Ending Homelessness in Toronto · Chapter 9
The Case Against: The Best Case Against This Plan, and Our Answers
A chapter of Ending Homelessness in Toronto — this project’s receipt-backed book on ending homelessness for less than the crisis costs now. Where a figure could not be confirmed against a primary source, the text says so on its face. How the checking works: how we verify · corrections.
A plan this ambitious deserves its most serious critics, not its weakest ones. This chapter sets out six real lines of argument against the plan in this book — each built from named, dated sources arguing in good faith, not straw versions of the other side — followed by this book's own answer to each. Where a critique lands, this chapter says so plainly. Where it proves less than it claims, this chapter says that too.
Does this actually pencil out? The fiscal case, led by the City's own auditor
The sharpest version of the "offsets aren't savings" argument comes from Stephen Eide of the Manhattan Institute: "Governments that invest heavily in Housing First programs should expect the overall cost of government to rise. For some individuals, or some service systems, there may be cost offsets, but cost offsets are different from savings. A $1 investment in Housing First may be offset by 30 cents in savings on other service systems, but that still means that the government is 70 cents larger... cost-effectiveness arguments should not lead anyone to think that Housing First investments will lead to tax reductions or somehow free up money" (Stephen Eide, Housing First and Homelessness: The Rhetoric and the Reality, Manhattan Institute, April 2020). Eide also names the "wrong pockets" problem directly: savings booked to a police or hospital budget do not automatically appear as cash the housing budget can spend. He cites Dennis Culhane — a pro-evidence homelessness researcher, not a Housing First critic — cautioning in 2008 against overstating the cost-savings case after reviewing more than forty cost studies and finding that the larger and more rigorous the sample, the lower the measured savings tend to be, meaning small, favorable pilot studies tend to overstate the case.
On Toronto specifically: the City's own Auditor General audited the Modular Housing Initiative — the exact delivery tool behind a meaningful share of this book's own capital-cost estimate — and found actual and forecast costs $33 million, or 63%, over the adjusted budget, only 159 of a planned 275 units completed at three of five planned sites, occupancy readiness taking five to thirteen months after construction start rather than the roughly five months this book cites elsewhere as the demonstrated pace, and, most tellingly, an audit team that stated it was "unable to conclude whether modular housing construction results in faster delivery and lower costs than traditional construction due to a lack of complete overall cost information and benchmarking information" (Auditor General of Toronto, City of Toronto's Modular Housing Initiative: The Need to Balance Fast Delivery with Stronger Management of Contracts and Costs, June 21, 2023). This is not an ideological critique. It is the City's own oversight office declining to certify the very cost-effectiveness claim a rapid-build strategy depends on.
Our answer. This is very likely the single most serious vulnerability in the entire plan, and it should be treated as exactly that, not argued away. Eide and Culhane are right that an avoided cost is not the same as cash in hand, and this book should never imply otherwise. On the Auditor General's own findings: Toronto's most comparable prior large-scale rapid-build program really did run 63% over budget and deliver barely more than half its planned units, and that is the load-bearing precedent for how skeptical any reader should be of a new capital-delivery promise. This book's answer is not to defend a single point-estimate cost, but to do exactly what an earlier chapter in this book does: build a range from the City's own reported and audited unit costs, price the 63% overrun in explicitly rather than assuming it away, and describe every claim of net fiscal benefit honestly as an avoided-cost argument, not a claim that government gets smaller. It does not, and this book should never say that it does.
Does Housing First actually work?
Eide's report makes four findings in its own executive summary: that Housing First "has not been shown to be effective in ending homelessness at the community level, but rather, only for individuals"; that it "may save taxpayers money" for a small segment of high-need people, but making it "the organizing principle of homeless services systems... will not save taxpayers money"; that its record on "untreated serious mental illness and drug addiction is far weaker than its record at promoting residential stability"; and that its record on "employment and addressing social isolation... is also weaker." He quotes the U.S. National Academies of Sciences, Engineering, and Medicine's 2018 consensus review, which "found no strong evidence of Housing First and improvement of mental disorders," and HUD's Family Options Study, a three-year study of over 2,000 homeless families, which found housing subsidies "increased rates of housing stability... but not self-sufficiency," and in some cases "actually led to diminished work effort."
Kevin Corinth of the American Enterprise Institute — the same researcher this book cites elsewhere for the Utah debunking — made the underlying objection earlier and more plainly: "Housing First goes wrong when it says that housing for the chronically homeless must be permanent. And, that supportive services should be offered, but cannot be required... Housing is a good start, but it's not enough. If we house every chronically homeless person but don't address their other needs, then we have failed them" (Kevin Corinth, "Homelessness Can Require Complex Solutions," AEI/New York Times, February 19, 2015; quotation abridged). The Cicero Institute's 2024 report adds that the At Home/Chez Soi trial — this book's own headline evidence — found "no significant difference between rates of substance use between participants of housing-first versus treatment-first programs," and cites a Denver supportive-housing study for a "50 percent higher mortality rate compared to the control group" — a citation this book checked against the underlying trial and could not support: the study reports mortality of roughly 10 percent against 9 percent, a difference that is not statistically significant. The critique is quoted as made; the check is this book’s own.
