uniteTOlove.ca
This is a living research library. Claims carry verification statuses. A documented set of about 234 findings, where a derived page carries a source's claim without its original hedge, is known and queued for remediation, so read any affected line alongside its source backgrounder and check the corrections log for status. Found a mistake? Let us know.

Homelessness Economic ROI — The Full Cost Case by Order of Government — Backgrounder

DRAFT Research backgrounder

Briefs for this backgrounder

Status: DRAFT Version: v1.1 · Date: 2026-07-13 (a page under the broader homelessness-encampments / C2★ issue index row) · Domain: H — Homelessness & Housing-First Deep Dive What this page draws on: carried-forward (facts carried from that page's own internal recordsthis library's prior synthesis document (Economic ROI) and this library's prior executive synthesis (ROI Lock Memo), both sampled fully CLEAN at promotion, this project's later) and NEW (facts added this review, each with inline source quote: exact quote, source, date accessed). Citations-or-silence applies to both classes; no private individuals are named. Cui Bono: 0 beneficiary entities identified directly in this page (pointer to homelessness-political-economy.md's 2 ESTABLISHED rows) — added 2026-07-14, a later review; see "Cui Bono" section below.

Scope

The neutral scope question this document answers: what does the evidence show about the fiscal return on investing in Housing First and related interventions, broken out by which order of government pays and which order of government saves, and how confident should each figure be treated? This document covers: the federal, provincial, and municipal cost structures of the status quo; the modelled savings each order of government would realize from Housing First at scale; the "who pays vs. who saves" fiscal misalignment that the inherited research identifies as the structural reason underinvestment persists; honest uncertainty ranges around the headline totals; and one newly-surfaced 2025–2026 provincial-scale data point on whether recent funding growth has actually been landing on the housing-first side of the claims register.

This document does not cover: the clinical/programmatic case for Housing First as a model (homelessness-housing-first-model); shelter operating costs and capacity strain in their own right (shelter-system-capacity-strain, shelter-cost-per-bed-night); encampment-specific enforcement costs (encampment-clearance-policy-costs); the federal funding architecture's programmatic detail (homelessness-federal-funding-architecture); or Ontario Works rate history (ontario-works-rate-history-erosion). Those slugs own that material; this document hands off to them by name rather than re-deriving it.

A hostile-fact-checker warning inherited directly from the backbone documents and binding on everything below: the corpus this document draws from contains three structurally different "dollars per homeless person" denominators (the Auditor General's per-bed operating cost, the Fraser Institute's blunt total-budget-per-person figure, and one-time capital costs per bed/unit) that must never be quoted against each other unexplained, and it contains figures that were investigated and formally removed after independent re-checking (an 8.1× ambulance-utilization figure and a $48M/year Toronto Police mental-health-call cost, both because the underlying source document could not be verified to exist). None of the removed figures appear anywhere below. Restored 2026-07-16 (a later verification pass) — the actual denominator behind the Fraser Institute's blunt figure was previously referenced but never stated as a number: Toronto Shelter and Support Services' (TSSS) gross operating budget (all funders) was $897.957M for 2025 (Budget Committee, Jan 2025, Tier 1/locked) and $796.40M actual for 2024 (TSSS 2024 Annual Report, Tier 1/locked) — the total-budget figure the Fraser Institute's blunt ~$51,000/person calculation divides by an estimated homeless population, distinct from, and not to be conflated with, the Auditor General's per-bed $49,640/year figure used throughout this document's own modelling below.

Current state

The fiscal misalignment — the "tragedy of the commons" this document exists to document

The inherited backbone document's master framing, reproduced because it is the organizing logic for everything that follows: Ottawa currently backstops shelter costs for refugee claimants through the Interim Housing Assistance Program (IHAP) and pays the Canada Health Transfer for excess hospital use; the Province pays OHIP for elevated emergency-department and hospitalization use and covers corrections costs; the City pays the shelter bill directly and funds front-line response (police, EMS, fire, bylaw, encampment management). The order of government that invests most in Housing First — the City, which pays the shelter bill it would stop paying — is not the order of government that captures the largest share of federal or health-system savings, and the order best positioned to fund large-scale income supports (the Province) sees most of the resulting savings land at the City and federal level instead. This is the standing explanation, carried forward from the backbone documents, for why coordinated investment has been chronically difficult to assemble even where the aggregate case is strong.

