It has been a jarring few years for leaders working to end homelessness.
Advocates and researchers have never had stronger evidence about the best way to most effectively house people who need it: a model known as “housing first.” As the name suggests, its focus is getting people into permanent housing and offering them support services, rather than requiring them to address mental health conditions, substance abuse, or job training first.
The housing-first model has enjoyed strong bipartisan support, and growing evidence about its effectiveness, for nearly three decades. But now it’s facing challenges on several fronts. Rising rents and a chronic shortage of affordable housing has meant it’s grown ever harder to find units for homeless people. The National Low Income Housing Coalition estimates a shortage of 7 million homes for the poorest renters in the US. Unreliable and shrinking budgets for social services have also meant people given housing are not always offered the support they’d need to really thrive under the housing-first model.
Homelessness on the streets and in shelters steadily declined throughout President Barack Obama’s second term, but the numbers began to tick back upward starting in 2017. In some high-cost cities like San Francisco, Austin, and Seattle, the highly visible crisis began sparking concern and political backlash.
Now “housing first” itself is experiencing new and unprecedented politicization. The model has been embraced by Congress, HUD, housing researchers, and national homeless organizations, but has faced growing criticism in the past few years.
The Republican-led backlash is leading to new punitive approaches, including in liberal cities. Last month, New York City Mayor Eric Adams announced a controversial new plan that could allow homeless individuals with mental illness to be involuntarily hospitalized. In April, the San Francisco Chronicle published an investigation into some of the dilapidated hotels where the city housed the formerly homeless. “This is Housing First policy in action,” declared a local journalist in a conservative policy magazine.
Some lawmakers have made clear that their new policies for homelessness should be seen as a rebuke to the housing-first model. A new Missouri law criminalizing sleeping outside on state-owned land was modeled on a template crafted by the Cicero Institute, an Austin-based conservative think tank founded in 2016 that’s opposed to housing-first policies. Stateline found nine bills introduced in six states in the past two years based on the Cicero template, but Missouri’s marked the first to pass that includes language preventing state and federal homeless dollars to be used on permanent housing.
“It’s easy to say there’s more people experiencing homelessness so ‘housing first’ must have failed, but the intervention isn’t causing homelessness,” said Ann Oliva, the CEO at the National Alliance to End Homelessness, who previously spent a decade working at HUD. “More people are struggling to afford and find housing, and homeless assistance services can’t keep up with the need.”
Still, the housing-first approach has notched successes — most notably in Houston, Texas, a city that has maintained a dedicated political commitment to housing first over the past decade. More than 25,000 people in Houston have been moved into housing over the past 10 years, yielding a remarkable 63 percent drop in homelessness since 2011.
“It’s challenging work every day, there’s a thousand barriers to overcome, but we have to do it because there is no other choice,” said Marc Eichenbaum, the special assistant to Houston’s mayor for homeless initiatives.
When Sam Tsemberis, the founder and CEO of the national organization Pathways to Housing, pioneered the housing-first model in New York City in 1992, it was a sharp departure from the previous consensus on homelessness policy.
Under the old approach, known as “housing readiness” or “treatment first,” people had to meet certain goals of stability and independence, like achieving sobriety or landing a job, to earn access to permanent housing. Housing first, by contrast, saw housing as integral to recovery, not a reward for achieving it.
In 2000, Tsemberis and his colleagues published results comparing Pathways clients with those housed through the “housing readiness” approach. Over a five-year period, 88 percent of Pathways clients remained housed compared to 47 percent of the control group.
A 2004 randomized controlled trial, the gold standard of social science research, backed up their findings. Among 225 homeless people with mental illness, participants assigned to housing-first programs obtained housing earlier and remained more stably housed compared to the control group. Another such trial, published in 2007, found that among 260 participants experiencing homelessness in New York City suburbs, those assigned to housing-first programs were more likely to maintain permanent independent housing than the control group over a four-year period. Other studies found the model was not only effective but was saving cities money.
The remarkably robust body of research has changed how policymakers and experts in the US and around the world view homelessness.
George W. Bush’s homelessness czar, Philip Mangano, promoted housing-first models around the country. Salt Lake City, Utah, was among the first large cities in the US to embrace the model in 2005, and by 2014 it was featured in numerous publications for its success in reducing homelessness.
