Perspective Map
Homelessness and Housing Instability: What Each Position Is Protecting
Sam Tsemberis spent the 1980s doing street outreach in New York, asking people who had been sleeping on the sidewalks of Manhattan for years what they wanted. The answer was not, he says, a bed-and-bible program that required sobriety first, then jobs, then a room if you earned it. The answer was almost always: a place to live. In 1992 he founded Pathways to Housing, which gave apartments to people experiencing chronic homelessness — people who were actively psychotic, actively using substances, and had been living outside for years — and then offered support services afterward, without making services a condition of tenancy. The results were studied repeatedly. The people stayed housed. That finding eventually became the federal policy framework called Housing First, which the Department of Housing and Urban Development has promoted since the early 2000s. Tsemberis had a theory: housing instability was not a symptom of the underlying conditions; it was the condition itself. You cannot treat anything when you are living on a grate.
Father Gregory Boyle, who founded Homeboy Industries in Los Angeles, has a different framing for what he observed over three decades of working with gang-involved youth, many of whom cycled between incarceration and the streets. He does not disagree that housing matters. He has watched what happens when someone has four walls and nothing inside them — no community, no structure, no reason to get up. The housing advocates, he thinks, are right about what people need first. He is more skeptical of what comes after. The hardest thing to give someone who has spent years outside society's frame is a sense that their life is worth building. That is not a service. It is a relationship. And the system, as designed, provides neither.
Elena Rosen owns three restaurant locations in Seattle's Capitol Hill neighborhood. She has watched the sidewalk in front of one of them become the site of a rotating encampment for three years. She is not unsympathetic. She knows that the people who sleep there are suffering. She has called social services more times than she can count. She has offered meals. What she does not have is a business that can absorb a thirty percent drop in foot traffic indefinitely, or an answer for the employee who was assaulted outside the back entrance last February. She is told that clearing the encampment would only displace people without helping them, and she accepts that this is probably true. She is also watching what it does to the neighborhood when the political consensus has produced no housing, no treatment, and no consistent presence except police who show up, take tents, and leave. She is not asking for cruelty. She is asking for a plan.
Isaiah Thomas — not the basketball player — works as a policy researcher at a housing justice nonprofit in Oakland. He grew up in East Oakland and has spent twelve years studying the data on eviction, displacement, and unsheltered homelessness. The thing he has learned that most reshapes how he sees the debate: homelessness is not a population problem, it is a pipeline problem. The same communities lose people to homelessness at high rates year after year. The same failures reproduce the same outcomes. Psychiatric deinstitutionalization began in the 1960s and 1970s without the community mental health infrastructure that was supposed to replace it. Affordable housing stock was demolished for urban renewal and never replaced. Wages stagnated for the bottom third of the workforce while rents rose. Eviction — a process that takes weeks, costs almost nothing to initiate, and marks someone permanently in landlord screening databases — became the default mechanism for managing poverty in the rental market. Any policy that does not address the pipeline will be overwhelmed by it. You cannot shelter your way out of a production problem.
These four people are not arguing about whether homelessness is a bad thing. They are arguing about what it fundamentally is — a housing problem, a services problem, a public order problem, or a structural production problem — and their answer to that question determines what intervention they think is worth trying. The surface debate, which is often about encampment clearings, shelter mandates, and Housing First funding, is downstream of a more fundamental disagreement about causation and responsibility.
What Housing First advocates are protecting
The Housing First position rests on a body of evidence that has accumulated over three decades and holds up across different cities, different populations, and different study designs. The core finding is consistent: when people experiencing chronic homelessness are given stable housing without preconditions — without sobriety requirements, treatment participation requirements, or demonstrated readiness — they stay housed at high rates, and their outcomes across nearly every other measured dimension improve or at least do not worsen. The most rigorous American study, the At Home/Chez Soi trial conducted in Canada by Tim Aubry and colleagues, randomly assigned more than 1,000 people with serious mental illness and chronic homelessness to Housing First or treatment as usual; Housing First participants spent significantly more time stably housed across the two-year study period. They did not, on average, reduce substance use more than the control group. They did stay housed. And housing stability is the precondition for everything else.