Federal government exposure and modelled ROI

Ottawa's financial exposure runs through three channels: IHAP reimbursement, the Canada Health Transfer's exposure to excess provincial health spending, and IRCC claimant-processing costs. IHAP paid Toronto $261.87M in 2024 (actual, confirmed against the TSSS 2024 Annual Report), a figure the backbone treats as Tier 1 — locked, citable without caveat. The 2026 IHAP budget for Toronto was cut to $97M, also Tier 1. Total estimated federal annual exposure across all three channels is modelled at ~$575M/year, and total estimated federal annual savings from Housing First at scale (12,026 chronically homeless) at ~$312M/year — both of these are Tier 2 figures in the backbone's own tiering (Meta-methodology, single-model, not independently cross-validated) and must carry that caveat whenever cited.

The federal 10-year NPV case is the widest range in the entire corpus, and this document states it as a range rather than a point estimate on the backbone's own instruction: the well-reasoned upper bound is +$1.8B NPV (3.4:1 ROI, ~2.2-year payback), built on a $500M capital investment against ~$312M/year in modelled savings; the conservative estimate is +$37M NPV, built on a much smaller assumed IHAP attribution share. The backbone is explicit that "the range is $37M–$1.8B depending on assumptions about IHAP attribution" and that the $1.8B figure should be presented as "a modelled estimate," never as settled fact — the ROI-lock memo companion document flags this exact figure as the one place a prior draft of the backbone had let the caveat slip, and records the correction.

A separate, independently-dated national anchor — not additive to the Toronto figures above, a different time period and geography entirely — comes from Gaetz, Gulliver & Richter's State of Homelessness in Canada 2014, which found homelessness cost the Canadian economy over $7 billion a year against a federal Homelessness Partnering Strategy investment of $119 million a year, roughly a 60:1 ratio of documented national cost to federal response. Its value here is establishing that Toronto's current federal-underinvestment pattern is the continuation of an academically documented decade-plus pattern, not a novel complaint.

A newer national anchor, independently confirmed by the backbone's own July 2026 corroboration pass: the Parliamentary Budget Officer estimated that halving chronic homelessness nationally (the National Housing Strategy's own target) would require an additional $3.5 billion per year — roughly a seven-fold increase over the current $561M/year Reaching Home program.

Provincial government exposure and modelled ROI

The provincial case is more structurally complicated than the federal or municipal cases, because the backbone document's own analysis concludes that Ontario's largest potential lever — raising Ontario Works to a living-income level — is not, on its own, an ROI-positive provincial investment: it costs the Province an estimated ~$2.4B/year (a $767/month per-recipient OW increase across roughly 248,000 single recipients, plus full Canada-Ontario Housing Benefit coverage for Toronto's homeless population) while the resulting savings land mostly at the City (~$521M in avoided shelter costs) and federally (~$312M in avoided IHAP/health costs). The backbone's own verdict, quoted directly: "Raising OW is not an ROI-positive provincial investment in isolation. It costs ~$2.4B/yr for Ontario while generating most savings at the City ($521M shelter) and federally ($312M IHAP/health). The 'tragedy of the commons' in full effect."

The separate, Housing-First-specific provincial case is smaller but more directly ROI-positive: gross Housing First operating cost for 12,026 people (~$240M/yr) against a provincial cost-share of roughly 40% (~$96M/yr), set against estimated provincial savings of ~$165M/yr in avoided OHIP, corrections, and long-term-care costs, nets to an estimated ~$69M/year provincial saving, or a ~$591M 10-year NPV — again a Tier 2, single-model figure, not independently cross-validated.

