Homelessness Policy: What Each Position Is Protecting

Ana has worked as a case manager in a Housing First program for eleven years. She has placed hundreds of clients — people who came to her with active addiction, untreated schizophrenia, years on the street — directly into apartments with no sobriety requirement, no treatment mandate, no readiness test. She has watched clients who failed every previous program, who were told over and over that they were not "housing ready," get a key, close a door behind them, and begin to stabilize. Not all of them. Not quickly. But the retention rates in her program are three times what she observed in the staircase model she worked in before. She is not naive about what housing doesn't fix. She has held her clients' hands through relapses, hospitalizations, deaths. But she has watched the research bear out what she sees every week: stable housing is the prerequisite for addressing everything else, not the reward for having already addressed it.

Margaret worked as a psychiatrist for thirty years, the last fifteen in a community mental health center that served people cycling between the street and the shelter system after deinstitutionalization. She watched patients she had stabilized on medication return to her office six months later — worse than before — because no one had ensured they were taking it. She watched people who needed involuntary treatment spend years in encampments, getting older and sicker, because the legal standard for compelling care had been raised high enough that only the acutest crises qualified. She believes housing matters. She believes her patients, once housed, still need their lithium administered and their psychosis treated. She is not arguing against Housing First because she doesn't care about her patients. She is arguing that placing someone in an apartment without addressing the underlying illness is not an act of respect. It is an act of abandonment wearing the costume of autonomy.

Ana and Margaret are not fighting about whether homeless people deserve stable housing. They are fighting about what stability means, and what intervention actually produces it.

What housing-first advocates are protecting

The Housing First model — originating with Sam Tsemberis's Pathways to Housing program in New York City in 1992 — holds that permanent, independent housing should be offered immediately to people experiencing homelessness, without prerequisites for sobriety, treatment participation, or demonstrated "housing readiness." Services — mental health care, substance use treatment, employment support — are offered alongside housing but are voluntary. The model emerged as a direct challenge to the "continuum of care" or staircase approach, which required people to demonstrate recovery progress before advancing through emergency shelter, transitional housing, and finally permanent housing. Housing First advocates are protecting several interconnected goods.

They are protecting evidence over ideology. The randomized controlled trial Tsemberis and colleagues published in the American Journal of Public Health in 2004 — assigning 225 homeless individuals with dual diagnoses (mental illness and addiction) to immediate housing or to the treatment-contingent staircase model — found that the Housing First group achieved housing stability at dramatically higher rates (80 percent versus 30 percent at two years) without showing worse outcomes on substance use or psychiatric symptoms. The staircase model consumed more treatment resources and produced less housing stability. This was not a modest finding. It was a direct refutation of the assumption that requiring sobriety or treatment compliance before housing was the responsible approach. The advocates' position is grounded in this evidence base, which has been replicated across five Canadian cities, multiple European countries, and dozens of American jurisdictions.

They are protecting the principle that housing is a human need, not a reward. The staircase model treated stable housing as the goal to be earned through behavioral compliance. This logic, Housing First advocates argue, is backwards: you cannot meaningfully address addiction, manage medication, maintain sobriety, or hold a job without the basic cognitive stability that shelter provides. Chronic sleep deprivation, constant threat assessment, and exposure to trauma — the conditions of street living — degrade executive function and make behavioral change neurologically harder. Asking people to demonstrate readiness for housing before receiving it is asking them to perform recovery under conditions that make recovery nearly impossible. It is the institutional equivalent of telling someone to prove they can swim before letting them into the pool.

They are protecting cost-effectiveness as a policy standard. Dennis Culhane's research at the University of Pennsylvania documented that a small percentage of homeless individuals — those with chronic homelessness and serious mental illness — consumed a disproportionate share of emergency services. His work showed that a chronically homeless, mentally ill person cycling through shelters, emergency rooms, and jails costs significantly more per year than a permanent supportive housing unit. Culhane's finding — that placing people in housing could be cost-neutral or cost-saving when emergency service utilization was included — reframed Housing First not as a charitable expenditure but as a rational allocation of resources already being spent on chronic homelessness. This argument helped bring Housing First from advocacy into mainstream policy, including bipartisan adoption at the federal level under the George W. Bush administration's initiative to end chronic homelessness.

What treatment-first advocates are protecting

The critique of Housing First from the treatment-first side is not that housing doesn't matter. It is that housing alone, without adequate treatment for the underlying conditions that make housing retention difficult, addresses the symptom while leaving the disease. This position is often marginalized in policy debates because it can sound like a demand for compliance or a denial of rights. At its most serious, it is neither.

They are protecting the possibility of genuine recovery. Stephen Eide's Homelessness in America: The History and Tragedy of an Intractable Social Problem (Rowman & Littlefield, 2022) argues that Housing First's success at maintaining housing stability is not the same as its success at producing flourishing. A person who is housed but still in active psychosis, still using in ways that harm them, still unable to form stable relationships or engage meaningfully with the world — that person is not well. Housing as a "platform" for recovery only works if the platform is being used. Eide's critique is that Housing First's emphasis on voluntary services reflects an ideological commitment to autonomy that, in practice, leaves the sickest people without adequate care. The people who most need structured treatment — those with severe, treatment-resistant mental illness — are least likely to voluntarily engage with services once housed.

