The Full Cost of Homelessness Across Government

Hospital stays, jail time, deaths — the homelessness costs that never show up on a shelter budget line, tallied up.

Claim coverage as of 2026-07-16 (a later verification pass): 1 carried-forward document group (this library's prior synthesis document (Full Cost Accounting), itself a merge of three source documents — this library's prior synthesis document (Funding Stakeholder ROI Matrix) and this library's prior derived-findings document (Follow The Money) — per its this page’s own scope note), mined for its costed, sourced findings — its campaign/advocacy voice and any political-strategy framing excluded per this project's editorial-scope rule — plus 2 new 2026 primary-source findings from the 2026-07-14 pass's live discovery (NEW-2026-1, NEW-2026-2), cited inline with source quotes, not yet through this library’s formal verification process. This review (2026-07-16) restored several genuine coverage gaps identified in the G3 rebuild-covenant coverage adjudication — see that page's coverage checklist for the full per-fact claims register. Coverage evaluated this review — see coverage checklist. Cui Bono: 0 beneficiary entities identified this review — this page’s own scope is cross-departmental cost accounting, not a "who profits" inquiry; per this template's own guardrail, an empty table with an honest explanation is the correct output here, not a defect. See "Cui Bono" section below.

Written per this library's standard page structure, a later review, 2026-07-14. The carried-forward sources document is itself unusually disciplined for a v1-era source — it already distinguishes sourced figures from estimates, flags its own prior corrections, and explicitly declines to monetize several categories (mental-health/trauma burden, intergenerational effects, community cohesion) rather than manufacturing dollar figures for them. This backgrounder inherits that discipline rather than loosening it, and adds only what a fresh 2026 live-discovery pass could independently confirm or update.

Scope

This page’s neutral scope question, per that page's own internal recordsthis page’s own scope note: what does homelessness cost across every government level and ministry, financial and non-financial, that isn't already captured in a shelter-operations budget line — health system costs, corrections/justice costs, municipal legal exposure, mortality (monetized via an official government methodology), and non-financial costs that resist monetization. This document covers: municipal-level costs beyond shelter operations (encampment enforcement, legal liability, fire services); provincial-level costs by ministry (health, social services, corrections, education); federal-level costs; mortality monetized via Canada's official Value of a Statistical Life methodology; and non-financial costs explicitly presented as real but not monetized. It hands off, rather than duplicates: the specific dollar savings/ROI case for Housing First interventions to homelessness-economic-roi (this page cites that page’s own figures rather than re-deriving them); Toronto shelter operating/capital budget mechanics to shelter-system-capacity-strain; and funding-structure incentive questions to homelessness-political-economy.

Current state

The visible system: who funds it now, and the cost hiding in a different budget line

Restored 2026-07-16 (a later verification pass). Toronto Shelter and Support Services' 2024 actual budget was $796.40M CL-150141, funded federal (IHAP) $261.87M/33%, provincial (HPP base + one-time "New Deal" top-up) $314.76M/39%, City of Toronto tax base $194.74M/24%, and other/in-kind ~$25M/3% — meaning Toronto taxpayers fund less than a quarter of the visible shelter system directly, with the remaining three-quarters arriving as federal and provincial transfers that were, as of the carried-forward sources's writing, both shrinking (the IHAP cliff and the COHB cliff) while the City absorbs the operational consequences [This library's prior synthesis document (Funding Stakeholder ROI Matrix) Part 1.1]. A separate, genuinely different cost hides in the Province's own healthcare budget rather than any homelessness-specific line: the carried-forward sources's own derived (not primary-sourced) arithmetic applies an excess-healthcare-cost figure of $10,440/year/person (the $12,209/year-homeless vs. $1,769/year-housed differential, ICES/Homeless Hub 2024) CL-150153 to the 15,418-person shelter-engaged population, producing approximately $160.96 million/year in excess provincial OHIP spending that does not appear in any homelessness budget document [carried-forward, ⚠️ this $161M figure is original synthesis/derived arithmetic on top of a primary-sourced per-person figure, not itself a primary-sourced total — label as such in any public use]. The carried-forward sources frames this as roughly half of the Province's entire HPP contribution to TSSS, arriving through a different ministry where it is harder to see and politically easier to ignore.

Restored 2026-07-16 (a later verification pass). A further cost category the carried-forward sources identifies as "completely externalized": Toronto's estimated 20,000–30,000-person hidden/provisionally-accommodated population (couch-surfing, informal overcrowding) is not costing any government anything directly, because their housing cost is currently being absorbed by host families and informal networks rather than the $49,640/year shelter-bed alternative [This library's prior synthesis document (Funding Stakeholder ROI Matrix) Part 1.4]. The carried-forward sources is explicit this is not a finding about generosity but about who is subsidizing a housing-affordability failure: a real, unmeasured, uncompensated cost (lost privacy, overcrowding-related stress, host-household financial strain, foregone informal rental income) disproportionately borne by lower-income households themselves, and a fragile arrangement that can collapse into the formal (and far more expensive) shelter system at any time. Neither the excess-healthcare figure nor the informal-subsidy population is included in the "visible system" total above; both remain unquantified-into-a-single-total by design, consistent with this document's discipline against fabricating a combined figure the underlying sources don't support.