The backbone's most targeted provincial recommendation is a narrower instrument than a blanket OW increase: a "Housing Stability Supplement" — a $500/month provincial top-up paid only to a recipient with a signed lease and confirmed COHB — modelled at roughly $72M/year for 12,000 people, versus $2.28B/year for a full OW increase covering the same population's income adequacy gap. This is presented in the backbone as an unranked proposal worth further costing, not a confirmed program.

Municipal government exposure and modelled ROI — the most direct case

Toronto's case is both the largest in dollar terms and the most directly evidenced, because it rests on two Tier 1 (locked, audited) figures rather than a modelled projection: Toronto's audited shelter cost is $136/night = $49,640/year (Toronto Auditor General, 2025), and the Housing First (ACT model) net annual cost, after a 69% cost offset from avoided emergency-service use, is $6,311/year (gross cost $20,367/year before the offset), from Latimer et al. 2020 in Psychiatric Services — a figure the backbone's own audit pass re-confirmed exactly, including correcting a wrong DOI across five documents in the same pass. The arithmetic difference, $43,329/year saved per person housed rather than sheltered, is itself Tier 1 (arithmetic from two locked figures).

Applied to Toronto's 12,026 estimated chronically homeless population, the shelter-cost-avoidance line alone is modelled at ~$521M/year — the single largest line item in the entire consolidated case. Sourcing note added (a later verification pass, 2026-07-13): the 12,026 figure was not independently derived or sourced anywhere in this page’s prior draft. The most plausible derivation, checked directly this review: 15,418 (the City's October 23, 2024 Street Needs Assessment total homeless-population count, the same figure the sibling shelter-system-capacity-strain page uses) × 78% (the homelessness-housing-first-model backgrounder's cited figure for "78% of shelter users meeting TSSS's own (180+ day) chronic-homelessness threshold") = 12,026.04, an exact arithmetic match. This derivation is stated here as the most likely explanation, not confirmed as the backbone's own original methodology — no source document in this page actually shows this calculation being performed, so it is presented as a plausible reconstruction, not a confirmed citation. ⚠️ Still being checked: this reconstruction also conflates two different populations — "78% of shelter users" (a shelter-system-specific rate) applied to 15,418 (the total homeless population count, which includes unsheltered/outdoor people not necessarily captured in the shelter-user rate) — a methodological imprecision worth flagging rather than silently accepting as exact, even though the arithmetic itself matches to the decimal. Adding non-shelter municipal cost lines (EMS, fire, bylaw, encampment response — each modelled, Tier 2, and each excluding the specific ambulance and police figures that were formally removed after failing independent verification) brings the modelled municipal total to ~$563M/year, revised down from an earlier ~$599M estimate specifically because of that removal — the backbone frames this revision as evidence the correction process works, not as a reason to distrust the remaining figure.

The Housing First operating-cost comparison ($6,311/yr vs. $49,640/yr) implies a 7.9:1 operating-cost leverage ratio, and the backbone separately models a prevention case with the widest ROI ratio in the corpus: a $2,000–$5,000 eviction-prevention intervention (e.g., a Rent Bank grant) avoiding a full year of shelter cost ($49,640) implies a 10–25:1 prevention ROI — the highest single return identified anywhere in this cost case, though built on a smaller and less rigorously audited evidence base than the Housing First comparison itself.

Productivity and GDP effects

The weakest-evidenced line in the consolidated case, and treated that way in the backbone: an estimated 25–30% of Toronto's homeless population (roughly 3,855–4,625 of 15,418 people, per the October 2024 Street Needs Assessment) is modelled as workforce-ready if housed, at an assumed entry-level/part-time contribution of $20,000–$35,000/year, yielding a modelled annual lost-GDP range of $77M–$135M, with ~$100M/year used as the campaign-safe midpoint. This figure excludes lost income-tax revenue, lost CPP/EI contributions, and the human-capital cost of the documented 3.5× elevated mortality rate among homeless populations (Hwang 2009) — it is a lower bound on the productivity case, not an estimate of the full economic cost.

