Reimagining the Policy Approach to Homelessness
Identifying the several causes of homelessness and applying solutions that fit each one
Executive summary
A person walking down the streets of any major American city — or one of many smaller ones—is likely to encounter people struggling with homelessness. The homelessness crisis erodes community wellbeing, causes economic hardships, and reduces the quality of life for communities. More profoundly, homelessness is also a human tragedy for the homeless and their families.
Localities, states, and the federal government have all been searching for solutions to homelessness. Working with many state and local governments, the Biden administration has pursued a “Housing First” strategy. Housing First policy assumes that providing permanent housing to the homeless without preconditions is the best way to reduce homelessness.
Meanwhile, many localities are over-reliant on aggressive law enforcement tactics to force people out of targeted public areas without addressing the underlying causes of homelessness or providing alternative resources.
Unfortunately, the evidence suggests that neither of these approaches has worked. This is because the homeless population is highly heterogeneous. Some people are simply poor and unable to afford a place to live. Many others are homeless due to psychiatric and/or substance abuse disorders. Yet others are homeless for reasons that do not fit into either of these categories.
In this paper, we seek to dramatically improve America’s approach to homelessness, by ending the one-size-fits-all ideology of homelessness policy and by developing discrete strategies to address different subpopulations of homeless people.
We suggest five strategies:
- Institutional care. Enact stricter conservatorship laws to require homeless people who are unable to care for themselves to seek shelter or treatment, including involuntary institutionalization where necessary, while also protecting their autonomy and civil liberties.
- Invest in treatment. More money should be directed to shelters, mental health programs, and substance abuse services that lack the resources to provide for local populations.
- Build more housing. Reduce housing costs by removing barriers to new construction such as exclusionary zoning rules so local inventory can meet demand.
- Enable state and local experimentation. Give local communities broad discretion to experiment with solutions that best meet their unique circumstances and resources.
- Broader social mobility policies. Evaluate efforts to deal with homelessness in the broad context of reducing poverty.
Homelessness is a serious problem that hurts both the people experiencing homelessness and negatively affects the surrounding community. Fortunately, proven solutions can make a real difference.
Introduction
Homelessness is not a new phenomenon in the United States. The earliest accounts in North America date to the 1640s. Today, homelessness is growing and imposing significant costs on states and localities.
According to government estimates, at least 653,000 Americans were experiencing homelessness in January 2023: the highest number since the government began collecting data in 2007, when there were 647,000. Of those who are homeless at a given point in time, roughly 143,000 could be considered chronically homeless, having been homeless for a year or longer.
These point-in-time estimates may understate the actual size of the problem. More than one million Americans may experience homelessness at some point during any given year. Millions more live in fragile housing situations in which they could easily fall into homelessness if they suffered an interruption in their income or some other crisis.
Perhaps more telling, a significant increase in unsheltered homelessness—i.e., those not living in shelters or other temporary arrangements—has offset a decline in the number of sheltered homeless people. This is one reason that homelessness is so much more visible in recent years than in the past.
While a growing problem everywhere, homelessness is not equally distributed. As has been the trend for many years, five states — California, New York, Florida, Washington, and Oregon — account for more than 55 percent of all people experiencing homelessness. California also has the highest percentage of unsheltered homeless, with more than two-thirds of its homeless population living on the street. Oregon, Hawaii, Arizona, and Nevada also have more than half of their homeless populations unsheltered. In contrast, Vermont, New York, Maine, Massachusetts, and Wisconsin shelter more than 90 percent of their homeless residents.
New York has seen the largest increase in homelessness both over the past year (up 39 percent) and since 2007 (up 65 percent). Other states with significant increases over the past year include Colorado, Florida, Massachusetts, and, of course, California. Still, while this handful of states continues to be most affected by homelessness, 41 states saw an increase in homelessness from 2022–23. Major cities, notably New York, Los Angeles, Seattle, San Diego, and Denver have the largest homeless problems, but the homeless populations are also growing in the suburbs and rural areas as well.
African Americans and Latinos are disproportionately likely to be homeless. Roughly 37 percent of American homeless are African American, and 24 percent are Latino. Men are more likely than women to be homeless and also far more likely to be unsheltered, although the proportion of homeless women is rising.
Families comprise roughly 37 percent of the homeless population, but most are sheltered. Individuals over 50 now make up nearly half of the homeless population, and are the fastest growing homeless group. LGBTQ individuals, particularly LGBTQ youth, also make up a disproportionate segment of the homeless population.
It is also important to note that most homeless people are U.S. citizens, and are living in the same city or state where they lived before they lost their housing. However, there is more mobility among the homeless population than some might think. In San Francisco, for instance, 30 percent became homeless outside the city and another 17 percent moved there a year before they became homeless. In Austin and Los Angeles, about a third of their homeless residents moved from elsewhere. One study in Seattle found a majority had come from outside the city. However, even those from “outside” the city tend to be from neighboring communities, not from across the country.
Of course, all these statistics can vary significantly from state to state or city to city.
So why are people homeless?
Not all people suffer homelessness for the same reasons. Solving homelessness will thus require different solutions for different populations.