Municipal-level costs beyond shelter operations

Encampment clearance and enforcement. The City's own disclosed figures for a 2021 clearing of three parks (Trinity Bellwoods, Alexandra Park, Lamport Stadium) total approximately $2 million, covering policing/security, fire, paramedics, transportation, waste management, and other operational costs, with two specific line items broken out: fencing at approximately $357,000 (since removed) and turf/landscaping restoration at $792,668, with the City's own acknowledgment that "full turf restoration will take more than one season" CL-150281 [From this library’s earlier research]. This single-event figure is not an annual recurring total; the carried-forward sources separately identified a standing security-guard-services contract specific to encampment support, confirmed via a City procurement report at $2,937,168/year net of tax ($3,319,000 including HST) CL-150583 as of a September 2024 Bid Award report — a genuinely different kind of figure (an ongoing annual contract, not a one-time event cost) [From this library’s earlier research]. This review's live discovery directly confirms and updates this finding: as of December 2025 reporting, Toronto's complete shelter/encampment-relevant private-security contract portfolio totals $109 million in contracted ceiling value across three shelter/encampment-specific contracts (shelter respite centre security, $44.8M; shelter core site security, $15M; Garda Canada Security Corporation "encampment support," $11.9M), with $35.2 million actually spent in 2025 alone across the City's full eight-contract private-security portfolio [NEW-2026-1]. This confirms the carried-forward sources's own tentative identification of Garda Canada as "very likely" the same contract previously known only by its dollar value — this review's directly-fetched source independently corroborates that identification by naming the same $11.9M figure directly against Garda Canada Security Corporation. Restored 2026-07-16 (a later verification pass): a further, live cost driver the carried-forward sources names but does not resolve: Council (2024.EC13.8) directed TSSS to report through the 2025 Budget process on the cost of "up to three concurrent 24/7 Community Safety Teams for large encampments," confirming this is treated internally as a real, distinct, budgetable cost category separate from base shelter operations [From this library’s earlier research]. Whether that costed report was ever produced, and what figure it contained, was not found by the carried-forward sources or independently checked this review — carried forward as the source's own honest open item rather than presented as resolved.

Restored and independently verified 2026-07-16 (a later verification pass) — a fifth large named City contract. The Canadian Red Cross Society holds Non-Competitive Blanket Contract 47024284 for lodging services for asylum seekers and refugees; City Council authorized increasing its value from $60,137,629 to $74,095,341 (both net of HST) CL-150221 and extending its term to December 31, 2025, "to maintain service levels in the winter and early spring" [This library's prior derived-findings document (Follow The Money) Tier 1 Gap Progress section; independently re-verified this review directly against Toronto City Council's own decision record, Item 2024.EC17.4, adopted December 17-18, 2024, which confirms this exact contract number and both dollar figures]. This is comparable in scale to Fred Victor's own confirmed government funding and larger than Dixon Hall's, and — notably — it is non-competitive, reflecting the emergency nature of the refugee-lodging response rather than the competitive procurement process used for most other shelter-operator agreements.

Municipal legal liability. Five plaintiffs are suing the City and Toronto Police Services Board over the July 2021 Lamport Stadium clearance (traumatic brain injuries, lacerations, police-baton injuries, PTSD alleged) CL-150282, with damages reported inconsistently across outlets (more than $100,000 each per one outlet, versus $300,000-$350,000 each for four of five plaintiffs per another) — a discrepancy the carried-forward sources states rather than resolves [From this library’s earlier research]. A separate $50 million class action was filed in 2025 over the systemic exclusion of refugee claimants from shelter beds (Nov 2022-Oct 2023), alleging negligence, breach of duty, and Charter s.7/s.15 violations CL-150213 [From this library’s earlier research]. Two judicial findings — Black et al. v. Toronto, 2020 ONSC 6398, and Sanctuary et al. v. Toronto, 2020 ONSC 6207 — represent realized litigation costs (City legal-defence resources) independent of any damages [From this library’s earlier research]. No aggregate, homelessness-specific municipal legal/settlement cost total exists in any source found by the carried-forward documents or this review; individual exposures are presented separately rather than summed into a fabricated total, consistent with this project's citation discipline. Restored 2026-07-16 (a later verification pass), general context, not itself a homelessness-specific cost: the carried-forward sources separately notes Ontario municipalities generally are reporting general liability insurance premium increases "of more than 20%" in recent years, driven by a hard insurance market, climate change, and rising litigation, under a joint-and-several-liability framework that can leave a municipality responsible for damages "even if they are deemed just one per cent responsible" [Citing AMO]. This is cited here only to explain why a municipality carrying multiple live, homelessness-related claims faces compounding, not merely additive, insurance-market exposure — not as a homelessness-specific cost line in its own right.

Fire services. Toronto Fire Services operates a dedicated Encampment Strategy Team conducting daily fire-risk site visits and fire-safety education, confirmed as a real, ongoing, named program, but with no cost isolated from Toronto Fire Services' overall budget ($574.3 million gross 2025, per the carried-forward sources) in any source checked [From this library’s earlier research].

Restored 2026-07-16 (a later verification pass) — COVID-era cost structures that became permanent. The carried-forward sources traces, through the City's own budget documents, a pattern of pandemic-era operational spending that was never wound back rather than temporary: the Temporary Hotel Program, created explicitly "as a response to the COVID-19" pandemic to provide physical-distancing capacity the base shelter system couldn't, required $278.052 million to sustain by 2024 — still framed as pandemic-response funding roughly six years after the acute pandemic ended, with the 2026 TSSS budget showing only $50.469 million in projected savings from planned hotel-site closures CL-150308, meaning the bulk of the program's cost structure persisted well past its original justification [This library's prior synthesis document (Full Cost Accounting) Part 13]. The City's own 2024 budget notes state plainly: "Beginning in 2024, the sustained impacts of the pandemic on Shelter Services is now being funded through the City's property tax base" CL-150309 — a direct admission that pandemic-era costs were absorbed into permanent, ongoing municipal operating funding, on the record, in the City's own words [From this library’s earlier research]. In 2023 the City reduced physical-distancing bed spacing to unlock roughly 400 additional beds under capacity pressure, while explicitly stating infection-prevention-and-control (IPAC) protocols "will remain in place" — one COVID-era measure rolled back under pressure, another deliberately retained, which the carried-forward sources reads as a genuine ongoing choice rather than blanket inertia [From this library’s earlier research]. One specific related claim the carried-forward sources researched directly and could not confirm: whether Toronto's shelter system switched from washable to disposable dishware/linens as a distinct, named COVID-era cost driver — general infection-control literature and a comparable Toronto hospital-sector finding (reusable isolation gowns saving hospitals an estimated $70 million over two pandemic years) CL-150333 make the claim plausible, but no Toronto shelter-specific source confirms it, and the carried-forward sources flags this explicitly as unconfirmed rather than assumed true because it fits the broader pattern [carried-forward, ⚠️ still being checked — not independently re-checked this review].

