Did 'Housing First' Fail America's Streets? What $24 Billion in Audits, Randomized Trials, and the 'West Virginia Paradox' Actually Prove

In June 2024, the U.S. Supreme Court issued its most consequential urban-policy decision in four decades. By a 6–3 vote in City of Grants Pass v. Johnson, the justices overturned the Ninth Circuit's 2018 Martin v. Boise doctrine, ruling that the Eighth Amendment's Cruel and Unusual Punishments Clause does not bar cities from enforcing anti-camping ordinances, issuing fines, or arresting unhoused people for sleeping on public sidewalks—even when a municipality lacks enough shelter beds to accommodate them ^[8]. Within months, Republican legislatures, conservative mayors, and even Democratic governors in California and Oregon began dismantling sprawling urban tent encampments, while the Trump administration moved to overhaul federal homelessness policy under its domestic urban agenda ^{[8], [9]}.

At the center of this policy revolution is a fierce ideological and fiscal indictment of "Housing First"—the federal policy framework initially adopted under the George W. Bush administration and codified nationwide by the Obama administration's 2009 HEARTH Act and 2013 Department of Housing and Urban Development (HUD) Continuum of Care (CoC) guidelines ^{[3], [9]}. Under Housing First, federal and state homelessness grants prioritize moving unhoused individuals directly from the street into permanent subsidized apartments without preconditions such as completing drug rehabilitation, maintaining sobriety, or adhering to psychiatric medication ^{[3], [4]}.

To conservative policy scholars—including Stephen Eide and Judge Glock at the Manhattan Institute, Devon Kurtz at the Cicero Institute, and officials in the Trump White House—the real-world results of this federal orthodoxy speak for themselves ^[9]. According to HUD's Annual Homelessness Assessment Report (AHAR) to Congress, homelessness in the United States surged by 18.1% in a single year to reach 771,480 people on a single night in January 2024—the highest total ever recorded since federal Point-in-Time (PIT) tracking began in 2007 ^[1]. Even more damning to critics, a landmark April 2024 investigation by the nonpartisan California State Auditor revealed that California spent $24 billion across more than 30 homelessness programs over five fiscal years (FY 2018–19 through FY 2022–23) under a state law mandating Housing First compliance for all state grants—yet California's unsheltered homeless population grew to over 187,000 people (accounting for nearly half of all unsheltered Americans nationwide), while state officials failed to track whether the billions spent actually reduced homelessness ^{[1], [2]}.

"For over a decade, Washington and blue-state governments forced every dollar of homelessness funding into a rigid 'Housing First' monopoly that treats fentanyl addiction and untreated schizophrenia as mere housing-affordability issues. The result of handing out 'housing only' without sobriety requirements or public-camping bans has been $24 billion wasted in California, state-subsidized overdose traps, and record numbers of Americans dying on urban sidewalks." — Synthesis of Conservative Critique of "Housing First" (Manhattan Institute, Cicero Institute & Federal Urban Policy Agenda) ^{[2], [9]}

Does the empirical evidence support the conservative charge that "Housing First" failed America's cities and that street homelessness is primarily a crisis of untreated addiction and permissive camping laws rather than housing supply? Or are critics blaming a clinically proven intervention for the self-inflicted zoning and construction bottlenecks of high-cost coastal cities? To separate partisan rhetoric from hard evidence, The Dispatch audited the clinical trial literature, state fiscal audits, federal GAO econometric models, and municipal case studies from San Francisco to Houston ^{[1], [2], [3], [5], [6], [10]}.

What the Data Shows Part I: Clinical Trials of "Housing First" vs. "Treatment First"

To evaluate whether "Housing First" is a failed progressive experiment, it is first necessary to remember why the George W. Bush administration's homelessness czar, Philip Mangano, originally championed the model in the early 2000s ^{[3], [9]}. Throughout the 1980s and 1990s, the standard American response to chronic homelessness was the "Linear Continuum" or "Treatment First" ("staircase") model: an unhoused person living on the street with co-occurring schizophrenia and alcohol or drug dependence had to enter an emergency congregate shelter, achieve total sobriety in a detox program, stabilize in transitional housing, and demonstrate "housing readiness" before qualifying for a permanent apartment ^[3].