International leaders took notice. In 2008, Canada launched the world’s largest randomized controlled trial of housing-first programs in five cities, following more than 2,000 individuals with serious mental illness over two years. The results, published in 2014, found the housing-first model was remarkably successful compared with Canada’s traditional “treatment first” approach to homelessness. Over the past decade, studies in several European countries have found similarly positive outcomes for participants. The most empirically rigorous study tracked hundreds of individuals in four French cities between 2011 and 2016, and found the randomized housing-first participants were more stably housed, spent significantly fewer days hospitalized, and saved cities money compared to those in traditional homeless treatment programs.
The evidence has found housing first works better than other approaches for all kinds of unhoused people — from individuals who are chronically homeless and experiencing serious mental health conditions to families in emergency shelters and suffering primarily from economic hardship.
Earlier this year, researchers in California announced results from a seven-year randomized trial of chronically homeless individuals in Santa Clara. The study found those in the housing-first group spent 90 percent of their nights housed on average since the study began, and made less use of psychiatric emergency services and more use of outpatient mental health services compared to the control group. “The experiment intentionally sought to try housing first for the very most complicated patients — those who society says are most hard to house — and it worked,” said study co-author Margot Kushel, who directs UCSF’s Benioff Homelessness and Housing Initiative.
But in recent years, conservative think tanks like the Manhattan Institute and the Heritage Foundation began ramping up their criticisms of housing first. Although Trump’s HUD Secretary Ben Carson praised the model publicly on multiple occasions, Trump’s Council of Economic Advisers released a report on homelessness in 2019 casting doubt on its effectiveness.
The Trump White House then appointed Robert Marbut, a longtime critic of housing first, to lead the US Interagency Council on Homelessness; the agency later published a report focused on the “drawbacks” and “concerning results” of housing first, and encouraged leaders to reconsider requiring sobriety and other treatments in exchange for housing assistance. (Twelve national homelessness organizations called the report “ineffective and dishonest.”)
President Joe Biden has renewed the federal government’s commitment to housing first, campaigning on the model and making it a priority for his administration with funds from the American Rescue Plan. His team has also emphasized eviction prevention, a typically ignored cause of homelessness.
Some conservatives are now drawing attention to a study published in 2021 looking at 73 chronically unsheltered individuals in Boston over 14 years. While 82 percent of program participants receiving permanent supportive housing were still housed after one year, only 36 percent still had housing after five years, and only 12 percent after 10 years. Almost half of the participants died while housed, including from conditions like cirrhosis, heart disease, and cancer.
“The human wreckage wrought by Housing First was revealed last year in Boston,” conservative commentators proclaimed this summer in The Hill.
Boston study subjects were offered fewer support services than is endorsed under the Pathways model, such as a multidisciplinary team that’s available 24/7 with full wraparound services. Study co-author Jill Roncarati told me their findings should not be used as an argument against housing first. “We firmly believe everyone should have housing, and a continuum of housing, where individuals can enter independent living options and then have access to whatever else they need in support,” she said.
“Giving folks the keys to affordable housing and leaving [them alone] will work for most families, and quite likely for individuals without serious psychiatric and medical problems … but people with psychiatric and medical problems need more,” added Beth Shinn, a Vanderbilt professor and leading national researcher on homelessness.
The housing-first model calls for providing individuals with permanent housing, but it doesn’t claim that housing alone is enough. Regular check-ins by trained case managers are required, as are making social and medical supports readily available. Having the services be voluntary, experts say, preserves an individual’s sense of autonomy, leading generally to higher uptake.
Perhaps unsurprisingly, some cities say they’re implementing housing-first policies, but leaders then lack the funding, services, or available housing units to make the model work sustainably and well.
Likewise, HUD, though it has embraced the housing-first model, has often deemphasized the importance of follow-up services that Pathways to Housing stresses.
“The housing-first language can be so sloppy, but the data suggests that fidelity to Sam [Tsemberis]’s model matters,” said Shinn. In the Canadian study, for example, researchers found that participants in programs that adhered more closely to the Pathways model had better outcomes in terms of housing stability, quality of life, and community integration than those in programs with looser definitions of housing first.
HUD has shifted away from spending its budget on supportive services, including health care, drug treatment, and education. As of 2021, according to a HUD spokesperson, 22 percent of HUD’s competitive homeless assistance funds went to supportive services, compared with 68 percent for housing. (In 1998, 55 percent of HUD’s budget was spent on supportive services and 45 percent was awarded for housing). The agency’s hope is that other local, state, and federal programs — like Medicaid — can fund supportive care.
This vision has proved effective for helping houseless veterans. In 2008, a housing first program began combining housing vouchers for veterans provided by HUD with case management and clinical services provided by the VA. Experts agree that the program has been successful, with homelessness among veterans declining some 55 percent since 2010.