Provincial-level costs, by ministry

Health (Ministry of Health). New this review, directly fetched and independently confirmed: the Canadian Institute for Health Information's own national data (2022-2023 hospitalization data, released April 25, 2024) finds the average cost of a hospitalization for a patient experiencing homelessness (identified via ICD-10-CA code Z59.0) is more than double the cost of an average hospital stay in Canada: $16,785 versus $7,803 CL-150151, with average length of stay at 15.4 days versus 8.0 days nationally, and 93% of homeless-patient admissions arriving via the emergency department (a proxy for inadequate primary-care access), the most common reasons for hospitalization being substance use disorders (18%), schizophrenic disorders (11%), and cellulitis (7%) [NEW-2026-2]. Precision note relative to the carried-forward sources: the carried-forward documents characterizes this relationship as "2.1× national average," which is directionally correct (the $16,785 figure is indeed roughly 2.15× the $7,803 figure) but should be read precisely as CIHI's own national-level comparison, not an Ontario-specific or Toronto-specific figure — this review did not locate a CIHI or ICES breakdown isolating Ontario or Toronto from the national $16,785 figure, so any Ontario-specific dollar claim beyond this national figure should be treated as unconfirmed pending a further ICES-specific search.

Children, Community and Social Services (MCCSS). The Financial Accountability Office of Ontario's finding of a structural underfunding gap relative to the Ministry's own cost-driver projections is carried-forward at $1.5 billion (2025-26) rising to $3.1 billion (2027-28) [From this library’s earlier research]. This review's live discovery independently confirms this figure directly from FAO reporting: the FAO projects MCCSS spending will decrease at an average annual rate of -0.5% (from $20.7 billion in 2024-25 to $20.4 billion in 2027-28) against a needed cost-driver growth rate, producing a funding shortfall of $1.5 billion in 2025-26, $2.4 billion in 2026-27, and $3.1 billion in 2027-28 CL-150215 — an independent, directly-sourced re-confirmation of the carried-forward documents's own figure, with the added precision of the interim $2.4 billion 2026-27 figure [NEW-2026-2 — note this label is reused per this document's own numbering convention below; see source quote for the FAO-specific citation distinct from the CIHI citation above].

Attorney General / Tribunals Ontario. Landlord and Tenant Board adjudicator costs are carried-forward at 133 adjudicators, approximately $19.5 million/year payroll as of March 2025, with the carried-forward sources's own finding being a productivity collapse rather than a headcount shortage — not independently re-verified this review [carried-forward, ⚠️ still being checked].

Solicitor General / Corrections. A provincial incarceration cost of $326/day ($118,990/year) is carried-forward, alongside the John Howard Society's figure of 7,455 annual releases with no fixed address CL-150187 [carried-forward, ⚠️ still being checked]. This review's live discovery could not independently confirm a current Ontario-specific per-diem figure — Ontario's own open-data portal for this specific metric currently states the per-diem data "is not available right now" and is under review for public release, and the most recent comparable figures this review located were a Canada-wide (not Ontario-specific) $326/day average (2022, Statistics Canada) and a federal (not provincial) $436/day figure (2025, Correctional Service Canada, a different jurisdiction from Ontario's provincial system) [NEW-2026-2]. This is a genuine data-access gap independently confirmed this review, not merely inherited — Ontario's own per-diem figure for its provincial correctional system is not currently publicly available in the sources checked, which affects confidence in the carried-forward sources's own $326/day figure without more clearly establishing whether that figure is Ontario-specific or itself the Canada-wide average being applied to Ontario.

Education, and Children's Aid/child welfare. Both remain genuinely unquantified for Toronto/Ontario specifically — the carried-forward sources flags both as real but uncosted gaps rather than estimating unsupported figures, and this review did not locate a source closing either gap [From this library’s earlier research].

Bill C-12 (federal-provincial policy development). Received Royal Assent March 26, 2026, introducing new eligibility bars for refugee claimants filing more than 12 months after entry or crossing irregularly and filing more than 14 days later; IRCC had sent procedural fairness letters to 30,000 asylum seekers as of April 2026, against a cited 196,000-case national asylum backlog CL-150284 [carried-forward, ⚠️ still being checked — not independently re-checked this review]. Given refugee claimants' documented share of Toronto shelter users (cited elsewhere in this project's leaves), this is flagged as a real, forward-looking cost driver rather than a currently-realized cost.

Federal-level costs

Interim Housing Assistance Program funding ($261.87 million 2024 actual, declining to a cited $97 million 2026 budget figure — independently consistent with this review's own finding elsewhere in this project's Batch-3 leaves of a $97.078 million 2026 federal refugee cost-share figure), the Homelessness Services Capital Infrastructure Strategy's capital request ($674.5 million), and Reaching Home's Toronto allocation (cited at $58 million/year) are all carried-forward from the carried-forward sources [carried-forward, ⚠️ still being checked — not independently re-checked this review beyond the cross-consistency noted].