In practice, clinical evaluations in the 1990s found that the Treatment-First staircase suffered from catastrophic attrition: more than half of chronically homeless individuals either refused high-barrier congregate shelters or relapsed midway up the staircase, gets evicted back onto the sidewalk, and cycled endlessly through emergency rooms, psychiatric wards, and county jails ^[3]. In 1992, psychologist Sam Tsemberis founded Pathways to Housing in New York City, inverting the sequence: move chronically homeless individuals with severe mental illness directly into scattered-site apartments using rental vouchers, and wrap them in voluntary, intensive Assertive Community Treatment (ACT) teams (psychiatrists, nurses, addiction specialists, and peer counselors) ^[3].

Because Housing First vs. Treatment First was tested in multiple rigorous Randomized Controlled Trials (RCTs)—the gold standard of scientific evidence—researchers do not have to guess which model keeps people off the street:

• The Pathways & Multi-Site U.S. RCTs (Tsemberis et al., 2000/2004; CPSTF Systematic Review): In Tsemberis's landmark New York randomized trial published in the American Journal of Public Health (2004), chronically homeless adults with serious mental illness and substance use disorders were randomly assigned to either Housing First or a Treatment-First control group ^[3]. Over 24 months, 80% to 88% of Housing First participants remained stably housed, compared to just 39% to 47% of the Treatment-First group ^[3]. A subsequent comprehensive systematic review of 26 U.S. and international studies by the federal Community Preventive Services Task Force (CPSTF, published in the American Journal of Preventive Medicine, 2020) confirmed that across all randomized trials, Housing First programs increased housing stability by a median of 41% (and up to 88% in high-fidelity trials) while reducing homelessness by 88% compared to Treatment First ^[4].
• The At Home / Chez Soi Mega-Trial (2,148 Participants Across Five Cities): Between 2009 and 2013, the Mental Health Commission of Canada conducted the largest pragmatic randomized controlled trial of homelessness policy in history, randomly assigning 2,148 homeless adults with mental illness across Vancouver, Winnipeg, Toronto, Montréal, and Moncton to either Housing First (paired with ACT or Intensive Case Management) or Treatment as Usual ^[4]. Published in JAMA and Psychiatric Services (Stergiopoulos et al., 2015; Aubry et al., 2016), the trial proved that over two years, Housing First participants spent 73% of their days stably housed compared to just 32% for the Treatment-as-Usual group (with long-term 6-year follow-up in Toronto showing 85% stable housing retention among high-needs participants) ^[4].
• The Real-World HUD-VASH Proof of Concept (-55.6% in Veteran Homelessness): The strongest national proof that Housing First works when properly funded is the joint HUD-Veterans Affairs Supportive Housing (HUD-VASH) program, which combines HUD Section 8 rental vouchers with VA clinical case management under a strict Housing First directive ^[1]. Even as overall U.S. homelessness hit an all-time high in 2024, HUD's 2024 AHAR Part 1 report documented that veteran homelessness fell by another 7.6% in 2024 to 32,882—a cumulative 55.6% reduction since 2010 (down from 74,087 homeless veterans in 2010) ^[1].

Where the Conservative Critique Is Substantiated: The Addiction Blind Spot & California's $24 Billion Audit

If Housing First triumphs over Treatment First in randomized trials of housing retention, why do conservative analysts at the Manhattan Institute and Cicero Institute—as well as an increasing number of centrist mayors and clinicians—argue that the model has failed on the streets of San Francisco, Los Angeles, Seattle, and Portland ^{[5], [9]}? A rigorous audit of the clinical and fiscal data shows that conservatives are entirely right on three major blind spots of the Housing First orthodoxy:

1. The National Academies Verdict: Housing First Does Not Cure Addiction

Early advocates of Housing First frequently claimed that once an addicted or mentally ill person was placed inside an apartment, their substance abuse and psychiatric symptoms would naturally subside, generating massive net healthcare cost savings that would "pay for the housing" ^[5]. When the National Academies of Sciences, Engineering, and Medicine (NASEM) convened an expert committee in 2018 to audit the entire body of peer-reviewed literature (Permanent Supportive Housing: Evaluating the Evidence for Improving Health Outcomes Among People Experiencing Chronic Homelessness), the committee delivered a sobering reality check ^[5]:

• Zero Advantage on Substance Abuse or Psychiatric Recovery: Across both U.S. trials and the Canadian At Home/Chez Soi RCT, while Housing First dramatically improved housing tenure, it produced no statistically significant reduction in illicit drug use, alcohol consumption, or core psychiatric symptoms compared to treatment-as-usual control groups ^{[4], [5]}. As Stanford University addiction researcher Keith Humphreys and University of Alabama physician Stefan Kertesz have noted, unconditional housing keeps an addicted person off the sidewalk, but it is not a clinical treatment for fentanyl or methamphetamine addiction ^{[5], [9]}.
• The Single-Site Fentanyl Overdose Crisis ("Housing Only"): In Tsemberis's original New York model, Housing First meant scattered-site private apartments (where no more than 10% to 15% of units in an ordinary apartment building were occupied by program participants) backed by 24/7 clinical ACT teams ^[3]. In the 2010s and 2020s, however, cities like San Francisco, Los Angeles, and Seattle shifted to single-site congregate Permanent Supportive Housing (converting entire dilapidated Single-Room Occupancy [SRO] hotels into 100% formerly homeless buildings) while underfunding clinical staff—turning "Housing First" into "Housing Only" ^[9]. When synthetic fentanyl flooded West Coast drug markets after 2019, concentrating hundreds of active drug users in single-site buildings without drug-free floors or mandatory treatment options proved lethal: investigative audits in San Francisco revealed that residents inside city-funded permanent supportive housing buildings suffered fatal overdoses at rates far exceeding even the city's street population, while drug dealers actively preyed on recovering residents who had no option to request a sober-living building ^[9].

2. HUD's "Monopoly Rule" Banning Sobriety-Based Transitional Programs

Starting in 2013 (and reinforced during the Biden administration), HUD's Continuum of Care (CoC) funding rules went beyond supporting Housing First for the chronically homeless: they effectively required every program receiving federal CoC funds to adopt low-barrier Housing First rules, stripping federal funding from transitional housing programs and recovery residences that required sobriety, drug testing, or work/service participation ^[9]. In 2016, California went even further, enacting Senate Bill 1380 to mandate that every state-funded housing program across all state agencies strictly adhere to Housing First rules—making it illegal for state grants to fund sober-living recovery housing for homeless addicts who explicitly wanted a drug-free environment away from active dealers ^{[2], [9]}.

3. California's $24 Billion Audit & $800,000 Apartments

On April 9, 2024, California State Auditor Grant Parks released Report 2023-102.1, an exhaustive audit of the California Interagency Council on Homelessness (Cal ICH) ^[2]. The audit confirmed that between FY 2018–19 and FY 2022–23, California allocated $24 billion across more than 30 state homelessness programs, yet Cal ICH had stopped tracking comprehensive spending and outcome data after 2021 ^[2]. Out of five major state programs accounting for $13.7 billion examined in detail by the auditor, only two (Project Homekey hotel conversions and CalWORKs family prevention) could be shown to be cost-effective, while for the remaining three programs the state had zero reliable data on how many people permanently exited homelessness ^[2]. Worse, in Los Angeles (under the $1.2 billion Proposition HHH bond) and San Francisco, prevailing-wage mandates, environmental review delays, and local zoning bottlenecks drove the average cost of building a single new unit of permanent supportive housing to $600,000 to over $837,000 per studio apartment—meaning billions of taxpayer dollars produced only a trickle of actual units while tent cities expanded ^{[2], [6]}.

What the Data Shows Part II: The "West Virginia Paradox" and Why Homelessness Is a Housing Supply Crisis

While conservative critics are justified in indicting California's bureaucratic waste and the lack of addiction recovery in "housing-only" SRO buildings, the broader conservative talking point makes a much larger macro-level claim: that regional homelessness rates are driven by drug addiction, mental illness, and permissive blue-state welfare policies rather than housing costs ^{[6], [9]}.

In their landmark 2022 empirical study Homelessness Is a Housing Problem: How Structural Factors Explain U.S. Patterns (published by the University of California Press), University of Washington real estate professor Gregg Colburn and data scientist Clayton Page Aldern tested this exact hypothesis across every major U.S. metropolitan Continuum of Care ^[6]. Their statistical regressions uncovered what housing economists call the "West Virginia Paradox" ^[6]:

As Colburn and Aldern Explain through their "Game of Musical Chairs" analogy, confusing who becomes homeless within a city with why a city has a high rate of homelessness is a classic ecological fallacy ^[6]. Imagine 10 people playing musical chairs with only 8 chairs, and two of the players are on crutches. When the music stops, the two players on crutches will reliably be the ones left standing without a chair. If an observer asks, "Why are those two people standing?", it is tempting to say, "Because they are on crutches!" Yet if you take away two more chairs (leaving 6 chairs for 10 people), four people will be left standing—even though the prevalence of crutches did not increase. Conversely, if all 10 players are in West Virginia and there are 11 chairs available, even the two players on crutches get a seat ^[6].