They are protecting the lessons of deinstitutionalization. The collapse of the state psychiatric hospital system between the 1950s and 1980s — driven by a combination of medications that allowed outpatient management, civil libertarian reforms that raised the bar for involuntary commitment, and budget pressure — produced the current crisis, in which a substantial fraction of the chronically homeless population has serious mental illness that is either undiagnosed or untreated. E. Fuller Torrey's American Psychosis: How the Federal Government Destroyed the Mental Illness Treatment System (Oxford University Press, 2014) documents how the shift from institutional to community mental health care was never adequately funded, leaving millions of severely mentally ill Americans without the consistent, supervised treatment that keeps them stable. Treatment-first advocates argue that Housing First policies sidestep this structural failure by calling it a housing problem rather than a mental health infrastructure problem — and that the solution requires rebuilding treatment capacity, not only building more units.

They are protecting a more demanding standard of success. The RCT evidence for Housing First measures housing stability as an endpoint. Housing stability is meaningful, but treatment-first advocates argue it is not the same as health, employment, sobriety, or participation in community life. A framework that counts housing retention as success may be systematically undercounting unmet need. The HUD Family Options Study — the largest randomized trial of homeless family interventions — found that while rapid rehousing produced faster exits from shelter, longer-term outcomes on employment, well-being, and family stability were more mixed. Advocates for treatment argue that the policy community has concentrated its evidentiary standards on the outcomes Housing First is best at achieving while paying less attention to the outcomes it is less good at — and that those other outcomes represent the actual goal.

What enforcement advocates are protecting

The case for enforcement — clearing encampments, prohibiting camping in public spaces, using police and city regulations to prevent the permanent occupation of parks, sidewalks, and public transit — is regularly dismissed as callousness or the desire to make poverty invisible. The genuine case is more complicated and deserves a serious hearing.

They are protecting public space as genuinely public. Robert Ellickson's "Controlling Chronic Misconduct in City Spaces: Of Panhandlers, Skid Rows, and Public-Space Zoning" (Yale Law Journal, 1996) makes the legal and philosophical case that public spaces serve important community functions — as commons accessible to all, including people with nowhere else to go — and that the permanent occupation of those spaces by some individuals imposes real costs on others, including low-income residents who depend on those same spaces. The enforcement position is not that homeless people should be hidden or punished. It is that the absence of any legal structure governing public space allows the spaces shared by the poorest urban residents to be degraded in ways that harm those residents. The choice is not between enforcement and dignity; it is between competing claims on shared space, each with legitimate standing.

They are protecting the safety of homeless people themselves. Large encampments — a phenomenon that has grown significantly in West Coast cities over the past decade — are not safe places. Violence, theft, sexual assault, and drug overdose are more common in encampments than in shelters or housing. The enforcement advocate's argument is that tolerating permanent encampments in the name of compassion is, in practice, allowing the most vulnerable people to live in dangerous conditions indefinitely, rather than pressing them toward the shelter system, services, and eventual housing. This is not an argument that enforcement without services works. It is an argument that non-enforcement combined with inadequate services produces a visible, permanent, dangerous underclass that is ill-served by the tolerance ostensibly extended to it.

They are protecting the practical accountability of service systems. When people have a visible alternative to shelter — however dangerous — some will choose it, including people who would benefit from services but who resist engagement. Enforcement advocates argue that removing the encampment option, combined with adequate shelter and service capacity, does not harm the homeless. It changes the calculus. Outreach workers who can offer services and a guaranteed shelter bed to someone living in an encampment have a clearer and more actionable offer than outreach workers operating against a backdrop of permanent tolerance. The goal is not punishment but directed engagement — using the legal framework to create conditions where accepting services is the path of least resistance rather than an imposition to be avoided.

What structural advocates are protecting

The structural position on homelessness holds that the individual-level interventions — Housing First, treatment-first, enforcement — are all responses to a failure that operates at a different level entirely. The cause of homelessness, on this view, is not primarily mental illness or addiction or poor choices. It is a housing market that has failed to produce enough units at prices ordinary people can pay, combined with a safety net that has eroded to the point where a single adverse event — a job loss, a medical crisis, an eviction — has nowhere near enough cushion to catch people before they fall into homelessness. Individual interventions are band-aids on a wound that requires surgery.