Mental illness and substance abuse
The homeless have higher rates of both substance abuse and mental health problems than does the general population. According to 2018 point-in-time counts, roughly 20 percent of homeless people have a severe mental illness and 16 percent are chronic substance abusers. Other surveys put the numbers slightly higher, with some suggesting that between 25 and 40 percent of homeless individuals has a substance use disorder. Because most of those studies exclude families — who are far less likely to have such problems — the affected persentage is likely inflated.
There is also something of a chicken-and-egg problem with these surveys. Many homeless people begin to abuse drugs or alcohol due to the stress of living on the street. And homelessness tends to exacerbate existing substance use disorders. Thus, substance abuse is often both a cause and effect of homelessness.
Indeed, the degree to which substance abuse problems contribute to homelessness is uncertain. A large-scale survey of California’s homeless population found that just 13 percent of participants said that substance abuse was a reason for losing their last housing. However, this almost certainly underestimates the impact since the survey also shows that roughly a quarter reported that their substance abuse led to health, social, or legal problems in the six months prior to homelessness, issues which may have contributed to their becoming homeless. Sheltered individuals are less likely to suffer substance abuse problems than unsheltered homeless. Moreover, statistics that combine sheltered and unsheltered homeless populations may artificially dilute the problem of substance abuse, since sheltered homeless and families
Even so, research suggests that rates of substance abuse explain only about six percent of the variation in rates of homelessness between states.
There is no clear relationship between rates of mental illness in a state and levels of homelessness. Still, mental illness can contribute to homelessness. The homeless population is also more likely to struggle with mental illness than the population at large. Several studies estimate that 20–25 percent of homeless people suffer from a serious mental health problem, compared to about five percent of the general population. As with substance abuse, however, rates of mental illness are higher for unsheltered homeless individuals than for the sheltered homeless or families.
In many ways, the country continues to deal with the downstream effects of the deinstitutionalization of the 1960s, 70s, and 80s. Although deinstitutionalization is generally regarded as a failure today — at least as carried out — it had broad support both across the political spectrum and among health care professionals at the time.
Two important societal and political forces drove deinstitutionalization. First, a series of exposés revealed horrific abuses at many mental health institutions. As early as 1946, a Life magazine article, “Bedlam 1946,” showed photographs of patients huddled naked in the corridors of filthy, overcrowded facilities or strapped to beds, chairs, and benches. In the 1960s, Senator Robert Kennedy visited New York’s infamous Willowbrook State School and found patients “living in filth and dirt, their clothing in rags, in rooms less comfortable and cheerful than the cages in which we put animals in a zoo.” That institution returned to the headlines in 1972 when Geraldo Rivera, then a young reporter for WABC-TV, documented numerous cases of physical and sexual abuse at Willowbrook. These stories, among others, helped turn public opinion against institutionalization.
Second, new classes of antipsychotic medications were seen as providing an alternative to institutionalized care. As early as 1954, the FDA had approved chlorpromazine, better known as Thorazine, for the treatment of schizophrenia and bipolar disorder. Hope grew that patients could be treated pharmaceutically on an outpatient basis.
The move toward deinstitutionalization was reinforced by a series of court cases establishing civil rights for the mentally ill and parameters for state intervention. In the 1966 case Lake v. Cameron, for example, the U.S. Court of Appeals for the District of Columbia established the concept of “least restrictive setting.” This means a setting with “a minimum of limitation, intrusion, disruption, and departure from commonly accepted patterns of living.” In the 1975 case O’Connor v. Donaldson, the U.S. Supreme Court held that involuntary commitment was only permissible in cases where an individual was a clear danger to themselves or others. And, in Olmstead v L.C. in 1999, the Court held that mental illness was a disability under the Americans with Disabilities Act, and thus government agencies were required to make “reasonable accommodations” for the mentally ill, including moving them to community-based treatment where possible.
Perhaps the most significant state-based deinstitutionalization legislation, California’s Lanterman-Petris-Short Act, was signed into law by then-Governor Ronald Reagan in 1967. It sought to “end the inappropriate, indefinite, and involuntary commitment of persons with mental health disorders,” by limiting involuntary commitments to a maximum of 14 days.
Deinstitutionalization was supposed to ensure that people who suffer from serious mental illness would continue to receive care, but in a safer, local, and less restrictive setting. Community health centers would replace large institutions. Care would shift from warehousing the mentally ill to treating their individual problems. However, that is not what happened.
Deinstitutionalization precipitively drove down the number of patients being treated in mental health hospitals, from a high of nearly 500,000 in the 1950s to under 100,000 by the mid-1980s. Today only about 50,000 people in the United States primarily reside in mental health hospitals. But without treatment, and with nowhere to go, the newly deinstitutionalized mentally ill ended up either in jail or on the street. According to the Department of Justice, roughly half of people incarcerated in jails today, and nearly a third of those in prison, suffer from at least one mental illness.
Reformers sought to build community mental health centers to replace the asylums. In 1963, President John F. Kennedy signed the Community Mental Health Act, under which the federal government would fund at least 1,500 community-based treatment and prevention centers. However, fewer than half that number were actually built. The advent of Medicaid in 1965 gave states a powerful fiscal incentive to close state-funded mental hospitals and shift patients to voluntary facilities that were at least partially paid for with federal funds. In addition, the construction of community centers faced local resistance. In an effort to address these problems, the Mental Health System Act of 1980 doubled federal funding for these community-based mental health centers. Under the Omnibus Budget Reconciliation Act of 1981, that funding was consolidated into state-based block grants. Today, there are more than 2,500 community mental health centers and nearly 5,000 additional outpatient mental health facilities, serving 3.6 million individuals. But, by their very nature, voluntary facilities are only effective when patients are highly motivated to comply with their treatment regimens. This is inherently challenging for those harboring mental illness and substance abuse.