What the Housing First framework challenges is a moral premise that has organized homeless policy in the United States for most of the twentieth century: that stable housing is an outcome you earn by demonstrating readiness, and that offering it without conditions removes the incentive to address the underlying problems that produced the homelessness. The "staircase" or "continuum of care" model that preceded Housing First required people to move through emergency shelter, to transitional housing, to permanent housing — at each step demonstrating sobriety, compliance, and rehabilitation. The problem is that the stairs are very hard to climb when you are in active psychosis, in active addiction, in chronic physical illness, or in the sustained stress response that living outside for years produces in the human nervous system. The readiness model, advocates argue, was not compassionate discipline. It was a mechanism that selected for the people who needed the least help and excluded everyone else from the front of the housing queue.
They are protecting the dignity of unconditional shelter — the recognition, grounded in evidence, that stable housing is not a reward for rehabilitation but the precondition that makes rehabilitation possible, and that a system which requires people to earn housing through behavioral compliance is asking something that the neurobiology and social circumstances of chronic homelessness make very difficult to deliver. Housing First advocates are sometimes accused of enabling addiction or ignoring public safety. The genuine version of this position is not indifferent to those concerns. It argues that trying to address addiction or mental illness without first addressing homelessness is addressing the symptoms of instability while maintaining the instability — and that the evidence supports this judgment.
What treatment-mandate advocates are protecting
The strongest version of the treatment-mandate position begins not from opposition to housing but from an observation about what unsheltered homelessness actually involves for a large share of the people experiencing it. The Substance Abuse and Mental Health Services Administration has estimated that roughly a third of people experiencing chronic homelessness have a serious mental illness, and a larger share have co-occurring mental illness and substance use disorders. The people who have cycled in and out of shelters, jails, hospitals, and the street for years — the population that housing policy focuses on most intensively — are often in active psychiatric crisis, often dependent on substances in ways that put their lives at immediate risk, and often unable to navigate the bureaucratic steps involved in obtaining and maintaining housing even when it is offered.
California's experience with CARE Court — the Community Assistance, Recovery, and Empowerment Act, signed by Governor Gavin Newsom in 2022 — reflects this argument in policy form. CARE Court creates a legal pathway by which family members, certain health providers, and public officials can petition a court to create a care plan for someone with schizophrenia or other psychotic disorders who is living on the street and refusing treatment. The plan can include housing, but it is structured around mental health treatment as the primary intervention. The argument is that offering housing to someone in active psychosis without structured support often results in that person losing the housing through behavior that the landlord cannot manage and the tenant cannot control. Compassion that offers the option of help to someone who cannot, by virtue of their illness, meaningfully evaluate options is, on this view, a form of abandonment dressed as respect for autonomy.
They are protecting the understanding that severe mental illness and addiction can make autonomous decision-making genuinely impaired — and that a policy framework premised on unconditional choice may, for the most acutely ill fraction of the population, defer the intervention that could most improve their lives by granting a nominal freedom they do not have the neurological capacity to exercise. Treatment-mandate advocates are often accused of coercive paternalism, and the history of forced psychiatric treatment in this country warrants caution. The genuine version of this position is not indifferent to that history. It argues that the alternative — watching someone deteriorate on the street out of respect for their stated preference — is also a form of harm, and that the policy conversation needs to reckon honestly with the fraction of people for whom Housing First alone is not sufficient.
What public order advocates are protecting
The public order position is often dismissed as a polite wrapper for NIMBYism, and sometimes it is. But there is a genuine version that is worth taking seriously. The visible presence of unsheltered homelessness in American cities has grown substantially over the past decade. The 2023 annual homeless count found more than 650,000 people experiencing homelessness on a given night, a thirty-six percent increase over 2007. In Los Angeles County, roughly 75,000 people are homeless on any given night. In Seattle, the proportion of public park visits that involve an encounter with open drug use or a tent encampment is high enough that usage of certain parks has measurably declined.
The genuine public order concern is not simply that wealthy residents find encampments aesthetically displeasing. It is that the effective withdrawal of public space from civic life — parks that are not used, sidewalks that are not walkable, transit stops that drive away riders — harms people who are not wealthy, including the low-income workers and residents who depend on those spaces and cannot simply drive around them. It is also that the visible concentration of people in acute crisis, with inadequate support and in conditions that are genuinely dangerous, creates harms to the people in encampments as well as to those near them. The argument is not that encampment clearing is an adequate response to homelessness — most public order advocates accept that it is not. It is that the alternative offered by the political system — tolerating encampments indefinitely while debating policy at the state and federal level — has its own costs, which are borne unevenly and are not ethically neutral.
