently released a report demonstrating that, despite nearly doubling the nation’s stock of Housing First units since 2007, there is no compelling evidence that overall homelessness has been reduced.REF Additionally, according to the National Academies of Sciences, there is “no substantial published evidence as yet to demonstrate that [Housing First] improves health outcomes or reduces health care costs.”REF The new director of USICH, Robert Marbut, has signaled strong opposition to Housing First, releasing a chart that shows a dramatic rise in unsheltered homelessness after the widespread adoption of Housing First policies.REF

What went wrong? Despite Housing First advocates’ insistence that their approach would be “research-and-data-driven, performance-based, and results-oriented,”REF the policy failed to produce results even on its own terms.

In the early 2000s, homelessness service providers and a coalition of activists successfully pitched Housing First based on a limited number of studies that purported to show high rates of housing retention and significant cost savings.REF However, as the literature has accumulated over the past 15 years, the outcomes are not as rosy as activists first claimed. Some of the early studies—for example, of the Downtown Emergency Service Center facility in Seattle—omitted construction and capital costs of Housing First units, which artificially inflated the claimed “savings.”REF Furthermore, as University of Alabama at Birmingham professor Stefan Kertesz demonstrates, the potential savings of Housing First programs would “not apply to the 82% of homeless individuals who are not chronically homeless” (emphasis added), which indicates that Housing First has significant diminishing returns for less severe cases.REF A recent metanalysis of Housing First programs lays it out plainly: In the most rigorous studies, Housing First increases overall costs—the precise opposite of what Housing First activists have claimed.REF

However, the real problem with Housing First is not that it increases costs, but that it does not help human beings. In study after study, residents of Housing First programs demonstrate reasonably high rates of housing retention, and consistently do not demonstrate any improvement in overcoming substance abuse, reduced psychiatric symptoms, or improved general well-being—the “human outcomes.” For example, in HUD’s foundational study of three nationally recognized Housing First programs, there was no significant reduction in impairment related to substance abuse; in fact, among residents with co-occurring substance abuse and mental health disorders, the percentage of people with moderate to severe impairment actually increased from 88 percent to 93 percent after entering permanent supportive housing.REF Even in the “gold standard” Pathways to Housing program, which provides residents with 24-hour access to “nine-person interdisciplinary teams consist[ing] of social workers, a substance abuse specialist, nurse practitioner, part-time psychiatrist, family systems specialist, wellness specialist, employment specialist, and administrative assistant,” the number of individuals with impairment related to substance abuse increased over a 12-month period; and none of the individuals with substance-abuse disorders achieved recovery.REF

The most devastating evidence against the Housing First approach, which is used in many Western countries, is a recent control group study in Ottawa, Canada, which compared outcomes between a Housing First population and a nonintervention control group. Although the Housing First group showed higher rates of housing retention after 24 months, it also showed higher rates of substance abuse, mental illness, and death than the control group—which consisted of people who were simply left on the streets.REF The Housing First cohort was provided with free housing, medical care, case management, substance abuse treatment, and mental health services. And yet, the control group, which was not provided anything, reported lower rates of substance abuse and psychiatric symptoms, and higher rates of family connectedness and overall quality of life after 24 months.REF

How is this possible? One plausible explanation is that Housing First, following the “harm reduction” model, does not require residents to participate in services, take medication, or reduce substance abuse. As a result, many Housing First programs end up concentrating the worst aspects of street homelessness—drug abuse, psychosis, and social pathology—into physical housing complexes. This can create a vicious cycle and, tragically, make it nearly impossible for the homeless to overcome their personal challenges.REF

In the end, the tragic flaw of Housing First is that it reduces human beings to housing statistics. The only metric of success for Housing First is “housing retention”—if people are indoors, it is a success, even if they fall deeper into addiction, psychosis, and despair. In some studies, even overdose deaths in permanent supportive housing units do not count as a negative outcome; they simply reduce the denominator for analysis, which, perversely, increases the apparent success of the program.REF From this perspective, Housing First is a profoundly pessimistic philosophy, prioritizing physical shelter spaces above the people living inside them. Housing First policies disconnect the critical link between compassion and responsibility, leaving the homeless to cope in a network of human pathologies, under shelter that does nothing to address them.