Tax revenue and workforce transition: the two questions a full-cost accounting has to ask honestly

Restored 2026-07-16 (a later verification pass). A full cost accounting has to test, not assume, whether housing outcomes generate offsetting tax revenue. The carried-forward sources researched this directly rather than assuming a simple "housed → employed → paying taxes" model, and found the opposite of what that model would predict: the At Home/Chez Soi trial — the same large, five-city Canadian RCT this project's Housing First evidence otherwise relies on — separately tested employment and income effects (Poremski et al. 2016, Psychiatric Services) and found Housing First participants had lower odds of obtaining competitive employment than the control group (unadjusted odds ratio 0.68 high-needs, 0.73 moderate-needs, both statistically significant), with no statistically significant effect on income from any source CL-150296 [This library's prior synthesis document (Full Cost Accounting) Part 10]. The carried-forward sources's own conclusion is that a direct tax-revenue claim from formerly-homeless individuals' future earnings is not currently supportable by the evidence and should not be made — the defensible fiscal case remains avoided cost, not new tax revenue. Two narrower, evidence-based exceptions the carried-forward sources did find support for: (1) adding Individual Placement and Support (IPS), a supported-employment model with its own separate evidence base, more than doubled the odds of securing employment in the At Home/Chez Soi data during periods of high implementation fidelity (odds ratio 2.42, 95% CI 1.13–5.15, p=0.02) CL-150299 — rated low-certainty evidence, conditional on fidelity to the model being funded and measured, not assumed [carried-forward, ⚠️ still being checked]; (2) new jobs created by implementing recommendations themselves (peer navigator positions, expanded case management, capital-program construction employment) generate real, distinct income-tax revenue independent of any claim about formerly-homeless individuals' own earnings — for one costed example, 200 peer-navigator positions at $55,000/year were calculated to generate $1,602,611/year in new combined federal/provincial income tax CL-150602 [From this library’s earlier research].

Restored 2026-07-16 (a later verification pass) — workforce transition. A related honest question: if this system's own cost drivers were addressed and fewer people cycled through emergency shelter, what happens to the shelter-sector workforce currently employed running it? The carried-forward sources's answer is transition, not elimination — Housing First shifts the type of labour needed (toward case management and housing-location support) rather than eliminating the need for it, since permanent supportive housing is itself a staffed model [This library's prior synthesis document (Full Cost Accounting) Part 11]. The one primary-sourced anchor figure: TSSS's own 2025 Budget Notes confirm 1,497.5 total approved divisional positions for 2025 CL-150665 (up from 1,064.9 in 2023), covering the entire division (directly-operated shelters, POS oversight, outreach, encampment response, administration) rather than directly-operated-shelters staff specifically — a scope distinction the carried-forward sources flags explicitly, and which does not include the larger purchase-of-service contractor workforce (see shelter-operator-financial-profiles and the sector-wide employment analysis it and this project's an earlier synthesis document substrate separately document) [carried-forward, ⚠️ still being checked — this review did not independently re-check the 1,497.5 figure or its 2026 trend]. The carried-forward sources is explicit that a precise, confirmed headcount of who transitions where remains unresearched and should not be asserted with false precision.

Mortality, monetized via Canada's official Value of a Statistical Life

This review's live discovery independently re-confirms the carried-forward sources's central mortality-monetization figure. The Treasury Board of Canada Secretariat's Cost-Benefit Analysis Guide for Regulatory Proposals — the standard all federal departments and agencies are instructed to use — sets the Value of a Statistical Life (technically, value of mortality risk reduction) at $6.5 million in 2007 dollars, updated in the current guide to approximately $9.1 million in 2020 dollars, and independently cross-calculated (via the Bank of Canada inflation calculator, consistent with the carried-forward sources's own method) to approximately $9.8 million in 2026 dollars CL-150491 [NEW-2026-1]. A 2022 peer-reviewed meta-analysis in Canadian Public Policy recommends an updated, higher figure (cited in the carried-forward sources at $13 million, 2020 CAD) — flagged in both the carried-forward sources and this review as the more current academic view, while the $9.8 million figure remains the one government bodies are actually instructed to use, which is why the carried-forward sources (and this backgrounder) leads with it [carried-forward, corroborated NEW-2026-1].

Applying this figure to Toronto shelter-system mortality data cited elsewhere in this project (this page does not re-derive that mortality count, which is shelter-system-capacity-strain's and homelessness-mortality-overdose-data's own owned figure): the carried-forward sources's own calculation, using a 2024 figure of 59 shelter-resident deaths CL-150672, produces $578.2 million for that single year alone at the $9.8 million VSL rate [carried-forward, ⚠️ still being checked — this page does not independently re-verify the underlying 59-death count, which belongs to a sibling leaf; it re-states the carried-forward sources's own arithmetic using that count]. The carried-forward sources's own necessary caveats are inherited in full: VSL is a willingness-to-pay regulatory construct, not a literal price on a human life; a cumulative multi-year total applying a single present-day VSL rate retroactively is a simplification that should be read as an illustrative order-of-magnitude, not a precise total; and these are not deaths solely attributable to a specific policy choice, but documented mortality within a population this project has separately shown is failed by current policy [From this library’s earlier research].

Restored 2026-07-16 (a later verification pass) — the carried-forward sources's own standard for when an extrapolation is legitimate. The mortality-monetization calculation above is, by the carried-forward sources's own account, a worked example of a broader discipline worth stating explicitly: an extrapolation (multiplying independently-sourced inputs together, such as a documented death count and a government's own VSL rate) is legitimate only when every multiplied input already has its own primary source — multiplying two unsourced guesses together does not produce a sourced estimate, it produces a compounded guess [This library's prior synthesis document (Full Cost Accounting) Part 8]. The carried-forward sources applies this standard against itself: a worked attempt to extrapolate police-time costs for encampment-related calls (using a per-event cost, an assumed call volume, and an assumed hourly compensation rate) was found to fail this test, because no reliable source exists for the annual call-volume input specifically — the same conclusion this project reached when an earlier $48M/year Toronto Police mental-health-call figure and an 8.1× ambulance-utilization figure were both independently found unverifiable and formally removed. The carried-forward sources declines to manufacture a police-time cost figure for the same reason, flagging it as a genuine, currently unquantifiable cost category rather than estimating one into existence [From this library’s earlier research].