Pharmaceutical treatment also fell short of its lofty promise. Many of the most prevalent treatments turned out to have side effects. More significantly, outside of institutions or community-based alternatives, many mentally ill individuals simply stopped taking their medications.
The legacy of deinstitutionalization continues to complicate the response to the homeless mentally ill. Many homelessness advocates remain hostile to institutionalization. And even if we were to return to an era of mass institutionalization, it would be expensive. From 1970 to 2018, there was an 84 percent reduction in state hospital beds. Of those beds that remain, roughly half are being used by individuals who have been charged with or convicted of crimes. Inpatient psychiatric hospitalization is estimated to cost nearly $280,000 per patient per year.
Many mentally ill individuals can be treated in non-institutional settings. Yet, for those who require institutionalization, policymakers have failed to develop effective alternatives. As a result, communities are now left without any effective mechanism for dealing with mentally ill individuals who are unable to care for themselves.
Poverty and housing costs
Although individuals who suffer from substance abuse or addiction and other mental illnesses are often the most conspicuous in American cities, people with mental illness and substance use disorders make up less than half of American homeless.
A 2021 University of Chicago study estimated that 53 percent of people living in homeless shelters and 40 percent of unsheltered people were employed, either full or part-time. We lack good data on how many hours they actually work — many may work just enough to produce a few hours on a W-2 — nor do we have a good breakdown on differences between housed and unhoused homeless. Still, this suggests that most homeless people were capable of making reasonable decisions, caring for themselves, and otherwise functioning in society. Some factor beyond mental illness or substance abuse was, therefore, likely to have led to their becoming homeless. The evidence overwhelmingly points to housing costs and the interaction of high housing costs with poverty as the driving factors for a large portion of homelessness.
At least as far back as 2001, James Quigley and Stephen Raphael were pointing out that “modest changes in housing market conditions can have substantial effects on the incidence of homelessness.”
In the survey of California’s homeless population, low incomes and high housing costs were the primary elements leading to homelessness for much of that state’s affected population. According to the report, 22 percent of the homeless blamed a loss of income for loss of their housing, while 10 percent cited an increase in non-housing costs (such as food or medicine), and 8 percent said there had been a rent increase that they could not pay. “Participants living on the economic margin, with high housing costs, low incomes, and little savings, had little margin for error. Loss of income or decrease in work propelled many living on the economic margins into homelessness.”
Numerous studies have shown a relationship between rising housing costs and the rate of homelessness. For instance, a 2017 study by Kevin Corinth of the American Enterprise Institute found that a one percent increase in median rent was associated with a one percent increase in the rate of homelessness. Similarly, Maria Hanratty of the Humphrey School at the University of Minnesota concluded that a one percent increase in median rent is associated with a rate of homelessness that is 0.9 percent higher. And an earlier study of California data by John Quigley, Steven Raphael, and Eugene Smolensky of the University of California–Berkeley likewise concluded that a one percent increase in median rents correlated with a 0.9 to 1.2 percent rise in the rate of homelessness.
The costs of homelessness
Regardless of the underlying causes, there is no doubt that homelessness is a tragedy. Living on the streets or even in shelters can be traumatic and dangerous. Homeless people can face extreme heat and cold, unsanitary conditions, and violence from a variety of sources.
The homeless population suffers from diseases like AIDS, tuberculosis and influenza, cancer, heart disease, diabetes and hypertension at rates three to six times higher than the general population. Nearly 8,000 individuals — nearly 20 per day — died in 2020 while living on the streets. Some of those deaths were likely due to COVID, but the number of preventable deaths among homeless people has been rising for several years.
At the same time, homelessness can impose significant costs on communities, financial and otherwise. California spends $7.2 billion annually fighting homelessness. New York City alone spends $2.2 billion per year on homelessness. And, as discussed below, the Biden administration is asking for $8.7 billion in federal spending per year. On average, according to the National Alliance to End Homelessness, each homeless person costs taxpayers roughly $35,000 per year. Other studies put the annual cost as high as $100,000 per homeless person. Most likely the real cost lies somewhere in between.
These figures represent only direct spending on homelessness. The health disparities mentioned above mean that homeless people are more likely to need medical treatment and access hospital emergency rooms where care is often provided to those without ability to pay. This means that uninsured homeless people impose additional costs on the health care system.
Communities must also deal with downstream costs, including quality of life issues like crime and disorder. Homeless individuals are more likely to be the victims of crime than perpetrators, (frequently at the hands of other homeless persons) but they are also disproportionately likely to be involved in criminal activity. This is true even if you take out substance abuse and “status offenses” related to the state of being homeless. A study by the San Diego District Attorney’s Office, for example, found that homeless individuals were substantially more likely to be arrested for such crimes as robbery, burglary, assault, arson, and vandalism than the population as a whole. Recidivism was also notably high among homeless individuals who were arrested.