On Los Angeles’ Skid Row and in San Francisco’s Tenderloin district, the conditions surrounding many Housing First facilities are reminiscent of the Third World.REF There are tents on the sidewalks and open-air drug markets on the street corners. Many Housing First residents spend their days on the streets to be “closer to the action,” then cycle back to their Housing First units when they have finished a cycle of drug taking. One HUD report found that 41 percent of Housing First residents in the study disappeared for weeks or months at a time, choosing to live on the streets despite having access to secure housing.REF Tragically, Housing First is often little more than homelessness within a residential setting—it does nothing to reduce the human pathologies; it merely contains them temporarily within four walls.

“Treatment First” as a Successful Alternative

Fortunately, there is an emerging body of research that supports an alternative policy: the so-called Treatment First approach. While Housing First prioritizes the values of “harm reduction” and low barriers to entry, the Treatment First model prioritizes the values of addiction recovery, personal transformation, and self-sufficiency. In Treatment First programs, the goal is to rehabilitate the individual, then secure permanent housing. It is also called the “linear” model, because it relies on a guided progression through recovery programs, building human capacity and treating addiction and mental illness.

Researchers at the University of Alabama at Birmingham completed a series of four randomized controlled studies of Treatment First programs.REF The six-month programs—collectively dubbed the “Birmingham Model”—provided the homeless with abstinence-contingent housing and required participation in a rigorous full-time program of addiction recovery, behavioral treatment, work training, and recreational opportunities. In the final clinical trial,REF 64 percent of residents maintained sobriety at six months after entering the program, and “housing stability and employment rose from baseline to six or twelve months in all trials.”REF At the end of the study, the researchers concluded that programs that combine abstinence-contingent housing with drug treatment could “contribute to long-term housing and employment for as many as 40%” of homeless drug users who enter such a program.REF

It is important to keep in mind that the programs of the Birmingham Model served a population with numerous challenges: The majority of clients were homeless, crack-addicted, mentally ill men in one of the poorest regions in the United States. In the most successful trial program, 44 percent of men were stably housed and 53 percent were stably employed after 12 months—a remarkable achievement given the severe social, personal, and financial obstacles at the outset of the program.REF The key to success, according to researchers, was the combination of housing and treatment with obligations and enforcement; in other words, the relinkage of compassion and responsibility. The Birmingham Model programs consciously sought to align incentives with the “restoration of behavioral self-regulation and the capacity to interact in a constructive social environment,” leading to long-term self-sufficiency, restored relationships, and improved human outcomes.REF

This is not to say that the Birmingham Model is a panacea. As with any policy choice, there are inevitable trade-offs between Housing First and Treatment First. In simple terms, the evidence presents these conclusions: Housing First has housing-retention rates of 80 percent, but does not improve substance abuse, mental health, or employment outcomes.REF Treatment First has housing-retention rates of 40 percent, but significantly improves substance abuse, mental health, and employment outcomes—and moves many people into self-sufficiency and private-market housing.REF

Reverend Andy Bales, CEO of Union Rescue Mission in Los Angeles, which offers faith-based treatment programs for the chronically homeless, explains the different objectives between Housing First and comprehensive treatment programs: “Housing First [sometimes leads to] a life of addiction and actually overdosing in that unit…. We want more than that for our graduates. We want them to have a recovered life, a productive life, a life more abundant than it was when they entered our facility.”REF

While there is still a need for permanent supportive housing for the severely disabled and chronically homeless, the vast majority of the homeless would be better served in treatment and recovery programs that promote self-sufficiency. Furthermore, from a financial standpoint, policymakers must keep in mind that permanent supportive housing functions as an annuity—it is an expenditure for the duration of an individual’s lifetime—while recovery programs seek to achieve positive results in nine months to 18 months, then accommodate new residents. In short, recovery achieves better human outcomes and can serve a much larger number of people, given limited resources. Treatment First comes closer to fulfilling the real purpose of public assistance: to enable human flourishing, even for those who face the greatest challenges.

Regulatory Reform for HUD’s Homelessness Programs

For the first time in a generation, there is the political momentum to dramatically re-orient federal policy toward better outcomes on homelessness. Although the Housing First philosophy has become deeply entrenched among a coalition of service providers, political activists, and organized interest groups in both the U.S. and Canada, the evidence no longer supports their blanket policy of Housing First programs. Through concrete reforms, HUD can redirect existing resources away from Housing First and to programs that demonstrate a better record of improving lives, especially with regard to mental illness, drug and alcohol addiction, and other social pathologies.

Rather than continue to be guided by the ideological commitment to Housing First, policymakers must take this moment to re-evaluate the approach to homelessness from the bottom up. The intentions behind Housing First might be noble, but in practice, it cannot be scaled up to meet the needs of the hundreds of thousands currently sleeping on the streets—and will do little to transform their lives. Instead of doubling down on Housing First, HUD should shift CoC funding to Treatment First programs, such as those operated by rescue missions, rehabilitation centers, and recovery homes.