Restored 2026-07-16 (a later verification pass) — real costs the carried-forward sources deliberately does not monetize. Consistent with the VSL discussion above (a real cost made visible through careful monetization) is its necessary opposite: the carried-forward sources is explicit that not every real cost belongs in a spreadsheet, and forcing one into an invented dollar figure would be the same kind of overreach its own citation discipline exists to prevent [This library's prior synthesis document (Full Cost Accounting) Part 5]. Four categories are named as real, documented, and deliberately not monetized: the mental-health and trauma burden of homelessness itself (beyond the healthcare-utilization cost already counted above), which includes encampment clearances and the lived experience of chronic housing instability; intergenerational effects (children in family homelessness facing documented elevated risk of later-life homelessness, interrupted education, and child-welfare involvement, compounding across a generation); community and social cohesion (encampment visibility correlates with documented increases in 311 complaints and community anxiety, but no Toronto-specific quantifying data was found, and the source explicitly declines to develop this further given the risk of treating "community discomfort" as commensurable with unhoused people's basic survival needs); and reputational/intergovernmental-relationship costs (plausible but genuinely unquantifiable effects on tourism, talent attraction, and investment perception, with no methodologically sound way to isolate homelessness's specific contribution). [From this library’s earlier research]

Toronto: the case for and against

Section merged 2026-08-11 from a companion Toronto-specific brief (Lane L2a Toronto brief-merge pass).

FOR — the true cost of homelessness is far larger, and more government-legible, than shelter budgets alone:

• The roughly $109 million private-security contract portfolio directly tied to Toronto shelters and encampments, independently confirmed via December 2025 City-supplied data, demonstrates that homelessness-adjacent costs recur well beyond the shelter operating budget most public debate focuses on [NEW-2026-1].
• CIHI's own national data shows patients experiencing homelessness cost the healthcare system more than double the average hospital stay ($16,785 vs. $7,803), driven substantially by emergency-department-first access patterns (93% of admissions) consistent with inadequate primary-care access — a cost currently absorbed by the health system, not the homelessness-response system [NEW-2026-2].
• The FAO's own projection of a widening MCCSS structural funding gap ($1.5B → $2.4B → $3.1B, 2025-26 through 2027-28) is an independent government forecasting body's own figure, not an advocacy estimate, and directly implicates the social-services infrastructure homelessness prevention depends on [NEW-2026-2].
• Canada's own required regulatory cost-benefit methodology (Value of a Statistical Life, currently approximately $9.8 million in 2026 dollars) makes a $578.2 million single-year mortality cost visible in terms directly comparable to any other government budget line, using a number the government itself, not any advocacy project, says is the correct one [carried-forward, corroborated NEW-2026-1].

AGAINST — several headline figures carry real jurisdictional-precision or data-availability limits that caution against overstatement:

• The CIHI hospitalization-cost differential is a national figure, not an Ontario- or Toronto-specific one — no source located in this review isolates the Ontario or Toronto figure from the Canada-wide $16,785 average, and presenting it as an Ontario-specific cost (as opposed to a plausible-but-unconfirmed local proxy) would overstate what CIHI's own release actually establishes [NEW-2026-2].
• The $326/day Ontario correctional per-diem figure may itself be a Canada-wide average mistakenly presented as Ontario-specific — this review found Ontario's own official per-diem dataset is currently unpublished (the province's own open-data portal states the data "is not available right now" pending review), while an independently-located Canada-wide 2022 Statistics Canada figure is identical in value ($326/day) to the figure the carried-forward sources cites as Ontario's — a genuine, unresolved provenance question this brief flags rather than assumes away [NEW-2026-2].
• The cumulative multi-year mortality-monetization figure is explicitly a simplified illustrative calculation, not a rigorous total — it applies a single present-day VSL rate retroactively across many years without discounting or year-appropriate historical VSL figures, a limitation the carried-forward sources itself states plainly; a hostile reviewer's most obvious and valid critique of this project's mortality-cost figures targets exactly this calculation, and this brief does not present the cumulative figure with more confidence than that caveat supports [From this library’s earlier research].
• No aggregate municipal legal-liability total exists, and none should be constructed — the individual exposures (the Lamport Stadium suit, the $50 million refugee-exclusion class action, two prior judicial findings) are each real and separately documented, but summing them into a single "legal cost of homelessness" figure would fabricate a precision the underlying data does not support, and both the carried-forward sources and this backgrounder decline to do so.
• Several of this page’s costed categories (education, child welfare) remain entirely unquantified, meaning any total-cost figure built from this page’s contents is necessarily an undercount relative to the true full cost — a limitation worth stating plainly rather than implying the documented categories represent the complete picture.

Both sides here point to real findings; the honest synthesis is that the true cost of homelessness is very likely understated by any single figure in this page, not overstated — but several of the specific figures carrying the most rhetorical weight (Ontario-specific healthcare and corrections costs) currently rest on jurisdictional-precision gaps this brief will not paper over.

Toronto-specific figures:

Toronto-relevant precedents: No non-Toronto Canadian municipality's own comparable multi-ministry, VSL-inclusive full-cost-accounting exercise was identified or independently verified in this review. The Calgary Homelessness Foundation's cost-comparison literature (institutional response $66,000-$120,000/year vs. supportive housing $13,000-$18,000/year), documented in this project's sibling homelessness-political-economy leaf, is the closest comparable cost-accounting precedent found, but addresses a narrower comparison than this page’s full cross-departmental scope and is not repeated here per this page’s own scope discipline.

Toronto bottom line: Homelessness's true fiscal footprint spans at least five government levels and includes a mortality cost, calculated via Canada's own required methodology, larger than several major program budget lines already documented in this project — but two of the specific figures carrying the most weight in existing homelessness-cost literature (Ontario-specific healthcare and corrections per-diem costs) rest on jurisdictional-precision gaps this review could not close, and should be cited with that caveat attached rather than as confirmed Ontario-specific figures.