Concentrations of homeless people can negatively affect tourism, street-level retail businesses, and other economic activity. People are less likely to visit an area where they encounter petty crime and must avoid feces in the streets. Camping, sleeping, and aggressive panhandling discourages other citizens from using those spaces. San Francisco’s empty downtown storefronts have many causes, but the city’s enormous homeless population is certainly a contributing factor. Similarly, after the Austin City Council attempted to decriminalize public loitering, camping, and other behavior associated with the homeless, the city attracted large numbers of homeless from other jurisdictions, leading to significant problems in its downtown, and a public revolt. Ultimately, Austin voters reversed the city’s actions in a referendum.
How not to solve homelessness
The status quo of endemic urban homelessness is untenable, so policymakers at all levels of government are searching for solutions. The evidence shows, however, that most commonly suggested policies are unlikely to significantly reduce homelessness.
Bad Idea 1: ‘Housing First’
Twenty years ago, much of the response to homelessness was centered around local homeless shelters. Shelters were seen as part of a linear treatment program, in which the homeless graduate from shelters to temporary housing and then to permanent housing. Most homeless families or individuals started with emergency shelter to get them off the street. Then they could move on to transitional housing, which was usually available for up to 24 months. In addition to providing shelter, transitional housing would provide supportive services designed to help tenants find and secure permanent housing and address other issues. Generally, participants had to satisfy preconditions, such as sobriety, before they could move on to a permanent residence.
However, over time, the emphasis shifted to a concept known as “Housing First,” which emphasizes providing housing to homeless individuals without preconditions. The idea assumes that only after individuals are placed in a stable housing situation, will they be able to deal with other problems such as joblessness, substance abuse, or mental illness.
As Sara Rankin of Seattle University School of Law explains, “The Housing First approach views housing as the foundation for life improvement and enables access to permanent housing without prerequisites or conditions beyond those of a typical rent.” As she argues, “people need basic necessities like food, sleep, and a stable place to live before attending to any secondary issues, such as getting a job, budgeting properly, or attending to substance use issues.”
The natural corollary to this new approach has been a shift in emphasis from emergency shelter and transitional housing to “permanent supportive housing,” which combines low-cost (usually subsidized) long-term housing with voluntary services aimed at helping the homeless integrate back into the community. In general, permanent supportive housing falls into one of three broad silos:
- Purpose-built or single-site housing Apartment buildings designed to primarily serve tenants who were previously homeless. Supportive services are often available on site.
- Scattered-site housing People who are no longer experiencing homelessness lease apartments through the private market using rental subsidies, while accessing services from staff who can visit them in their homes.
- Unit set-asides Landlords agree to lease low-cost apartments to tenants who have exited homelessness or who have service needs, and partner with supportive services providers to offer assistance to tenants.
While Housing First has been discussed since at least the 1990s, Utah was perhaps the first state to widely implement the concept in 2005, followed by California, and then cities like Denver, Houston, and Seattle. Under President Obama, it became the centerpiece of federal homelessness policy. Housing First was deemphasized under President Trump, but is returning under President Biden.
The Biden administration recently issued its plan for combating homelessness, calling for a 25 percent reduction in homelessness by 2025. The $8.7 billion plan relies almost exclusively on Housing First.
Despite Housing First’s popularity among progressive activists and policymakers, a growing body of evidence suggests that it has failed. When Obama prioritized Housing First in 2013, he promised that the program could end veteran homelessness by 2015, chronic homelessness by 2016, and family homelessness by 2020. Obviously, it did not.
Too often, supportive services are either unavailable, or they are available but people fail to take advantage of them. Some studies have shown that Housing First programs have little or no impact on drug use, alcohol consumption, or psychiatric symptoms. Similarly, a National Institutes of Health review of five studies, involving a total of 7,128 people, found that Housing First programs did not reduce the likelihood of involvement in the criminal justice system. Metadata studies from the National Academy of Sciences and The Lancet also found little impact on these factors.
Homeless people are vulnerable to ending up back on the street when their underlying problems are not addressed. According to a Harvard University study of Housing First programs in Boston, retention of the homeless in subsidized housing was initially high, but declined significantly over time. After 10 years, just 12 percent of the previously homeless remained housed.
Since 2013, federal, state, and local governments have built at least 200,000 units of “permanent supportive housing” to support the Housing First model. Yet, the impact on homelessness has been marginal, at best. From 2009 to 2019, federal spending on homelessness increased 200 percent. From 2014 to 2019, the number of rapid rehousing and permanent supportive housing units increased by 42.7 percent. Yet over that period, the number of unsheltered homeless increased by nearly 21 percent.
Notably, those areas that have leaned most heavily into the Housing First approach have not seen a significant reduction in homelessness. One study by Vanessa Brown Calder and Jordan Gygi of the Cato Institute looked at California and Utah, two otherwise very different states that have both committed to the Housing First model.
Initially, Utah looked like a massive success. State officials credited Housing First with reducing homelessness by 91 percent between 2005 and 2015. However, a substantial portion of this decline appears to have resulted from a change in the state’s definition of homelessness: removing those in transitional housing from the homeless category. Moreover, since 2017, chronic homelessness has been rising rapidly in the Beehive State, up by more than 300 percent.
And, in California, chronic homelessness had been declining prior to the statewide adoption of Housing First in 2016. Since then, it has risen by 93 percent. Cities such as San Francisco, which prioritized Housing First for more than a decade before it was implemented statewide, continue to see high rates of homelessness.