The consolidated case, stated as a range, not a point estimate

Summing the four lines above (City ~$563M/yr, Province ~$165M/yr, Federal ~$312M/yr, Productivity ~$100M/yr) produces a headline steady-state total of ~$1.14B/year, against an annual net Housing-First operating cost of ~$75.9M/year for the same 12,026-person population — implying a net annual benefit of roughly $1.06B/year at steady state, with a modelled 10-year capital requirement of ~$5B implying break-even around Year 5.

This document states, per the backbone's own explicit instruction, that this total is a single-point steady-state estimate, not a range, and should not be cited as such without qualification. The City figure's shelter-savings component — $521M of the $563M total — has a documented sensitivity range of roughly $220M (conservative) to $580–620M (optimistic), with $521M as the base case; the backbone's own guidance is to lead with the conservative figure under skeptical questioning specifically because "it is still enormous, and it's more credible to start low." The Province, Federal, and Productivity lines have not yet had an equivalent sensitivity-range exercise applied to them in the underlying research — a gap this document states rather than papering over with a false symmetry.

A separate, more aggressive per-person model in the same backbone document produces a combined 10-year NPV of ~$13.8B (Federal $6.5B, Provincial $2.26B, Municipal $5.01B) by scaling to the full 15,418-person homeless population rather than the 12,026-person chronic subset and using a broader federal savings definition. The backbone is explicit that this figure is not the one to use in public-facing material — "the $1.8B-on-$500M-investment figure is the more transparent and defensible figure for advocacy use. The $13.8B combined figure is the upper-bound 'cost of inaction' argument" — and this document preserves that distinction rather than quoting the larger number without its context.

A cost line usually left out entirely: mortality, monetized at the government's own standard

A later integration pass into the backbone (from a dedicated full-cost-accounting companion document) applied Treasury Board of Canada's own required Value of a Statistical Life ($9.8M, 2026 dollars — the figure federal departments are themselves instructed to use in cost-benefit analysis) to Toronto's documented 59 shelter-resident deaths in 2024, producing a $578.2 million single-year loss figure — the single largest dollar figure in this cost case, larger than the IHAP budget or any capital program line, and one that had previously sat unmonetized in mortality statistics the wider corpus had already verified. Correction/primary-source citation added (a later verification pass, 2026-07-13, live-verified): the 59-death figure is directly confirmed by the City's own "Annual Review of Statistics on Deaths of Shelter Residents," which states: "59 deaths occurred in 2025, which remains unchanged from 2024" and "Since 2007, there have been 884 total shelter resident deaths reported to TSSS" (City of Toronto, https://www.toronto.ca/city-government/data-research-maps/research-reports/housing-and-homelessness-research-and-reports/deaths-of-shelter-residents/annual-review-of-statistics-on-deaths-of-shelter-residents/, page date modified 2026-04-28, accessed 2026-07-13) — this replaces the prior internal-only cross-reference to an earlier synthesis document with a directly checkable primary source, and independently confirms the 59-death figure for both 2024 and 2025. The backbone flags the $578.2M figure explicitly as requiring careful framing ("using the federal government's own required methodology") rather than being read as a literal price on a life, and as belonging in future headline ROI arguments rather than being treated as a footnote.