Toronto-specific uncertainties:

• Whether Ontario's correctional per-diem cost is genuinely distinct from the Canada-wide average is unresolved; Ontario's own dataset is not currently public.
• Whether the CIHI hospitalization-cost differential holds at the same magnitude specifically for Ontario or Toronto is unconfirmed; only the national figure is directly sourced.
• The cumulative multi-year mortality-monetization figure is explicitly illustrative, not a rigorous year-appropriate-discounted total (see Card a recommendation card for the proposed fix).
• All NEW-2026-# findings in this brief are directly quoted from primary sources but have not yet been run through this project's formal claim-verification and registration process.
• Education-system and child-welfare-system costs specific to Toronto/Ontario homelessness remain entirely unquantified in any source found.
• No aggregate municipal legal-liability total exists and none is asserted; individual exposures are stated separately.

Key tensions / tradeoffs

A single-event clearance cost against a now-confirmed standing security-contract total. The 2021 three-park clearance's ~$2 million one-time cost and the newly-confirmed $109 million multi-year private-security contract portfolio (of which $44.8M+$15M+$11.9M = $71.7 million is shelter/encampment-specific contract value) measure fundamentally different things — an incident cost versus an ongoing contract ceiling — and should never be summed into a single "encampment cost" figure, a discipline this document maintains explicitly rather than implicitly [carried-forward, corroborated NEW-2026-1].

A national healthcare-cost figure standing in for an Ontario-specific one. The CIHI $16,785-versus-$7,803 hospitalization-cost finding is a genuinely strong, current, directly-sourced figure — but it is national, not Ontario- or Toronto-specific, and this review found no source that isolates the Ontario or Toronto figure from the national average. Any downstream document citing this figure as an Ontario-specific cost differential would be overstating what CIHI's own release actually establishes; this document states the distinction explicitly rather than letting the "2.1× national average" framing imply more geographic precision than the source supports.

A provincial corrections per-diem figure whose own currency and jurisdiction could not be confirmed. The carried-forward sources's $326/day Ontario incarceration-cost figure is structurally identical in value to the Canada-wide (not Ontario-specific) 2022 Statistics Canada average this review independently found — raising a real, unresolved question about whether the carried-forward sources's "Ontario" figure is in fact the national average being applied to Ontario without a distinct Ontario-specific source. This is flagged as a genuine data-provenance concern, not resolved either way, since Ontario's own official per-diem dataset is not currently public.

Flagged, not silently reconciled (2026-07-16 G3 pass) — a genuine cross-document contradiction on "shelter cost per night." This page’s own merge-group source material contains three distinct per-night/per-year shelter-cost figures that must not be conflated: (1) the Toronto Auditor General's audited accounting cost, $136/night = $49,640/year, the figure this backgrounder and homelessness-economic-roi both treat as Tier 1/locked; (2) a separate, non-audited arithmetic estimate in this library's prior derived-findings document (Follow The Money) ("THE THREE FACTS THAT MATTER," Fact 1) of approximately $271/night ($68,000–$99,000/year in that same source's comparison table) as "Fred Victor's implied cost per night, when you include their overhead and capital servicing" — a single-operator, derived figure built by combining Fred Victor's T3010 overhead/revenue data with the City's per-diem rate, not itself an audited or primary-sourced figure; and (3) a separately-sourced $253/night figure for temporary shelter hotels specifically (distinct from base shelters), per this library's prior synthesis document (Full Cost Accounting) Part 13's own correction of an earlier, unconfirmed $352/night hotel figure. These three numbers answer three different questions (audited system-wide average; one operator's derived full-cost estimate; a hotel-specific per-diem) and this document discloses all three explicitly rather than picking one to present as "the" shelter cost — a hostile reviewer comparing any two of these without this context would find an apparent, unexplained contradiction that is actually a difference in what is being measured, not an error in either figure.

What the evidence does and doesn't support

Well-supported:

• The scale and category structure of Toronto's private-security spending tied to shelters/encampments, independently confirmed via direct City-supplied data reported in December 2025 [NEW-2026-1].
• The CIHI national hospitalization-cost differential for patients experiencing homelessness, directly sourced from CIHI's own April 2024 release [NEW-2026-2].
• The FAO's MCCSS structural-underfunding figures, independently re-confirmed via direct FAO-sourced reporting [NEW-2026-2].
• Canada's official Value of a Statistical Life figure and its 2026-dollar equivalent, independently re-confirmed via direct search of Treasury Board guidance and consistent secondary cross-validation [NEW-2026-1].

Thin or contested:

• Whether Ontario's provincial correctional per-diem cost is genuinely $326/day, as opposed to that figure being the Canada-wide average misattributed to Ontario specifically — Ontario's own official dataset is not currently public, and this review could not resolve the question either way.
• Whether the CIHI national hospitalization-cost differential holds at the same magnitude specifically for Ontario or Toronto — plausible, consistent with the broader literature, but not independently confirmed at that jurisdictional level in any source found this review.
• The cumulative multi-year mortality-monetization figure (applying a single present-day VSL retroactively across many years) remains, as the carried-forward sources itself states, a simplified illustrative calculation rather than a rigorous year-appropriate-discounted total.

International context

1. Treaties/frameworks touched. This page’s cross-departmental cost-accounting scope does not itself engage a specific international human-rights treaty article — the relevant instrument (ICESCR Article 11, the right to adequate housing) is more directly engaged by the homelessness-legal-rights-frameworks leaf, and this document does not manufacture a treaty connection that its own cost-accounting scope does not require.

2. 2-3 best global comparators. No global comparator for a comprehensive, multi-ministry, monetized full-cost-accounting exercise of this specific kind (cross-departmental, VSL-inclusive) was identified via live discovery in this review — the carried-forward sources's own comparator material is concentrated in the political-economy page’s case studies (Houston, Calgary, Finland), which address system-design comparators rather than cost-accounting methodology specifically, and this document does not duplicate that page’s international content. Flagged honestly as a gap rather than manufacturing a comparator that does not fit this page’s specific accounting-methodology scope.