They are protecting the legitimate claim that civic life requires functional shared spaces — and the recognition that the political consensus which has produced years of debate without producing housing, treatment, or alternatives has transferred the costs of that inaction onto the neighborhoods where people are living without shelter, not onto the policymakers who cannot agree on a solution. The public order position is frequently invoked in bad faith by people whose actual concern is property values. The genuine version is asking why the answer to that concern is always either "build nothing and clear encampments" or "build Housing First programs in someone else's neighborhood" — and whether a serious commitment to addressing homelessness would look different from both.
What structural critics are protecting
The structural critique of homelessness policy begins from a number that tends to get lost in debates about services and shelter: the vacancy rate for affordable housing in most major American cities is near zero. The number of people entering homelessness in any given year is larger than the number of housing units being created for them. Any service system operating in those conditions is, by definition, running to stay in place. The sociologist Matthew Desmond's study of eviction in Milwaukee, published as Evicted in 2016, documented the specific mechanics by which low-income renters cycle into homelessness: a medical emergency, a job loss, a broken car, an abusive partner — any single destabilizing event is enough to trigger eviction in a market where low-income renters have no margin for error. Eviction then marks someone in tenant screening databases, making the next apartment harder to find, which makes the next destabilizing event more likely to produce homelessness again. The pipeline is systematic, not random.
The historical dimension matters here. The deinstitutionalization of psychiatric patients, which began in earnest in the 1960s under the Community Mental Health Act, was not intended to produce homelessness. It was intended to replace large state hospitals with community mental health centers. The community mental health centers were never adequately funded. The urban renewal programs of the 1950s and 1960s demolished hundreds of thousands of low-income housing units — what were called SROs, single-room-occupancy hotels, which were not elegant but were cheap and plentiful — and replaced them with office towers and highways. The Fair Housing Act of 1968 outlawed explicit racial segregation in housing while leaving in place the zoning codes, restrictive covenants in spirit if not law, and investment patterns that continued to concentrate poverty in specific neighborhoods. These are not background conditions that homelessness policy must accept. They are policy choices that homelessness policy was built on top of.
They are protecting the recognition that homelessness is produced by specific, traceable policy decisions — about housing supply, psychiatric care infrastructure, eviction law, wages, and investment patterns — and that a service system, however compassionate and evidence-based, cannot absorb the output of a system that is producing housing instability faster than services can respond to it. Structural critics are sometimes accused of treating homelessness as purely a policy abstraction while real people sleep in the rain. The genuine version of this position does not dismiss the immediate need for shelter. It argues that a serious commitment to ending homelessness looks like a commitment to changing the conditions that produce it — not just managing the population that results.
Where the real disagreement lives
The debate about homelessness is, at its core, a debate about what kind of problem it is. If it is primarily a housing access problem — if the defining feature is that people cannot afford housing in the cities where they are — then the policy response is to build housing, subsidize housing, and prevent displacement. If it is primarily a services problem — if a large share of the people experiencing chronic homelessness have needs that unstabilized housing alone cannot meet — then the policy response centers clinical intervention, mental health services, and addiction treatment. If it is primarily a public order problem — if the central issue is the impact of unsheltered homelessness on civic space — then the policy response involves enforcement, shelter mandates, and clearing. And if it is primarily a structural production problem — if homelessness is less a population to serve than a pipeline to fix — then the policy response requires changing the conditions that produce housing instability, not just managing the people who fall through.
What is genuinely difficult about this debate is that all four framings are partially correct. Tsemberis is right that unconditional housing works for the population he studied. The neuroscientist and psychiatrist Thomas Insel, who reviewed California's mental health infrastructure after leaving NIMH, is right that a significant fraction of the chronically homeless population is in psychiatric states that make independent tenancy unsustainable without intensive support. Isaiah Thomas is right that the cities that spend the most on homelessness services have the fastest-growing homeless populations, and that this is not a coincidence — it is evidence that services cannot outrun supply failures. And Elena Rosen is right that the political stalemate that has produced neither housing nor services has transferred its costs onto the neighborhoods where people are actually sleeping.
The people holding each position are not, for the most part, confused about the facts. They are emphasizing different facts because they are protecting different things: the dignity of people in crisis, the lives of those too ill to advocate for themselves, the integrity of shared civic space, and the systems-level diagnosis that makes the smaller interventions sustainable. The deepest difficulty is that a serious response to homelessness requires all four — enough housing, adequate mental health services, maintained public space, and structural change to the conditions producing displacement — and that the political economy of each American city has, so far, made it impossible to pursue more than one or two at a time, at a scale far below what the problem requires.