But for unhoused people who aren’t veterans, consistent and reliable social services have been harder to guarantee. “Because the VA is a full-service health care provider, it was relatively easy to maintain” those supportive services, said Oliva, of the National Alliance to End Homelessness. “We haven’t really been able to systematically create that same level of connection nationwide with other types of mainstream systems.”
And while social services are voluntary for housing-first participants, Tsemberis told me that many mistakenly assume that no follow-up is then required for individuals after getting housing unless the tenant wants it. “The home visit is not optional, whether the person is agreeing to treatment or not,” Tsemberis said of his Pathways model. A trained case manager is supposed to regularly check in, help a person manage their life and apartment, and encourage them to consider health, mental health, and addiction support.
Roncarati, the Boston study co-author, said she agrees their results suggest more intensive supports are needed to help certain sub-populations of people experiencing homelessness. “The housing-first models that have been replicated are probably underfunded and there needs to be more support and different types of housing available,” she told me, though noted some acutely struggling individuals did “extremely well” in independent living, and researchers were “sometimes very pleasantly surprised.”
Houston, Texas, has stood out in the United States for its dedicated commitment to implementing housing-first policies, earning positive national media coverage this year in the New York Times Magazine, the Los Angeles Times, and Smart Cities Dive, among others. Earlier this month, the housing advocacy group California YIMBY published a report heralding Houston’s housing-first experiment, arguing California has not been able to replicate it primarily because Houston has more abundant housing. The group praised Houston’s land use policies — including its lack of a traditional zoning code — for substantially increasing Houston’s housing supply and lowering its costs.
Leaders in Houston agree their housing supply has helped them over the last decade, but cautioned against seeing their city as some housing utopia. Much of the credit, they say, goes to the slow, dogged work of earning trust from private-sector landlords, having a strong mayor system that remained all-in on housing first, and strategically leveraging federal dollars, including from the seven federally declared disasters the city has had in the last seven years. Houston puts no local general operating funds into homelessness receives scant funds for it from its state legislature, and Texas has not expanded Medicaid.
“The theory that Houston’s success at reducing homelessness is because of its lack of zoning is a red herring,” Eichenbaum, the special assistant to Houston’s mayor for homeless initiatives, told me. “The reality is while we might not have the typical zoning that many cities have, we do zone through ordinance and the hardest piece is still siting a location. We still have to deal with NIMBYism.”
Houston began its housing-first efforts in 2011, and at the time, the city had a higher apartment vacancy rate which helped leaders more easily move unhoused individuals into open units. But the secret wasn’t that Houston’s housing code allowed them to build new units for homeless individuals easily; it was the large supply of existing apartments considered moderately priced for vouchers.
“Even without zoning there can be a lot of backlash, and the neighborhoods can still prevent new housing,” said Ana Rausch, the vice president of program operations at Coalition for the Homeless of Houston/Harris County.
Things have gotten harder in Houston since the pandemic. Average apartment rents across the city have increased 12 percent since 2019, according to the real estate firm CoStar, and within older buildings, rents have jumped nearly 14 percent. “In many Houston neighborhoods, the days of a one-bedroom for $1,000 are long gone,” reported the Houston Chronicle in September.
Homeless advocates say they’re now something of a victim of their own success; Houston leaders were so successful at moving individuals experiencing homelessness into apartments that more apartment complexes now have low vacancy rates, making them newly attractive to investors. Some of these complexes have been purchased by new owners who now have little interest in renting to homeless people.
One asset Houston has is a strong mayor system, and two successive mayors who’ve maintained consistent support for the housing-first strategy. This makes housing first easier to coordinate than in a place like Los Angeles, for example, where 15 elected city council members all compete for power with each other and the city’s mayor.
But Houston Mayor Sylvester Turner is serving his final year now of his term-limited tenure, and advocates say they don’t know what their next mayor will want to do or what will happen when federal pandemic money dries up.
Though Houston’s future strategy is unclear, researchers agree that the city’s land use policies merit broader attention now. Increasing housing supply and helping people afford housing will be key to any successful effort to end homelessness. Zillow economists reported in 2018 that communities can expect a more rapid increase in homelessness in areas where people spend more than 32 percent of their income on rent. In 2020, the US Government Accountability Office found that every $100 increase in median rent is associated with a 9 percent increase in the estimated homelessness rate.
“It’s pretty clear that exclusionary zoning drives up housing costs, and if you have higher housing costs you’re going to have more homelessness,” said Shinn. “If we’re going to reduce homelessness, that means making housing affordable, and there’s a lot of things that go into housing affordability, and one of them is zoning.”