3. What Toronto/Ontario can steal shamelessly. Not applicable in the same descriptive-comparator sense as other pages, given the gap noted above — the closest transferable mechanism this document can point to is methodological rather than a named foreign program: Canada's own Treasury Board VSL methodology is itself the "mechanism worth stealing" for this page’s own purposes, already domestic, and the carried-forward sources's application of a government's own required regulatory tool to make homelessness mortality visible in budget-comparable terms is the transferable design element — using a government's own already-mandated cost-benefit standard against its own budget silence on this question, rather than importing a foreign model.

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. This section is required in every backgrounder per this library's binding 2026-07-14 standing rule; a leaf with a genuinely different scope than "who profits" may still produce a thin or empty table, and this template's own guardrail states explicitly that an empty table with an honest explanation is the correct output, not a failure to fill the template.

Table

(No rows this review.)

Why this table is empty this review, stated explicitly

This page’s own scope — cross-departmental cost accounting (how much homelessness costs across health, justice, corrections, municipal, and federal systems) — is structurally different from a "who profits from the current system" inquiry, which is this project's homelessness-political-economy page’s own binding scope. The two private-security findings surfaced in this review's live discovery (Garda Canada Security Corporation's $11.9M contract, the broader $109M portfolio) are exactly the kind of Cui Bono-relevant finding this template's Prime Rule discipline requires — and this backgrounder does not duplicate them here, because homelessness-political-economy's own backgrounder (dated the same day as this document) already carries the full, properly-graded Cui Bono table for those same findings (entity_id ENT-0009), per this repo's own "claim-anchored, not backgrounder-anchored" discipline against two documents independently re-deriving the same synthesis. A reader seeking this page’s Cui Bono content should consult our research file for that page's Cui Bono section directly, cited here by reference rather than duplicated.

Open questions / data gaps

• Not yet drawn into the claims register: the 2 items tagged NEW-2026-# in this document are drawn from primary sources fetched and quoted in this review (2026-07-14) but have not been run through this project's formal add_claim.py/registry pipeline. Treat as ⚠️ still-being-checked until a future verification pass formally promotes them.
• Genuinely uncovered: an Ontario- or Toronto-specific (as opposed to Canada-wide) hospitalization-cost differential for patients experiencing homelessness — CIHI's own national figure is the best available, and no source located this review isolates the Ontario/Toronto figure.
• Genuinely uncovered: whether Ontario's own provincial correctional per-diem cost is genuinely distinct from, or identical to, the Canada-wide average — Ontario's own official dataset is currently unpublished per its own open-data portal, a data-access constraint independently confirmed (not merely inherited) this review.
• Genuinely uncovered: education-system and child-welfare-system dollar costs specific to Toronto/Ontario homelessness — both remain flagged as real but uncosted gaps by the carried-forward sources, and this review did not locate a source closing either gap.
• Found but not yet formally registered: none identified this review beyond the NEW-2026-# items already flagged above.
• Scoped out by design: the specific Housing First savings/ROI case (steady-state avoided-cost figures, sensitivity ranges, phased build-out projection) belongs to homelessness-economic-roi, cited rather than re-derived here; the Cui Bono findings surfaced in this review's own live discovery are documented in full in homelessness-political-economy's backgrounder rather than duplicated in this one; individual shelter-operator financial profiles (T3010 overhead/reserves/Sunshine List compensation, including the DERIV-011 four-operator $154.5M/29%-overhead/$86.8M-reserves table and Covenant House's $48.5M reserve) belong to shelter-operator-financial-profiles, which carries that exact material in full, cited rather than duplicated here; the 32-of-50-recommendation costed ROI matrix (this library's prior synthesis Part 2, Tiers 1-4) is recommendation/campaign-cost framing quarantined project-wide per this project's recommendations-quarantine rule, not this page’s scope to restate.
• 2026-07-16 a later verification pass: restored into "Current state" — the TSSS visible-system funder breakdown ($796.4M by federal/provincial/city share), the derived ~$161M/year excess-provincial-healthcare figure, the 20,000-30,000-person hidden-population externalized-cost finding, COVID-era permanent cost structures (Temporary Hotel Program, property-tax-base absorption), the tax-revenue honest-treatment finding (Poremski et al. 2016 employment/income null result, the IPS conditional exception), workforce-transition findings (1,497.5 TSSS positions), and the carried-forward sources's own extrapolation-discipline standard (Part 8). All carry the master briefing's own citations, marked carried-forward.

Claim-index appendix

carried-forward (carried forward from this page’s own sources document, costed/sourced content only):