Of course, rising rates of homelessness cannot alone quantify how many people would be homeless absent permanent supportive housing or Housing First. It may just be that new people are becoming homeless faster than cities and states can move people into shelters. Still, at some point, policymakers need to turn off the spigot of homelessness rather than simply continuing to bail water. As Calder and Gygi conclude, “It is possible that California and Utah policies still reduced homeless counts over a counterfactual without the policies, but they are not delivering on lofty promises to end homelessness or even reduce homeless numbers from the baseline.”
Moreover, there is evidence that Housing First itself may contribute to rising rates of homelessness. For example, to the degree that Housing First takes housing off the market, it can exacerbate shortages of affordable housing and increase homelessness by inflating rents. And Housing First requires funding that often comes at the expense of other programs. Transitional housing funding, for instance, has fallen by nearly 50 percent since cities began to emphasize permanent supportive housing. Phoenix has lost 2,500 short term and emergency shelter beds over the past decade, even as the city’s total homeless population has risen steadily. And that is a typical outcome.
Housing First programs cost a lot of money. Permanent supportive housing is both usually expensive and extremely slow to build. For instance, in 2016, Los Angeles voters approved Proposition HHH authorizing $1.2 billion in bonds to build permanent affordable housing. As of 2018, developers had only completed 14 percent of projects, and the average cost was a whopping $600,000 per unit. Some units cost as much as $837,000.
Shifting resources from transitional housing and more traditional shelters to permanent supportive housing is particularly counterproductive when one considers that the chronically homeless make up only a small portion of the total homeless population. Cities are taking resources away from the types of programs that the majority of homeless people need.
In addition, by concentrating large numbers of the previously homeless in one location: many with substance abuse problems or mental illness. Housing First programs can both contribute to community quality of life issues and make it more difficult for individuals to deal with their personal problems.
The research on Housing First is not unanimously negative. There are studies suggesting positive outcomes, especially in the short term and among motivated participants. But the weight of the evidence strongly suggests that communities should not put all their eggs — or even many of them — in the Housing First basket. Certainly neither federal nor state governments should be mandating this approach above all others.
Bad Idea 2: Overreliance on law enforcement
Many localities have responded to the rise in homelessness by prioritizing the enforcement of laws limiting the behavior associated with homeless people, including sleeping, loitering, or panhandling in public places. Some 48 states and at least 187 cities currently have one or more laws that restrict such behavior. This represents a nearly 50 percent increase in such laws since 2006. Some of these laws carry penalties of significant jail time and fines of up to $5,000 — a sum few homeless people are likely to have. Other locales have mobilized police to clear homeless encampments from certain public spaces.
These enforcement efforts attempt to address legitimate quality of life issues associated with large public homeless populations. However, they do not solve the underlying problems causing homelessness. Policing homelessness isn’t providing treatment for the mentally ill or providing treatment for substance abusers. In fact, some studies show that jailing the homeless can make their mental health problems worse.
Our criminal justice system already plays a significant role in homelessness: in effect, partially substituting for the deinstitutionalization of the late 20th century. One (admittedly dated) study of homelessness in New York City found that roughly 17 percent of single adults in city homelessness shelters had spent time in jail during the previous two years; roughly eight percent had spent time in prison. More recently, the Department of Housing and Urban Development (HUD) reports that nine percent of those entering a homeless shelter in 2017 had been incarcerated. It is hard to see how cycling still more people through the criminal justice system is likely to help anyone rise up from homelessness.
Homeless people frequently encounter police on the street. According to one survey of people without housing in San Francisco, 45 percent of those who are homeless, 46 percent of those camping in parks, and 20 percent of those living in a vehicle report being approached by police at least once monthly. Most of these interactions don’t lead to injury or arrest. More often, police simply force homeless people to move from one location to another. In the process, homeless individuals often lose what few possessions they have. The police throw away tents, but also necessary medication, mobility devices, identification, and clothing. Police will often impound the vehicles of homeless people living in their cars. Roughly 46 percent of homeless people in San Francisco report having their belongings confiscated at least once and 38 percent reported those items were destroyed. And for many women experiencing homelessness, the arrest of a male protector can leave them even more vulnerable to rape, exploitation, or abuse.
A 2019 study of police actions on homelessness in San Francisco concluded that such tactics were counterproductive. “Anti-homeless ordinances play an instrumental role in contributing to homelessness by systematically [limiting] homeless people’s access to services, housing, and jobs, while damaging their health, safety, and well-being.”
In fairness, some states and localities have combined policing with a mandate to provide shelter or at least a location where the homeless can camp. But many others have no plan for providing shelter and services. Those localities are doing little more than hoping that the homeless go away or become someone else’s problem. As Lt. Lawrence Davis, who leads the Austin Police Department’s camping ban enforcement, told the Texas Tribune, “If someone says, ‘I want to come into compliance, where do I go?’ We have no answer.”
Even those jurisdictions that attempt to offer an alternative frequently fall short in practice.
Police interaction with the homeless can also escalate very quickly. In New York City, for example, the Civilian Complaint Review Board has received more than 2,700 complaints since 2017 stemming from police-homeless contacts. The police themselves report that they are not comfortable serving on the front lines of their cities’ homelessness crisis. They also recognize that they lack the expertise to differentiate between those who pose a genuine risk to themselves and others and those who may just be acting eccentrically.