Our answer. This book's return-on-investment case rests on one 2014 randomized trial's high-needs cohort, run in Toronto's own 2009–2013 housing market — a real, confirmed result, but a genuinely thin evidentiary base to carry both the headline claim and the cost arithmetic in an earlier chapter, and this book says so there rather than only here. Eide and Corinth are, however, attacking a "Housing First as universal cure" framing this book does not actually make: the rapid-rehousing plan in this book is explicitly paired with a root-cause and income-support chapter and a shelter-transformation chapter, precisely because housing stability alone was never being sold as a mental-health or employment fix. The honest fix is narrower language throughout this book — housing retention, yes; addiction and employment outcomes, no — and a plain statement, made here rather than left for a critic to introduce first, of exactly how much weight one 2014 Toronto trial is being asked to bear.
Finland's reversal: proof the model is fragile, or proof the funding was?
This book's own strongest "it works" case study is also the site of its most dangerous fact. Finland's Y-Foundation, reporting on the national housing statistics centre's 2025 data: "Homelessness in Finland increased more in 2025 than ever before. The rise was record-breaking both in percentage terms and in the number of people affected... There are now 4,579 single homeless people, and overall homelessness increased by 20 percent in one year." Long-term homelessness rose even faster, up 29% in a single year. This is the first consecutive-year rise in Finnish homelessness since 2009 — before the national program existed. The Foundation's own chief executive, Teija Ojankoski, names the mechanism directly: cuts to housing allowance and social assistance are pulling people who were successfully housed back out of their homes.
Our answer. This is a front where the critique has full force and this book should not minimize it — but it cuts in a direction that helps this book's actual policy prescription more than it hurts it. Y-Säätiö's own diagnosis is not "Housing First failed," it is "the service system reacted too late" after benefit cuts — a funding-withdrawal story structurally identical to the argument this book makes about Canada's own 1993–1995 withdrawal from housing funding. Finland is not proof that Housing First is fragile as a model. It is proof that Housing First is fragile as a budget line the moment austerity touches it. This book's own ongoing-cost ask, an unaudited author's estimate, has to be argued for as a permanent, inflation-indexed commitment across electoral cycles, not a one-term pledge — or Toronto risks the identical trajectory on a shorter timeline, given that Toronto never had Finland's multi-decade head start.
Can Toronto actually rehouse everyone in a year?
Canada Mortgage and Housing Corporation's own 2025 Rental Market Report puts the Greater Toronto Area's purpose-built vacancy rate at 3% and its condominium vacancy rate at 1% — tight by historical standards, and the tightest leg of a three-tool sequence that puts rent supplements into the existing market first. The Canada-Ontario Housing Benefit, the fastest tool this book recommends, is portable and formula-based — but a household still has to find a landlord willing to sign a lease before the benefit can do its work, and reporting on social-assistance tenants describes landlord stigma and discrimination as a live barrier at exactly that step. On the support-staffing side, the Canadian Mental Health Association's Ontario division has reported hundreds of unfilled community mental-health and addictions jobs provincewide — roughly 250 in its 2023 accounting, still about 200 in its 2026 pre-budget submission — driven by working conditions and pay, with community mental health receiving a smaller share of Ontario's health budget than the benchmark share the sector's own analyses call for — precisely the workforce a rapid-rehousing support model depends on scaling. And Toronto has a documented, recurring pattern of organized opposition to shelter siting in both higher-income and working-class neighbourhoods alike, including campaigns against specific proposed shelters that drew petitions and, in one case, accusations of harassment against the shelter's supporters.
Medicine Hat's own record, discussed in the previous chapter, adds a sharper edge to this front: Canada's smallest and most mature coordinated-entry system has now let its "functional zero" declaration lapse twice — once in 2015, once in 2021.
Our answer. Every element of this front is real, and none of it should be softened — but none of it is fatal to the plan, only to an unqualified one-year promise. The vacancy and landlord-participation numbers argue for treating rent supplements as the fast-but-not-unlimited leg of a three-tool sequence, which is already this book's own architecture: rent supplements, then acquisition, then new-build, in that order. The honest fix is to make the one-year timeline explicitly conditional on landlord-recruitment and staffing benchmarks, rather than presenting it as fixed regardless of market absorption capacity, and to cost workforce investment explicitly into the ongoing-support ask rather than assuming staff simply appear at the modelled price. On neighbourhood opposition, the evidence cuts back further than most fronts in this chapter: a small Toronto study of existing supportive-housing buildings, produced by a Toronto organization of psychiatric-survivor advocates and based on 54 interviews — a real limitation, stated plainly rather than hidden — found no evidence that the buildings studied had negatively affected either property values or crime rates in their neighbourhoods. The fair distinction is between organized objections to siting a new shelter, which are real, and the after-occupancy record of supportive housing, where the evidence is more favorable — this book should draw that line itself rather than let the two be treated as identical by whoever raises the objection first. On Medicine Hat, the double-lapse history strengthens rather than undercuts the case for permanent, not campaign-cycle, funding — the same lesson Finland teaches at national scale.