Across the United States, multiple independent econometric studies confirm that regional homelessness rates are governed by the math of the local housing market—not local rates of drug use, mental illness, poverty, climate, or "magnet" migration:

1. Zero Correlation With Regional Addiction or Poverty Rates (Colburn & Aldern, 2022): Across U.S. Continuums of Care, Colburn and Aldern found that local poverty rates actually correlate negatively with per-capita homelessness (poorer cities like Detroit, Cleveland, and Huntington, WV have cheaper housing and lower homelessness), while regional rates of serious mental illness and substance use disorder show virtually zero statistical explanatory power ($R^2 \approx 0.01\text{–}0.03$) for cross-city homelessness variation ^[6]. Instead, two structural variables—median rent levels and ultra-low rental vacancy rates—explain the vast majority of regional variation ^[6].
2. The GAO Econometric Audit (+9% Homelessness per $100 Rent Hike): In a comprehensive July 2020 econometric study of 382 Continuums of Care nationwide (Report GAO-20-433, Better HUD Oversight of Data Collection Could Improve Estimates of Homeless Population), the nonpartisan U.S. Government Accountability Office found that after controlling for local wages, unemployment, poverty, and demographics, every $100 increase in median monthly rent was associated with a 9% increase in the estimated homelessness rate ^[7]. Similarly, Zillow Research economists (Glynn, Byrne, & Culhane, 2021) demonstrated that once a metro area's median rent exceeds 32% of median household income, homelessness enters a steep, non-linear inflection point where small rent increases trigger explosive surges in street homelessness ^[6].
3. Refuting the "Homeless Magnet" Migration Myth (UCSF Benioff CASPEH Study, 2023): A pervasive talking point holds that California's high homelessness rate is driven by unhoused people migrating from across the country to take advantage of California's mild weather, permissive camping laws, and generous welfare benefits. In June 2023, the University of California, San Francisco (UCSF) Benioff Homelessness and Housing Initiative published the largest representative survey of homeless adults since the 1990s (tracking 3,200 participants and 365 in-depth interviews across California) ^[6]. The study found that 90% of people experiencing homelessness in California lost their last stable housing inside California, and 75% lived in the exact same California county where they last had a lease or mortgage ^[6].

The Full Picture: Do Camping Sweeps Work, and How Did Houston Cut Homelessness by 63%?

What happens when cities use their newly restored authority under City of Grants Pass v. Johnson (2024) to sweep street encampments and enforce camping bans ^[8]? And if California's $24 billion outlay failed to reduce street homelessness while Texas's largest city succeeded, what separates failure from success?

1. The Public Health Cost of "Sweeps Without Beds" (Barocas et al., JAMA 2023)

Conservatives and moderate urban residents rightly argue that allowing permanent tent encampments to colonize public sidewalks, parks, and school zones is inhumane both to neighborhoods and to camp residents trapped in open-air drug markets ^[9]. However, empirical studies show a sharp divide between encampment resolution paired with shelter/housing versus involuntary sweeps when no shelter beds exist ^{[8], [10]}.

In a landmark April 2023 study published in the Journal of the American Medical Association (JAMA), physician-researcher Joshua A. Barocas (University of Colorado School of Medicine) and colleagues modeled the population-level health effects of continual involuntary displacement ("sweeps") of unsheltered people who inject drugs across 23 U.S. cities ^[8]. Because police sweeps without shelter placements confiscate or scatter medications (such as buprenorphine and methadone), destroy naloxone (Narcan) kits and identification documents, sever ties to outreach workers, and force individuals to inject drugs alone in hidden alleys with unfamiliar drug supplies, Barocas et al. found that continual involuntary displacement increases fatal drug overdoses by 15% to 25% and increases life-threatening infections and hospitalizations, while contributing up to a 7% increase in 10-year all-cause mortality—all without reducing the total unsheltered population, which simply relocates a few blocks away ^[8]. Separate longitudinal evaluations of Los Angeles encampment cleanups by the RAND Corporation confirmed that within weeks of an uncoordinated sweep, unsheltered counts in the surrounding neighborhood returned to pre-sweep baselines ^[8].