In the end, reliance on law enforcement is more about managing homelessness than solving it. Policymakers should not ignore how concentrated mass homelessness negatively affects quality of life for all community members, housed and unhoused. But unless communities are willing to make the investments necessary to deal with the underlying issues that cause and prolong homelessness, aggressive policing will not work.
Better ideas
If neither Housing First nor “enforcement first” is likely to significantly reduce homelessness, what other policies might have more impact? While no single reform is going to be the sole answer, evidence indicates that the following five approaches could make a difference.
Institutional care
While the mentally ill make up only a minority of the homeless population, dealing with this subgroup is one of the most difficult aspects of homelessness policy. Some present a danger to themselves or others, if only temporarily. Some are too incapacitated to make sound decisions. In these cases, cities and states must have a legal mechanism for ensuring people do not become violent or endanger themselves through self-harm or self-neglect. These people need legal custodians until they are capable of safely caring for themselves.
As discussed above, law enforcement is poorly situated to deal with homelessness head on. However, there is a better way to help get people off the street: conservatorship.
Conservatorship is a legal mechanism whereby a court can give an individual responsibility for another person’s safety, well-being, and other personal affairs.
Many people conflate conservatorship with involuntary commitment. While conservatorship may ultimately be used to institutionalize some homeless people, it can cover a variety of less intrusive interventions, such as determining where a person should live or applying for government benefits on an individual’s behalf. Traditionally, the primary focus has been personal finances, especially for the infirm elderly. But a conservatorship can help in other areas, particularly in cases involving mental health incapacity.
Conservatorship statutes vary widely, but most mental health conservatorships require a legal finding that an individual is “gravely disabled” or a “threat to self or others.” Although these terms can be open ended and expansive, the courts have generally interpreted “gravely disabled” to mean “unable to provide properly for his or her personal needs for physical health, food, clothing or shelter.” Clearly, there are people living on the street who fit this definition.
A study of conservatorships for mentally ill homeless people in Los Angeles found that “conservatorships can benefit some patients with grave disability from mental illness.” Homeless patients being treated under conservatorship had longer inpatient stays and were significantly less likely to be homeless upon release.
Several states are reexamining their conservatorship laws with an eye towards dealing with the problem of mental illness-driven homelessness. One potential conservatorship model is California’s 2022 Community Assistance Recovery and Empowerment (CARE) Act. The CARE Act allows family members, mental health providers, and first responders who have a history of engagement with a mentally ill individual to refer that person to a special judicial proceeding known as a CARE court.
Importantly, the court assigns a public defender to anyone so referred to help protect their rights and interests. The court can order a clinical examination and, if an individual is judged unable to care for themselves, county health agencies are required to develop a plan providing for that person’s treatment, stabilization medication, and housing. The homeless person can reject the plan and treatment. But if he or she does so, the court can subsequently refer that person for conservatorship. Significantly, the conservatorship can include provisions for involuntary treatment and even involuntary commitment.
Given the history of institutionalization in the 20th century, states must exercise caution if they expand or strengthen conservatorships. Policymakers must respect individual autonomy and lawful lifestyle choices, even if they disapprove of them. When dealing with people suffering from both homelessness and debilitating, incapacitating mental illness and/or substance abuse, policymakers must carefully balance the need for personal autonomy with the recognition that some people are — at least temporarily — unable to sensibly make or appreciate their choices. Cities must also differentiate conservatorships from the law-enforcement approach discussed above. Conservatorships exist to ensure treatment and shelter, not to punish. Incarceration is not conservatorship.
On the other hand, as California State Sen. Scott Wiener (D., San Francisco), one of the most progressive members of the California legislature and a sponsor of conservatorship legislation, points out, “It is neither progressive nor compassionate to just sit by and let people deteriorate, fall apart and ultimately die on our streets.”
Cities need a mechanism for mandating treatment for individuals who are unable to care for themselves. Conservatorships, accompanied by appropriate safeguards, but with the authority to require treatment and even involuntary commitment, can offer a promising structure for such interventions.
Invest in treatment
It makes little sense to encourage homeless people to seek shelter or treatment, whether voluntarily or involuntarily, if those services aren’t actually available. As the Los Angeles study concludes, success “requires substantial resources from facilities that initiate such conservatorships and does not guarantee resolution of long-term supportive housing needs.”
Yet, for many localities, a combination of funding concerns and “not in my backyard” NIMBYism results in a severe shortage of shelter beds, mental health services, and substance abuse treatment.
As more people have joined the ranks of the unsheltered homeless in recent years, some states and localities have attempted to increase the number of shelter beds available. Overall, the total number of beds has increased by roughly seven percent since 2015. But that remains far short of the number needed. Nationally, there are only enough shelter beds available for about 55 percent of homeless adults on a given night. That’s a shortage of around 188,000 beds. Moreover, shelter capacity does not always line up neatly with need. Some communities may have excess beds available, while others are experiencing deep shortages. According to the Coalition on Homelessness, as many as 4,000 people are forced to sleep on the street each night in San Francisco because shelters are full.
Mental health and substance abuse treatment is also often hard to come by. For instance, a 2018 audit found that San Francisco failed to offer 38 percent of those discharged from psychiatric emergencies any continuing services, increasing their likelihood of ending up back on the street. San Francisco also reportedly has a waiting list of more than 500 people for methadone or residential treatment for substance abuse.