Correction (a later verification pass, 2026-07-13, live-verified): the same integration pass documents municipal legal-liability exposure as a real, previously uncounted cost category. Two genuinely separate pieces of litigation were previously conflated under one unsourced "$50M class action" mention: (1) a single 2021 encampment-clearance event cost the City approximately $2M directly [see encampment-clearance-policy-costs page]; (2) the five-plaintiff Lamport Stadium personal-injury suit over the 2021 encampment clearance, detailed in the encampment-clearance-policy-costs page, remains unresolved; and (3) a separate, distinct $50-million proposed class action, filed and reported June 2025, alleges the City violated refugee claimants' Charter rights by systemically denying them shelter beds between November 7, 2022 and October 1, 2023, even when beds were available — this is not an encampment-clearance case, and the two should not be conflated (CTV News/Joanna Lavoie [The Canadian Press wire], "Toronto hit with $50M class action lawsuit over decision to deny refugees shelter beds," published 2025-06-04, https://www.ctvnews.ca/toronto/local/article/toronto-hit-with-50m-class-action-lawsuit-over-decision-to-deny-refugees-shelter-beds/; Global News/Rianna Lim [The Canadian Press], "Proposed lawsuit alleges Toronto violated refugees' rights by denying shelter beds," published 2025-06-04, https://globalnews.ca/news/11213143/ont-toronto-refugee-lawsuit/; both live-fetched and quote-confirmed independently re-verified 2026-07-16 — original CP24/CBC citation retargeted to these two directly-fetchable outlets carrying the same Canadian Press wire story, since CP24 and CBC are systematically JS-gated on direct fetch). The lawsuit seeks $50 million plus special damages to be determined; the City has stated only that the matter is before the courts (city spokesperson Elise von Scheel: the city will respond "in due course" and has no further comment as the case is before the court). Both pieces of litigation remain unresolved as of this writing, and are genuinely separate proceedings over genuinely separate underlying facts (encampment-clearance use-of-force vs. refugee shelter-bed denial), not one case described two different ways.

New this review: does recent funding growth actually track the Housing First lever the ROI case rests on?

A live-discovery search (2026-07-13) surfaced a genuinely new, dated, province-wide data point not previously in this page’s backbone: the Association of Municipalities of Ontario and HelpSeeker Technologies' Municipalities Under Pressure One Year Later update, published 2026-01-13, updating a province-wide baseline first published January 2025 with a full year of 2025 data. Its most load-bearing finding for this document's fiscal-misalignment argument: "while the number of people experiencing homelessness increased by 49.1% between 2021 and 2025, total funding increased by 32.1% over the same period, with municipal funding increasing by 48.2%" (HelpSeeker Technologies, "Municipalities Under Pressure One Year Later," accessed 2026-07-13). The same report states that "program expenditures increased by 75.4% overall—rising by 88.0% for homelessness programs and 66.1% for housing programs—indicating that municipalities are increasingly absorbing the cost of managing higher and more persistent levels of homelessness through local service delivery."

This is province-wide Ontario data, not Toronto-specific, and it measures total system spending growth, not Housing-First-specific investment — it should not be summed with, or presented as confirming or refuting, the Toronto-specific dollar figures above, which model a specific counterfactual (housing people instead of sheltering them) rather than describing what has actually been funded. Read carefully, though, it is directly relevant to this document's own fiscal-misalignment argument in two ways: first, it independently corroborates, at province scale and with 2025 data, the backbone's claim that municipalities are structurally absorbing a disproportionate and growing share of system cost; second, it reports that within that growing spend, "emergency shelters have remained the largest area of homelessness-related expenditure, increasing by 51.6% since 2021... [while] community housing has remained the largest area of housing program expenditure, but spending declined by 0.6% over the same period" — meaning recent funding growth has disproportionately gone to the crisis-response side of the claims register (shelters) rather than the supply/exit side (housing), which is broadly consistent with, rather than contradicting, this document's own argument that the type of spending matters as much as the amount. ⚠️ Still being checked: no claim in this page’s evidence base yet directly quantifies how much of Toronto's own specific shelter-vs-housing spending mix has shifted over the same period; the AMO/HelpSeeker figure is province-wide and should not be read as a Toronto-specific number.

The same report also restates its own prior "two investment approaches" framing, worth carrying forward as an independent, differently-sourced comparator to the backbone's $500M federal-bond and $5B-over-10-years figures: "a longer-term investment of approximately $11 billion over 10 years to achieve functional zero chronic homelessness... or a targeted near-term investment of approximately $2 billion to reduce acute pressures such as encampments and stabilize conditions." This is a province-wide figure, built by a different organization using a different methodology than the backbone's Toronto-specific modelling, and this document treats it as an independent comparator rather than folding it into the Toronto totals above.