They are protecting an accurate diagnosis that leads to adequate solutions. Gregg Colburn and Clayton Page Aldern's Homelessness Is a Housing Problem (University of California Press, 2022) presents the most rigorous recent version of the structural argument: using county-level data across the United States, they find that rates of poverty, mental illness, and drug use do not predict homelessness rates across cities nearly as well as housing market conditions — specifically, the ratio of housing costs to median income. Seattle and San Francisco have high homelessness rates not because they have unusually high rates of mental illness or addiction but because their housing markets have produced a radical mismatch between cost and income. Memphis and Houston, with similar rates of poverty and addiction, have far lower homelessness rates because they have built enough housing to keep costs from outpacing incomes. The structural advocates' argument is not that mental illness doesn't matter — it is that mental illness doesn't explain the variance between cities. Housing markets do.

They are protecting the recognition that eviction is a cause, not just a consequence. Matthew Desmond's Evicted: Poverty and Profit in the American City (Crown, 2016) documented the mechanisms by which housing instability produces poverty rather than merely reflecting it. Families spending 60 to 70 percent of income on rent have no cushion for a medical bill, a missed shift, a broken car. Eviction erases their rental history, their school district, their social network, their accumulated possessions — and because an eviction record bars access to most subsidized housing, it traps them in the most precarious part of the market. Desmond's argument is that addressing homelessness through individual intervention — placing people who have been evicted into housing — is necessary but insufficient if the conditions that produce evictions at mass scale are unchanged. Universal housing vouchers that cap rent at 30 percent of income would, he estimates, make evictions rare rather than routine. That would prevent more homelessness than any amount of Housing First placements after the fact.

They are protecting the long-run sustainability of any intervention model. Finland is the only European country where homelessness has declined consistently over two decades — a 45 percent reduction in individuals experiencing homelessness since 2008 and a 75 percent reduction since 1987. Finland achieved this not only by implementing Housing First at national scale but by pairing it with robust social housing construction, income support systems that prevent people from falling to homelessness in the first place, and healthcare that does not leave mental illness untreated. Structural advocates read the Finnish case as evidence that Housing First works best as one component of a functional welfare state rather than as a substitute for it. When American policymakers adopt Housing First without addressing housing supply, wage stagnation, and safety net gaps, they are trying to extract the specific remedy from the system that makes it work — and finding that it scales less well than the original model because the pipeline from housing instability to homelessness keeps producing faster than individual placements can absorb.

Where the real disagreement lives

The homelessness debate often proceeds as if the four positions are in competition — as if accepting one requires rejecting the others. In practice, the deepest disagreements concern the right level of analysis, the right endpoint for measuring success, and the right distribution of responsibility between individuals, service systems, markets, and the state.

The level-of-analysis problem. Housing First is an individual-level intervention backed by strong evidence that it works better than alternative individual-level interventions. Structural advocates are not arguing that Housing First is wrong — they are arguing that it operates at the wrong level to address the problem. These are not contradictory positions. A city can have the best Housing First program in the country and still have growing homelessness if the housing market is producing new homeless people faster than the program can house them. The disagreement is not about what works for individuals who are already homeless. It is about what would prevent homelessness at scale — and those are different questions with different answers.

The autonomy problem. The treatment-first and Housing First positions encode different values about what it means to respect a person's autonomy. Housing First treats voluntary service engagement as evidence of respect: you offer services, you do not compel them, you trust the person to know their own needs. Treatment-first advocates argue that severe mental illness compromises the judgment required to make voluntary choices meaningful — and that deferring to the expressed preferences of someone in active psychosis is a form of abandonment that masquerades as respect. This is a genuine philosophical dispute, not a factual one, and it maps onto contested questions about when and how society may compel care on people who resist it. The answers have major implications not only for homelessness but for mental health law, guardianship, and the ethics of outreach more broadly.

The visibility problem. Enforcement debates are often conducted as arguments about hiding poverty vs. tolerating it, but the genuine enforcement case is about something more specific: whether the permanent encampment is a harm or a right, whether non-enforcement is compassion or abandonment, and whether the discomfort that homelessness creates for housed residents is a legitimate input into policy or something to be suppressed. The Supreme Court's 2024 ruling in City of Grants Pass v. Johnson — allowing cities to enforce camping bans even when shelter capacity is insufficient — shifted the legal landscape significantly. Whether that shift produces better outcomes for homeless people or worse ones depends almost entirely on whether enforcement is paired with adequate shelter and services or is simply used to move the problem out of sight. Enforcement without services is cruelty. Enforcement with services is a tool whose effects are contested and context-dependent. Most of the disagreement turns on which of those descriptions fits actual practice in any particular city.

The success-definition problem. The positions in this debate are measuring different things. Housing First counts housing retention. Treatment-first advocates count psychiatric stability, sobriety, and quality of life. Enforcement advocates count public order and safety. Structural advocates count total homelessness rates and housing cost ratios. None of these measures is wrong. All of them are real. The policy choices that follow from each depend on which metric you treat as primary — and that choice is not a technical one. It reflects a prior commitment about what homelessness policy is ultimately for: is it for housed individuals to maintain housing? For sick people to get treatment? For public spaces to be safely shared? For housing markets to produce enough supply that homelessness is rare? These goals are related but not identical, and they do not necessarily come from the same set of interventions.