This does create something of a conundrum for local governments. If they provide services, not only do they bear that cost, they may also attract homeless people from surrounding communities that do not offer such services. This creates an incentive for communities to shift responsibility to others. As one Grants Pass, Oregon city councilor said about a local anti-camping ordnance, “The point is to make it so uncomfortable for them in our city so that they will want to move on down the road.” It may be necessary, therefore, for the federal government to provide some level of equalization funding.
We cannot address homelessness without some public expenditure. Communities should invest in the resources necessary to provide treatment for mental illness and substance abuse, as well as short-term shelter.
Build more housing
While policies to deal with the mentally ill and substance abusers are essential to any strategy to combat homelessness, it is important to keep in mind that those groups represent a minority of those dealing with homelessness. The majority of the homeless are simply unable to afford housing. And, as discussed above, there is a remarkable consistency to the research that makes it clear that reducing housing costs should be an important component of any plan to deal with homelessness.
Basic economics teaches us that if the demand for housing increases, but supply remains low, prices rise. In a functioning market, suppliers respond to growing demand by increasing supply. When supply increases faster than demand, prices fall. However, government regulation artificially constrains the supply of housing in most growing cities in the United States.
There are several steps that federal, state, and local governments can take to help make housing more affordable. As FREOPP scholars Roger Valdez, Jon Hartley, and Avik Roy have described, this includes simplifying the Low-Income Housing Tax Credit; shifting from subsidized housing to Section 8 voucher-style rental assistance; reducing the systemic risks to housing markets caused by Fannie Mae and Freddie Mac; examining the role that the Federal Reserve has played in creating Housing bubbles; and, most importantly, eliminating exclusionary zoning regulations.
President Trump’s 2019 executive order establishing a White House Council on Eliminating Regulatory Barriers to Affordable Housing calls out federal, state, and local governments for erecting host of regulatory barriers to building more housing:
Overly restrictive zoning and growth management controls; rent controls; cumbersome building and rehabilitation codes; excessive energy and water efficiency mandates; unreasonable maximum-density allowances; historic preservation requirements; overly burdensome wetland or environmental regulations; outdated manufactured-housing regulations and restrictions; undue parking requirements; cumbersome and time-consuming permitting and review procedures; tax policies that discourage investment or reinvestment; overly complex labor requirements; and inordinate impact or developer fees.
According to the White House Council of Economic Advisers, if the 11 metropolitan areas with the most significantly supply-constrained housing markets were deregulated, overall homelessness in the United States would fall by 13 percent, and by even larger amounts in those areas.
For example, deregulating San Francisco’s housing supply would likely reduce that city’s rate of homelessness by 54 percent. Similarly, housing deregulation could reduce homelessness by 40 percent in Los Angeles and Oxnard, 38 percent in San Diego, 36 percent in Washington, D.C., and from 19 to 26 percent in Baltimore, Boston, Denver, New York City, and Seattle. These findings are generally in line with an earlier study by Stephen Raphael of the University of California–Berkeley, that estimated that housing market deregulation could reduce overall homelessness in the United States by 7 to 22 percent.
In addition, a study by William Ruger and Jason Sorens — as part of the Cato Institute’s Freedom in the 50 States index — found a direct relationship between greater land‐use regulation and homelessness. They found that a state’s relative land‐use freedom explains 38 percent of the variation in rates of homelessness between states.
Minneapolis provides a valuable case study. In 2018, Minneapolis abolished single-family only zoning and reformed other exclusionary zoning policies. The result was a near doubling in the number of multi-family housing permits issued monthly. Over the same period, the number of homeless in the city has declined steadily while rates of homelessness in similar midwestern cities have risen substantially. This includes its twin city St. Paul.
Whether or not policymakers support Housing First or other initiatives, cities cannot meaningfully reduce rates of homelessness without building additional housing. Consider Houston, one of the few cities to actually reduce rates of homelessness after implementing Housing First policies. Houston has implemented a number of anti-zoning, pro-housing policies keeping housing prices in Houston far lower than in California or Utah.
Because housing in Houston is cheaper, permanent supportive housing is also less expensive to build there than in comparable cities. Houston can provide one person with housing for one year for $17,000–$19,000, versus $40,000–$47,000 in San Francisco.
These lower costs have enabled Houston to provide permanent supportive housing for more than 1,000 people, provide another 3,200 people with rental assistance, and stabilize some 2,800 people close to homelessness for just $85 million. Meanwhile, Los Angeles spent more than $300,000 per unit to renovate hotel rooms to house homeless Angelenos. Given that funds are finite, reducing housing costs allows communities to help more people.
It is important to differentiate housing reform from Housing First.
Housing First is a bandage over an existing wound, moving homeless people into permanently subsidized housing. Substantive housing reform should be preventative, helping people avoid becoming homeless in the first place. Deregulating housing markets to allow for building more housing benefits everyone, regardless of economic status. But it is clearly critical to any successful effort to reduce homelessness.
Enable state and local experimentation
In the United States, homelessness is a complex problem that does not yield to easy answers or one-size-fits-all mandates. The composition of the homeless population, the causes of homelessness, and social, political, and economic circumstances vary greatly from community to community.
So it would be a mistake for Congress — or state governments for that matter — to dictate a universal approach to homelessness for every community.