New this review: a health-economics angle absent from the existing evidence base

A second live-discovery find, also dated 2026-07-13: a July 2025 JAMA Network Open mathematical-modelling study by University of Toronto engineering researcher Isabelle Rao and Stanford's Margaret Brandeau found that a housing-first intervention for a modelled population of 1,000 unhoused individuals with opioid addiction "reduced total overdoses and fatal overdoses over the five years by 11 per cent and nine per cent, respectively," at a modelled intervention cost of US$96,000 per person, for an incremental cost of US$26,800 per quality-adjusted life year (QALY) gained compared to no housing intervention, with the university's own reporting noting that "when factoring in savings to the criminal justice system, such housing programs may even save money" (University of Toronto, "'Housing-first' more effective than 'treatment-first'...," accessed 2026-07-13). ⚠️ Still being checked: this is a mathematical-modelling study, not an empirical Toronto trial — the researchers are, per the same reporting, still "seeking to... expand into Toronto" and have so far engaged with officials in Santa Clara County, California, where Stanford is located. The figures are in US dollars and should never be presented as Canadian-dollar or Toronto-specific figures. Its value to this document is narrow but real: it is a second, independently-derived, health-economics-framed (QALY-based) methodology reaching the same directional conclusion as the Latimer/At Home-Chez Soi cost-offset evidence already in the backbone — that housing-first interventions are cost-effective by standard health-economic measures — using a different currency, a different population (opioid-specific), and a different country's cost structure. It does not add or subtract from any dollar figure elsewhere in this document.

Key tensions / tradeoffs

Point-estimate confidence versus range discipline. The federal 10-year NPV case spans $37M to $1.8B depending on IHAP-attribution assumptions — a nearly 50-fold range — and the backbone is explicit that only the range, not the upper figure alone, should be presented publicly. The consolidated $1.14B/year total is itself a single-point steady-state estimate resting on Meta-methodology figures for three of its four lines, not independently cross-validated the way the municipal shelter-cost-avoidance figure is. This document surfaces both tensions rather than resolving them toward the larger, more dramatic figures.

Whether recent province-wide spending growth supports or complicates the ROI case. The AMO/HelpSeeker finding that Ontario-wide funding grew 32.1% while homelessness grew 49.1% between 2021 and 2025 could be read two ways: as evidence that "more money" alone does not solve the problem (a challenge to any simple ROI-based funding ask), or — per this document's own reading above — as consistent with the ROI case's actual claim, since the spending growth that did occur skewed toward emergency shelters (+51.6%) rather than the housing-first/exit-capacity side of the claims register (community housing, −0.6%). The claims register of evidence available to this document does not contain a claim that adjudicates between these two readings for Toronto specifically; both are stated here, neither is resolved.

Anti-poverty asks versus homelessness-specific ROI. A companion document elsewhere in this project's research base cautions against summing large anti-poverty asks (a blanket OW increase and related income-support proposals, together modelled around $5.25B/year) into any homelessness-specific ROI total, on the grounds that they are not homelessness-budget offsets. This document's own $1.14B/year consolidated figure does not include those costs, and this document flags the distinction explicitly so no derivative document conflates the two.

Modelled financial ROI versus unmonetized-until-recently costs. The mortality-monetization figure ($578.2M/year, using Treasury Board's own VSL standard) is larger than every other single line item in this cost case, yet it was only recently converted from raw mortality statistics into cost-benefit terms and is not yet integrated into the consolidated $1.14B/year headline total above. This is a live, acknowledged gap rather than an oversight silently carried forward.