2. The Houston Blueprint: Why "Housing First" Worked in Texas

The ultimate empirical test of homelessness policy is Houston, Texas—a city that simultaneously validates the clinical efficacy of Housing First and the conservative emphasis on free-market housing supply, centralized accountability, and public-order enforcement ^{[6], [10]}. In 2011, Houston had one of the largest unsheltered homeless populations in the nation (over 8,500 people on a given night) and was placed on HUD's "priority list" for dysfunctional oversight ^[10]. Over the next 13 years, under both Democratic and centrist civic leadership, the Houston/Harris County Continuum of Care—operating as The Way Home—slashed overall homelessness by 63% (down to roughly 3,200 people across a metro area of 7 million) and moved more than 30,000 homeless residents into permanent housing, with roughly 90% remaining stably housed after two years ^[10].

Why did Housing First cut homelessness by 63% in Houston while homelessness rose in San Francisco and Los Angeles? Housing economists identify three structural differences ^{[6], [10]}:

• Elastic Free-Market Housing Supply (No Municipal Zoning Code): Unlike California cities—where restrictive single-family zoning, CEQA lawsuits, and permitting delays make apartments scarce and drive supportive housing construction costs above $700,000 per unit—Houston famously has no municipal zoning code and permits tens of thousands of multifamily apartments every year ^{[6], [10]}. Because Houston maintains a healthy 7% to 9% rental vacancy rate and abundant, modest market-rate apartments renting for $850 to $1,100 per month, The Way Home did not have to wait five years and spend $800,000 per unit to build government housing; instead, landlord liaison teams leased thousands of existing private-market apartments within weeks using modest rental vouchers and landlord damage guarantees ^[10].
• Centralized Command-and-Control Accountability vs. Fragmented NGOs: Whereas California distributed $24 billion across dozens of fragmented state agencies and hundreds of competing non-profits without unified tracking ^[2], Houston forced over 100 public agencies and non-profits into a single command center with one real-time by-name database, tying every dollar of funding to verified housing retention metrics ^[10].
• Coordinated Encampment Decommissioning (Housing + Enforcement): Houston does not allow permanent tent cities to take over public underpasses, nor does it conduct punitive police sweeps that scatter people with nowhere to go ^[10]. Instead, under Houston's Encampment Decommissioning protocol, outreach teams map every resident of an encampment, secure actual apartment leases or bridge units for 100% of the residents simultaneously, move them indoors with wraparound caseworkers, and then permanently close and harden the public site—successfully decommissioning more than 100 major encampments across the city ^[10].

Conclusion: Combining Housing Supply, Clinical Recovery, and Public Order

When stripped of ideological dogma on both left and right, the empirical evidence on American homelessness points to a clear, pragmatic synthesis. Conservatives are entirely right that "Housing First" was never meant to be "Housing Only" ^{[5], [9]}. As the National Academies of Sciences, Engineering, and Medicine documented, handing an apartment key to an individual suffering from severe fentanyl addiction or untreated psychosis does not cure their behavioral health crisis—and when cities pack hundreds of vulnerable individuals into single-site SRO hotels without clinical staffing, security, or the option of drug-free recovery housing, the results are tragic ^{[5], [9]}. Reforming federal HUD Continuum of Care rules so that cities can fund both low-barrier Housing First units and high-accountability, sober-living recovery residences—while demanding strict outcome auditing like that called for by the California State Auditor—is an overdue, evidence-based correction ^{[2], [9]}.

At the same time, policymakers in Washington and state capitals must not allow the shortcomings of California's bureaucracy to obscure the foundational economic truth proven by Gregg Colburn, Clayton Page Aldern, the GAO, and the city of Houston: you cannot treatment-mandate or police-sweep your way out of a structural housing shortage ^{[6], [7], [8], [10]}. Sweeping tent encampments under Grants Pass when no shelter beds or affordable apartments exist merely pushes destitute people into the shadows while driving fatal drug overdoses up by 15% to 25% ^[8]. By contrast, when a city pairs deregulated, abundant private-sector apartment construction with a centralized Housing First pipeline, intensive behavioral health teams, and coordinated encampment resolution—as Houston and the federal HUD-VASH veteran program have done—homelessness does not merely move down the block; it falls by more than half ^{[1], [10]}.