Existing federal policies, however, routinely stifle local innovation. While the current funding mechanism provides a patina of local control and local agencies take the lead in fighting homelessness, they nevertheless receive their funding through grants provided by the federal Department of Housing and Urban Development. These grants include strings that allow HUD to explicitly or implicitly dictate how cities can use their funds. This significantly limits how much agencies can innovate. For example, as noted, the emphasis on Housing First has squeezed out more traditional approaches to providing shelter.
This also applies to so-called “continuum of care” (CoC) grants, which provide funding directly to local governments and nonprofits. While ostensibly a nod toward decentralization, the selection criteria and the interaction of local non-profits tends to lock out organizations that might try new or innovative approaches to solving homelessness. In many ways CoC organizations operate as cartels to protect both their own perquisites and the status quo generally.
Policymakers should give wide latitude to local communities that wish to experiment with and to develop case-specific responses to homelessness. However, that flexibility will be difficult to achieve as long as funding remains centralized in Washington. Congress, therefore, should make clear that Housing First is not the sole use for federally appropriated funds to fight homelessness. In fact, Congress should go further and convert such funding into additional block grants to the states with minimal strings attached (beyond reporting requirements and evidence of results). State governments should follow suit, setting broad policy parameters, but generally transferring funds to localities with few strings.
Broader social mobility policies
There is a tendency to think of homelessness as a discrete issue, separate from other forms of poverty. And, as noted above, it is true that many individuals suffering from homelessness have mental health or substance abuse issues that require a specialized approach to solve. However, for many or most homeless individuals, homelessness is simply a downstream result of more generalized poverty.
As John Quigley and Stephen Raphael put it, “the tough choices faced by households and individuals in the lower tail of the income distribution goes a long way towards explaining the problem [of homelessness].” Similarly, Barret Lee of the University of Pennsylvania concludes that, “The poverty component, though implicit, is fundamental. Affluent individuals who unexpectedly lose their housing (to fire, flood, and the like) can replace it quickly and avoid a prolonged homeless situation.”
More evidence that a significant portion of homelessness is an income problem, rather than a behavior problem, is provided by a new study published by the National Academy of Sciences. The Vancouver-based experiment provided a one-time unconditional $7,500 cash transfer to individuals experiencing homelessness. This one-time payment reduced homelessness and generated net financial savings of $777 per participant per year. The study also showed that, contrary to stereotypes, many homeless people are able to properly manage their funds.
However, when discussing poverty and homelessness, it is important to draw a distinction between individual poverty and broad geographic areas of poverty. Indeed, some of the highest rates of homelessness occur in areas that are highly prosperous overall, like Silicon Valley, in large part because the price of housing in those areas is too high.
Still, a long-term solution to homelessness will require addressing larger issues of poverty in America. Obviously, policymakers must deal with those people living on the streets today. But they should also deal with the long-term issues that lead more people to fall into homelessness every year. In Los Angeles, roughly 207 previously homeless people find housing every day. But 227 people become homeless daily. Bailing water from a sinking boat without fixing the hole in the hull will not keep it afloat, no matter how furiously one bails.
Solving the entrenched poverty issues that contribute to homelessness does not mean simply throwing more money at the problem. To state the obvious, welfare spending is far from the only component of social mobility. The federal government already spends roughly $1.1 trillion per year on more than 130 anti-poverty programs. (This includes 34 separate housing programs, managed by seven different cabinet departments.) State and local governments provide an additional $744 billion in anti-poverty funds annually. All this spending helps reduce material deprivation — though homelessness is evidence that it does not eliminate it — but does a much worse job of helping people escape poverty long-term. Yet, it is precisely that sort of long-term change that will be necessary to reduce homelessness.
Fortunately there are reforms that can have a long-term impact on poverty and economic mobility. These reforms include eliminating exclusionary zoning, reforming the police and reducing mass incarceration, giving parents greater choice and control over schools, and encouraging entrepreneurship, innovation, and job creation. These changes would help generate new opportunities for low-income Americans, and would lead to less poverty, more prosperity, and. as a result fewer people living in homelessness.
Conclusion
Homelessness is a complex problem. Tragically, the two most common strategies for significantly reducing homelessness, Housing First and increased policing, have not worked. However, that does not mean that the problem is hopeless. There are evidence-based solutions that can make a difference.
For that subset of the homeless population unable to care for themselves, cities need stricter conservatorship laws to get them into shelter or treatment. This must include the possibility of involuntary institutionalization or other mandated care, while also protecting their autonomy and civil liberties. Moreover, such efforts must be accompanied by an investment in the facilities and services necessary to provide shelter or treatment.
For the larger set of individuals and families whose homelessness results from an inability to afford housing, the answer isn’t more subsidies, but to reduce housing costs by removing barriers to new construction.
Importantly, rather than impose a one-size-fits all approach from either Washington or state capitals, local communities need broad discretion to experiment with solutions that best meet their unique circumstances and resources.
Finally, policymakers should recognize that homelessness is ultimately part of a larger problem of poverty in America. It will take progress on that larger issue to meaningfully reduce homelessness.
Homelessness is a growing problem with negative impacts on both people experiencing homelessness and the surrounding community. Fortunately, however, there are proven solutions that can make a real difference — if policymakers and activists focus rigorously on the evidence for what works.