What the evidence does and doesn't support

Well-supported (audited or independently cross-validated):

Thin or contested:

International context

1. Treaties/frameworks touched. This page’s own scope — modelled fiscal ROI by order of government — does not itself engage a specific international human-rights treaty article the way a rights-framework or homelessness-measurement page would; the underlying right to adequate housing (ICESCR Article 11(1), domesticated via the National Housing Strategy Act, per homelessness-federal-funding-architecture.md's own primary-source-fetched treaty section) is the relevant instrument for the population this document costs out, but this page’s own contribution — cost-benefit modelling, fiscal misalignment across orders of government — is a fiscal/economic question, not a treaty-compliance one. Named plainly rather than manufacturing a treaty connection that does not fit this page’s actual scope, consistent with homelessness-political-economy.md's own International context section making the identical scope-based call for its adjacent page.

2. Best global comparators. This document's own "New this review" sections already carry this page’s two most load-bearing international/comparator findings in full depth: the Rao & Brandeau JAMA Network Open mathematical-modelling study (a US$96,000-per-person modelled Housing First intervention, US$26,800 per QALY gained, a second independently-derived health-economics methodology reaching the same directional conclusion as the Canadian dollar-offset evidence) and the AMO/HelpSeeker Ontario-wide funding-versus-need finding — the latter a domestic, not international, comparator, restated here only to note it is not being double-counted as an international one. For a genuinely international fiscal/cost-effectiveness comparator, Finland's Housing First model is this corpus's most consistently cited example (see homelessness-housing-first-model.md for the authoritative EUR 21,000/15,000-per-person-per-year savings-figure reconciliation, and homelessness-federal-funding-architecture.md for the Y-Säätiö institutional model) — this document does not repeat that figure set, since its own scope is Toronto's fiscal misalignment specifically, not a comparative national cost-effectiveness case, and cites it here only by name per this template's claim-anchored, not backgrounder-anchored, discipline.

3. What Toronto/Ontario can steal shamelessly. Stated descriptively, not as a recommendation: this page’s own Rao & Brandeau finding models cost-effectiveness using quality-adjusted life years (QALY) — a standard health-economics measurement approach not otherwise used anywhere in this page’s Canadian dollar-offset evidence base (which uses net-cost-after-offset figures, not QALY). The specific, nameable transferable element is the measurement methodology itself — QALY-based cost-effectiveness analysis — as a second, methodologically distinct way of stating the same underlying finding (Housing First is cost-effective), which could in principle be applied to Toronto-specific data if a future study attempted it; the researchers themselves are, per this document's own citation, "still seeking to expand into Toronto," meaning this is a live, nameable opportunity rather than a closed comparator.

Cui Bono — who profits from this problem persisting

Draft note: the sourced findings below are published pending independent legal review, which is currently under solicitation. Every row is a pointer to a named, already-published source finding — never this document's own allegation. This note is removed when legal review completes.

Per the Accountability Observatory's charter (Prime Rule: pointer, never author) and this library's standard page structure's binding requirement — this document identifies zero beneficiary entities directly, and points instead to homelessness-political-economy.md's own Cui Bono table, which is the corpus's binding source for this question. This page’s own scope is explicitly fiscal-modelling and cost-benefit accounting across orders of government — "who pays, who saves, and by how much" — not "which named entity profits from the current system's design," a distinct question this page’s own Scope section hands off, alongside the clinical/programmatic Housing First case and shelter operating costs, to sibling pages. homelessness-political-economy.md's own Cui Bono table identifies two ESTABLISHED entities directly relevant to this page’s own cost lines — Garda Canada Security Corporation/GardaWorld (an $11.9 million encampment-support contract, per ACL-0009) and two anonymized hotel operators ("Hotel C"/"Hotel D") found by the Toronto Auditor General's 2022 audit to have overcharged the City for vacant emergency-shelter rooms (no registrable entity — the Auditor General's own report anonymizes the operator(s), so no registered accountability claim is registered for this row) — both bearing directly on this document's own "mortality, monetized" and "municipal legal-liability exposure" cost lines (the same 2022 hotel-program audit this document's own integration pass draws its cost figures from is the source of the political-economy page’s hotel-operator row). This document does not re-list those rows as its own table, per this template's pointer-never-author discipline against re-deriving another page’s already-graded findings — a reader looking for the entity-level "who profits" answer specific to Toronto's homelessness-response spending should go to homelessness-political-economy.md directly, not here, exactly as that page’s own Cui Bono section already directs readers away from itself for the comparative-cities question in reverse.