As a starting point for reform, President Trump recently issued a series of executive orders that provide guidance for improving HUD’s homelessness programs, among other public assistance programs. In Executive Order 13828, President Donald Trump recognized that the welfare system, which would include many Housing First programs, “still traps many recipients…in poverty and is in need of further reform and modernization in order to increase self-sufficiency, well-being, and economic mobility.”REF In Executive Order 13831, the President recognized that faith-based organizations, such as rescue missions and rehabilitation programs, “lift people up, keep families strong, and solve problems at the local level,” and recommended that the federal government “partner with such organizations through innovative, measurable, and outcome-driven initiatives.”REF

Following these principles, HUD policymakers should immediately remove the incentives for “low-barrier” programs and orient its funding toward outcomes of recovery, rehabilitation, and self-sufficiency. The Homeless Emergency Assistance and Rapid Transition to Housing (HEARTH) Act of 2009, which amended the McKinney–Vento Act, explicitly states that HUD should seek a “balance among strategies targeting homeless individuals, families, and other subpopulations.”REF For the past decade, HUD’s approach has been oriented too far toward Housing First and permanent supportive housing; any new funding must rebalance HUD’s overall portfolio toward Treatment First and transitional housing.

In order to begin these reforms, HUD Secretary Ben Carson should immediately declare Treatment First a “proven strategy” based on the evidence from evaluations and submit this finding through the notice-and-comment process. Second, Secretary Carson should submit a finding that HUD has funded at least 150,000 new units of permanent housing, which under Section 428(a)(5) of the HEARTH Act, will terminate the requirement for a 30 percent minimum allocation for permanent housing.REF Taken together, these actions will establish the regulatory foundation for shifting funds from Housing First programs to Treatment First programs.

Finally, under the Secretary’s leadership, HUD officials should conduct a targeted overhaul of the funding formula in next year’s CoC Notice of Funding Availability (NOFA), which allocates competitive grant funding to local CoC and nonprofit service providers. The Secretary, in conjunction with HUD staff, should remove all disincentives for recovery and self-sufficiency in the current formula. Specifically, HUD should amend the current NOFA scoring system as follows:

  • Eliminate the reward for CoC programs that demonstrate that “at least 75 percent of all project applications…provide low barriers to entry without preconditions,”REF which, in effect, penalize recovery and faith-based programs;
  • Eliminate the reward for projects that have “low barriers to entry and [prioritize] rapid placement and stabilization in permanent housing,”REF which de-prioritize treatment-based programs;
  • Eliminate the reward for demonstrating “an increase in income from non-employment cash sources,”REF which may incentivize work-capable individuals to permanently exit the workforce through disability programs; and
  • Eliminate the reward for “specific strategies to ensure homelessness is not criminalized,”REF which undercut local governments’ enforcement of municipal ordinances for public safety and order.

To replace these points, HUD should add new rewards for programs that lead to self-sufficiency, as described in President Trump’s Executive Order 13828. HUD should design its competitive grant process into a pay-for-outcomes system that rewards service providers that deliver tangible human outcomes. Specifically, HUD should create point incentives in its upcoming NOFA for:

  • Treatment-based programs that demonstrate improvements in substance abuse, mental health, and physical health outcomes;
  • Programs that demonstrate an increase in employment, earned income, and financial independence; and
  • Programs that successfully move the homeless into long-term private housing.


With national attention drawn to the problems associated with public encampments and a record $9 billion in CARES Act funding, policymakers have a critical opportunity to adjust the national strategic approach to reducing homelessness.

As legislators and HUD officials contemplate their response to the problem, they should take inspiration from the origins of federal homelessness policy—the McKinney–Vento Act. When President Reagan signed the bill into law, he made clear that federal homeless assistance was deeply connected to the mission of providing support to “public or private nonprofit organizations for health and substance abuse services” and to “assist individuals in developing marketable skills that bring economic independence.”REF

This should still be the guiding light of homelessness policy: to improve human lives meaningfully. The most urgent challenge for policymakers is not to build new apartment units, but to rehabilitate the individuals who will live inside them.

Christopher F. Rufo is Visiting Fellow in Domestic Policy Studies, of the Institute for Family, Community and Opportunity, at The Heritage Foundation, and the Director of the Center on Wealth & Poverty at Discovery Institute.


Christopher Rufo
Christopher Rufo

Former Visiting Fellow for Domestic Policy Studies