See also
- Who bears the cost? and Who gets to decide? — the framing essays for the two questions running underneath housing instability: which people and neighborhoods absorb upstream scarcity, and who gets political power over the land-use and service decisions that shape that scarcity.
- Housing Affordability: What Each Position Is Protecting — the supply-side debate about zoning, density, and development that sits directly upstream of housing instability; understanding why housing is expensive is the structural precondition for understanding why people lose it.
- Mental Illness and Treatment: What Each Position Is Protecting — a large share of people experiencing chronic homelessness have serious mental illness; the debate about involuntary treatment, community care, and psychiatric authority maps directly onto the treatment-mandate vs. autonomy debate within homelessness policy.
- Drug Legalization: What Each Position Is Protecting — addiction and housing instability are deeply intertwined; the debate about criminalization vs. harm reduction in drug policy has a close parallel in the debate about preconditions vs. unconditional housing.
- Criminal Justice: What Both Sides Are Protecting — homelessness and criminalization are linked through both the criminalization of homelessness itself (camping bans, loitering laws) and the high rate at which formerly incarcerated people exit prison without housing.
- Wealth Inequality: What Each Position Is Protecting — housing instability is concentrated in the bottom income quintile, where small income shocks produce homelessness because there is no financial margin; understanding homelessness requires understanding the distribution of economic precarity.
- Urban Planning and Development: What Each Position Is Protecting — zoning law, the history of urban renewal, and the politics of neighborhood change are all upstream causes of the housing supply conditions that structural critics identify as the root of housing instability.
Further reading
- Matthew Desmond, Evicted: Poverty and Profit in the American City (Crown Publishers, 2016) — an ethnographic study of eviction in Milwaukee based on eighteen months of fieldwork in rooming houses and trailer parks; Desmond, a sociologist at Princeton, follows eight families through the eviction process, documenting how a single destabilizing event — job loss, domestic violence, medical emergency — can trigger an eviction that permanently marks someone in tenant screening databases and breaks the social networks and neighborhood ties that might otherwise absorb the shock; the book won the Pulitzer Prize in general nonfiction and is the most important contemporary account of the mechanics by which poverty produces housing instability; Desmond argues that eviction is not a consequence of poverty but one of its primary causes, and that the political economy of low-income rental housing depends on high turnover in ways that make stability actively counterproductive for some landlords.
- Sam Tsemberis, Housing First: Ending Homelessness, Transforming Systems, and Changing Lives (Hazelden Publishing, 2010) — the primary text by the founder of the Housing First model; Tsemberis describes the Pathways to Housing program he built in New York from 1992 onward, its theoretical foundations in the consumer/survivor movement and consumer-directed care, and the research evidence that accumulated from studies of the model; the book is most valuable for understanding what Housing First actually is, as opposed to what critics and advocates often claim it is — not a philosophy of non-intervention but a specific service model that provides unconditional housing and then offers (without requiring) an integrated array of clinical and social supports; Tsemberis's account of the "readiness" model it replaced explains why Housing First was experienced as radical even though its premises are, in hindsight, somewhat obvious.
- Thomas Insel, Healing: Our Path from Mental Illness to Mental Health (Penguin Press, 2022) — a former director of the National Institute of Mental Health examines the gap between scientific progress in understanding mental illness and the failure of the American mental health system to reduce its burden; Insel's account of California's homelessness crisis is particularly important for this debate: he argues that Housing First, while effective for the population Tsemberis studied, is not adequate for the fraction of chronically homeless people with untreated serious mental illness, for whom stable housing without intensive clinical support often ends in eviction or deterioration; Insel supports Housing First while arguing that the policy debate has been distorted by treating it as a complete solution rather than one component of a system that also needs robust involuntary treatment pathways, forensic mental health services, and long-term supportive housing for people who need it; the book is the clearest statement of the treatment-mandate position by someone who accepts the Housing First evidence base.
- Dennis Culhane and Stephen Metraux, "Rearranging the Deck Chairs or Reallocating the Deckchairs? Cost Studies of Services for Homeless People" (Journal of the American Planning Association, vol. 74, no. 1, 2008) — a review of cost studies examining the relationship between supportive housing and public expenditures; Culhane and Metraux's earlier 2002 study for the Psychiatric Services journal, which found that Housing First reduced shelter use and Medicaid costs enough to partially offset housing costs, became the most-cited piece of evidence in Housing First advocacy; this later review contextualizes those findings in a broader analysis of when and for whom cost offsets are observed; important for understanding the economic argument for Housing First and its limitations — the cost offsets are real but concentrated in the highest-need populations, and Housing First at population scale requires public investment rather than paying for itself.