Open questions / data gaps

Source / citation index

Grouped by section; carried-forward sources carry the trust tier assigned in this library's prior executive synthesis (ROI Lock Memo); NEW sources carry their own inline source quote above.

Fiscal misalignment framing — carried-forward, this library's prior synthesis document (Economic ROI) "The Master Argument" / "The jurisdictional tragedy of the commons."

Federal ROI — carried-forward Tier 1: IHAP Toronto 2024 actual ($261.87M, TSSS 2024 Annual Report); IHAP 2026 budget ($97M, TSSS 2026 Budget Notes). carried-forward Tier 2: federal exposure/savings/NPV modelling (Meta methodology). carried-forward, independently corroborated: Gaetz/Gulliver/Richter State of Homelessness in Canada 2014; PBO "Federal Spending to Address Homelessness," May 2024.

Provincial ROI — carried-forward Tier 2 throughout (OW cost modelling, Housing First provincial share, Housing Stability Supplement proposal).

Municipal ROI — carried-forward Tier 1: Toronto AG 2025 shelter cost ($136/night); Latimer et al. 2020 Psychiatric Services (net/gross HF cost, 69% offset). carried-forward Tier 2: non-shelter municipal savings lines, prevention-ROI modelling.

Productivity/GDP — carried-forward Tier 2/single-model average; October 2024 Street Needs Assessment population figure (Tier 1).

Mortality monetization / legal liability — carried-forward from the full-cost-accounting integration pass; Treasury Board of Canada Secretariat VSL figure is itself Tier 1 in the backbone's own tiering. Corrected/added (a later verification pass, 2026-07-13): 59-death figure now directly cited to City of Toronto, "Annual Review of Statistics on Deaths of Shelter Residents" (accessed 2026-07-13); $50M class-action figure now traced to CTV News/Canadian Press (2025-06-04) and Global News/Canadian Press (2025-06-04), both accessed 2026-07-13 — a refugee-shelter-bed-denial suit, distinct from the Lamport Stadium encampment-clearance suit. Independently re-verified 2026-07-16: both sources live-fetched directly (literal URLs added above, replacing the prior CP24/CBC citation which named the correct underlying wire story but pointed at two systematically JS-gated outlets); all facts confirmed exact — filed June 4/5 2025, $50M + special damages, Nov 7 2022–Oct 1 2023 denial period, lead counsel Linette King, City spokesperson Elise von Scheel's "in due course" statement, and the disproportionate-impact-on-Black-community allegation.

AMO/HelpSeeker provincial funding-growth finding — NEW. HelpSeeker Technologies, "Municipalities Under Pressure One Year Later: An Update on the Human and Financial Cost of Ontario's Homelessness Crisis," published 2026-01-13, https://www.helpseeker.org/reports/municipalities-under-pressure-one-year-later, accessed 2026-07-13.

Rao & Brandeau QALY finding — NEW. University of Toronto, "'Housing-first' more effective than 'treatment-first' for addressing addiction, homelessness: Study," published 2025-07-11, https://www.utoronto.ca/news/housing-first-more-effective-treatment-first-addressing-addiction-homelessness-study, citing Rao & Brandeau, JAMA Network Open, accessed 2026-07-13.

International context / Cui Bono (added 2026-07-14) — ICESCR Art. 11(1)/NHSA, scope-based non-engagement stated per homelessness-political-economy.md's own identical call; Rao & Brandeau QALY methodology (already cited above) as this page’s own international comparator; Cui Bono findings pointed to homelessness-political-economy.md's table (ENT-0009/ACL-0009; the "Hotel C"/"Hotel D" row carries no registrable entity, since the Auditor General's own report anonymizes the operator(s)) rather than re-derived here.