The public-order case: encampments, crime, and "shelter-resistant" homelessness
The legal terrain around encampment clearing is less settled than a single citation suggests. Heegsma v. Hamilton (City), 2024 ONSC 7154 upheld Hamilton's by-law enforcement against a Charter challenge, but on a narrower basis than a plain reading implies: counsel for the City of Hamilton itself summarized the ruling as declining "to extend Charter protections to daytime or indefinite encampments" while requiring municipalities to still permit overnight sheltering where shelter beds are insufficient. The decision was under appeal as of this book's research, with the Court of Appeal for Ontario having heard arguments on February 10–11, 2026 and the decision still under reserve as of this book's last check — presenting the ruling as settled law overstates the certainty of the ground Toronto's own encampment strategy stands on.
Crime effects near shelters are a real, peer-reviewed, Canadian finding, not an American import. A study of Vancouver's emergency winter shelters, using a difference-in-differences design with statistical corrections for multiple comparisons, found "the presence of a shelter appears to cause property crime to increase by 56% within 100m of that shelter, with thefts from vehicles, other thefts, and vandalism driving the increase" (Faraji, Ridgeway & Wu, "Effect of Emergency Winter Homeless Shelters on Property Crime," Journal of Experimental Criminology, 2018). And the "shelter-resistant" homelessness debate is a genuine fault line argued in good faith on both sides: one line of research holds that chronically unsheltered people sometimes refuse available shelter for identifiable reasons; the advocacy-side answer holds that nearly all unhoused people would accept housing that actually met their needs, and that apparent refusal usually reflects untreated mental illness, trauma responses to congregate settings, and a strong preference for private space over dormitory-style shelter, not a preference for street life as such [flagged: not yet independently verified].
Our answer. The crime-effect finding deserves real weight and should not be minimized: it is Canadian, peer-reviewed, and methodologically strong. But its own authors bound the effect tightly — concentrated within 400 metres and converging toward a null effect beyond that radius — and the paper's own policy conclusion is "greater security" near shelters, not an argument against siting them. This book should cite that study directly rather than leave it for a critic to introduce first, and pair it with Toronto's own, more limited, supportive-housing study described above: emergency and winter shelters and permanent supportive housing appear to have different neighbourhood-crime profiles in the Canadian evidence base, and this book should not treat the two kinds of siting objection as interchangeable, because the evidence does not support treating them that way. On the pending appeal, this book states plainly that the legal question is not fully settled. On "shelter-resistant" homelessness, the low-barrier, trauma-informed, twenty-four-hour, no-turn-away shelter system this book calls for elsewhere is, functionally, the policy answer the advocacy side is asking for — and this book says so directly rather than leaving that connection for someone else to draw.
Is the evidence base itself too thin?
The critique that this book's evidence base leans on small-sample pilots, contested trial generalizability, and advocacy-adjacent research is fair — and it cuts symmetrically across both sides of this argument, not just this book's. On this book's own side: a single 2009–2013 randomized trial, run over a decade ago in a different housing-cost environment, is asked to carry both the headline Housing First return-on-investment claim and a meaningful share of the cost arithmetic elsewhere in this book — a weight this book names honestly rather than hides. Dennis Culhane's 2008 warning about small, favorable-looking cost studies is a methodological caution from within the pro-housing research community, which makes it harder to dismiss than an outside attack would be. Kevin Corinth's analysis of Utah's claimed 91% reduction — this book's own cautionary case in the previous chapter — found the number was driven by changing annualization formulas and shifting definitions of "chronic," precisely the kind of count-methodology question a skeptical reader would be right to ask of Toronto's own count trend as well. And the National Academies of Sciences, Engineering, and Medicine's 2018 consensus report concluded there is "no substantial published evidence as yet to demonstrate that PSH improves health outcomes or reduces health care costs," with the single exception of HIV/AIDS populations.
Our answer. The fairest response here is procedural, not rhetorical, because the critique is not wrong in kind, only in degree, and only if applied to one side alone. The sources marshaled against this book have their own unexamined small-sample and advocacy-adjacent problems, just as this book's own counter-evidence does — neither side's evidence base is as solid as its own advocates tend to present it. What should actually distinguish this book is process, not rhetoric: this book corrects its own numbers in the open when a claim turns out not to hold up, rather than defending them by default, and it should hold every new claim in this chapter — for or against its argument — to that same standard before treating any of it as settled.
The chapters that follow set out what this book concludes despite everything in this chapter, and what it would actually take to build.