- National Academies of Sciences, Engineering, and Medicine, Permanent Supportive Housing: Evaluating the Evidence for Improving Health Outcomes Among People Experiencing Chronic Homelessness (National Academies Press, 2018) — a systematic review of the evidence on permanent supportive housing commissioned by HUD; the panel reviewed 119 studies and found strong evidence that Housing First interventions increase housing stability, reduce shelter use, reduce psychiatric inpatient stays, and increase participation in mental health treatment; evidence for improvements in mental health and substance use outcomes was more mixed; the review is the most authoritative summary of the evidence base as of 2018 and is the closest thing to scientific consensus on what Housing First does and does not demonstrate; freely available online.
- Robert Rosenheck et al., "Cost-Effectiveness of Supported Housing for Homeless Persons with Mental Illness" (Archives of General Psychiatry, vol. 60, no. 9, September 2003) — one of the foundational studies of the economics of Housing First, using data from the HUD ACCESS demonstration project; Rosenheck and colleagues found that supportive housing reduced institutional costs (hospitals, jails, emergency services) substantially, and that those reductions partially offset housing costs for the highest-need populations; this paper, along with Culhane's work, established the evidence base that allowed Housing First to be framed not just as a moral choice but as a cost-effective one — an argument that was decisive in winning federal support.
- Forrest Stuart, Down, Out, and Under Arrest: Policing and Everyday Life in Skid Row (University of Chicago Press, 2016) — an ethnographic study of policing on Los Angeles's Skid Row based on eighteen months of fieldwork; Stuart, a sociologist, embedded with both the LAPD's Safer Cities Initiative and the people who lived on Skid Row, documenting how the strategy of intensive "quality of life" enforcement — misdemeanor arrests for camping, urination, and other survival behaviors — shaped the lives of people experiencing homelessness and how it interacted with the service ecology of the neighborhood; the book is the most detailed account of what enforcement-centered homelessness policy looks like in practice, and of the specific ways it generates collateral consequences — warrants, criminal records, disentitlement from benefits — that make future stability harder; essential for understanding why the public order position generates the resistance it does from service providers and advocates.
- Peter Rossi, Down and Out in America: The Origins of Homelessness (University of Chicago Press, 1989) — an early sociological account of the structural conditions that produced homelessness as a mass phenomenon in the 1980s; Rossi, a sociologist, conducted street surveys in Chicago and documented the demographic profile of people experiencing homelessness — disproportionately single men, disproportionately Black, disproportionately with mental illness and addiction, but primarily extremely poor; his analysis connected the emergence of mass homelessness to deinstitutionalization, the collapse of the SRO market, the end of industrial employment, and cuts to public assistance; the book established the structural framework that subsequent researchers have refined and remains an essential reference for understanding why homelessness is a production problem as well as a services problem.
- Margot Kushel, "Tackling Homelessness as a Medical Challenge" (New England Journal of Medicine, vol. 383, September 2020) — a summary by one of the country's leading homelessness researchers of what medicine now knows about the health consequences of homelessness and the evidence for housing-based health interventions; Kushel documents that homelessness produces mortality at roughly triple the age-matched rate in the general population, with people experiencing homelessness dying at median ages in the 40s and 50s; she reviews the evidence that housing interventions improve health outcomes across multiple domains and argues that addressing homelessness should be understood as a public health intervention, not just a social services one; the article is a useful entry point into the medical literature and is representative of how the clinical research community has come to understand the relationship between housing stability and health.
- Jennifer Wolch and Michael Dear, Malign Neglect: Homelessness in an American City (Jossey-Bass, 1993) — one of the foundational urban geography studies of homelessness, based on Los Angeles; Wolch and Dear documented how the spatial concentration of homeless service facilities in specific neighborhoods (Skid Row in LA, the Tenderloin in San Francisco) emerged from zoning decisions, NIMBYism, and the political economy of land use — and how that concentration simultaneously made services accessible and made those neighborhoods functionally ungovernable as residential areas; the book introduced the concept of the "service-dependent ghetto" and raised the distributional question that underlies the public order debate: why do some neighborhoods bear the full costs of a city's homelessness crisis while others bear none?