• this library's prior synthesis document (Funding Stakeholder ROI Matrix) Part 1.1 · carried-forward · TSSS 2024 visible-system funder breakdown ($796.40M: federal 33%/provincial 39%/City 24%/other 3%)
• this library's prior synthesis document (Funding Stakeholder ROI Matrix) Part 1.2 · carried-forward · derived ~$161M/yr excess provincial healthcare cost (15,418 × $10,440), restored 2026-07-16
• this library's prior synthesis document (Funding Stakeholder ROI Matrix) Part 1.4 · carried-forward · 20,000-30,000-person hidden/externalized-cost population, restored 2026-07-16
• this library's prior synthesis document (Full Cost Accounting) Part 1.2 · carried-forward · 2021 three-park clearance cost (~$2M), standing security-contract figure ($2.94M/year, pre-this-pass anonymized)
• this library's prior synthesis document (Full Cost Accounting) Part 1.3 · carried-forward, ⚠️ still being checked · municipal legal liability (Lamport Stadium suit, $50M refugee-exclusion class action, Black/Sanctuary cases)
• this library's prior synthesis document (Full Cost Accounting) Part 1.4 · carried-forward · TFS Encampment Strategy Team, cost not isolated
• this library's prior synthesis document (Full Cost Accounting) Part 2.1-2.4 · carried-forward, ⚠️ still being checked · CIHI/ICES health costs, MCCSS FAO underfunding, LTB adjudicator costs, corrections per-diem
• this library's prior synthesis document (Full Cost Accounting) Part 2.7 · carried-forward, ⚠️ still being checked · Bill C-12 eligibility bars, IRCC procedural fairness letters
• this library's prior synthesis document (Full Cost Accounting) Part 3.1 · carried-forward, ⚠️ still being checked · IHAP, HSCIS capital request, Reaching Home
• this library's prior synthesis document (Full Cost Accounting) Part 4 · carried-forward, corroborated NEW-2026-1 · VSL methodology and mortality monetization
• this library's prior synthesis document (Full Cost Accounting) Part 8 · carried-forward · extrapolation-discipline standard (legitimate vs. compounded-guess extrapolation), restored 2026-07-16
• this library's prior synthesis document (Full Cost Accounting) Part 10 · carried-forward, ⚠️ still being checked · tax-revenue honest treatment (Poremski et al. 2016 null employment/income result; IPS conditional exception; peer-navigator income-tax calculation), restored 2026-07-16
• this library's prior synthesis document (Full Cost Accounting) Part 11 · carried-forward, ⚠️ still being checked · workforce transition (1,497.5 TSSS positions), restored 2026-07-16
• this library's prior synthesis document (Full Cost Accounting) Part 13 · carried-forward, ⚠️ still being checked (disposables claim unconfirmed) · COVID-era permanent cost structures (Temporary Hotel Program $278.052M, property-tax-base absorption), restored 2026-07-16

New load-bearing findings (this review, source quotes below, not yet through this library’s formal verification process):

---

Source quotes (NEW-2026-1, NEW-2026-2)

NEW-2026-1 — Value of a Statistical Life (Treasury Board) and private-security contract portfolio.

"The value of statistical life (VSL) is the additional cost that individuals would be willing to pay for small reductions in risks that, in the aggregate, reduce the expected number of fatalities by one. Departments must use the VSL specified in TBS's Cost-Benefit Analysis Guide... the updated guideline for cost-benefit analysis suggests that government departments use the average value of mortality risk reduction (VRMR) of $6.5 million (in 2007 dollars) which is $9.1 million in 2020 Canadian dollars after adjustment for inflation... equivalent to $9.8 million in 2026 Canadian dollars."

Source: Treasury Board of Canada Secretariat, Policy on Cost-Benefit Analysis, https://www.canada.ca/en/treasury-board-secretariat/services/federal-regulatory-management/guidelines-tools/policy-cost-benefit-analysis.html; Canadian Cost-Benefit Analysis Guide for Regulatory Proposals, https://www.tbs-sct.canada.ca/rtrap-parfa/analys/analys-eng.pdf. Accessed 2026-07-14.

"The City of Toronto is facing questions about the cost of guards it deploys at locations like homeless shelters and encampments, with various private security companies signed up to contracts worth a total of roughly $109 million... Three contracts for security at shelter respite centres are worth $44.8 million in total, while another contract for shelter core sites runs to $15 million. There's also an agreement with Garda Canada Security Corporation to provide 'encampment support' at $11.9 million... So far this year, across all of those agreements, city hall has shelled out $35.2 million."

Source: Global News (Isaac Callan), "'No plans to change': Toronto spent $35M on private security contracts this year," December 18, 2025, https://globalnews.ca/news/11582413/toronto-private-security-guard-costs/. Accessed 2026-07-14.

NEW-2026-2 — CIHI hospitalization-cost differential and FAO MCCSS underfunding.

"Nearly 30,000 hospitalizations in Canada included code Z59.0 Homelessness in 2022–2023. The average cost of a hospitalization for a patient experiencing homelessness was more than double the cost of an average hospital stay in Canada ($16,785 versus $7,803). The most common reasons for hospitalization for PEH were substance use disorders (18%), schizophrenic disorders (11%) and cellulitis (7%). Admission via the emergency department was very common (93%) for PEH."

Source: Canadian Institute for Health Information, "Homelessness and hospital use," released April 25, 2024, https://www.cihi.ca/en/homelessness-and-hospital-use. Accessed 2026-07-14.

"[The FAO] projects that MCCSS spending will decrease at an average annual rate of -0.5 per cent, from $20.7 billion in 2024-25 to $20.4 billion in 2027-28... Compared to the FAO's cost driver projection needed to maintain current service levels, this results in a funding shortfall of $1.5 billion in 2025-26, $2.4 billion in 2026-27 and $3.1 billion in 2027-28."

Source: Financial Accountability Office of Ontario, Ministry of Children, Community and Social Services: 2025 Spending Plan Review, https://fao-on.org/en/report/estimates-2025-mccss/; corroborated by CBC News reporting, "Ontario underspending on social services by $3.7B, financial watchdog says," https://www.cbc.ca/news/canada/toronto/ontario-fao-mccss-report-1.7225423. Accessed 2026-07-14.

Ontario's own correctional per-diem open-data portal: "This data is not available right now. We are reviewing the data to determine if it can be made open to the public." Canada-wide (not Ontario-specific) comparator: "daily inmate expenditures for adult federal, provincial, and territorial correctional services in Canada remained at around 326 current dollars in 2022" (Statistics Canada); federal (not provincial) 2025 comparator: "$436 a day" (Correctional Service Canada, per Blacklock's Reporter, July 2025). [independently re-verified 2026-07-16: corrected from a paraphrased composite to the portal's exact two-sentence text; substance unchanged.]

Source: Government of Ontario open-data portal (dataset listing, per-diem data unavailable), https://data.ontario.ca/dataset/average-cost-per-day-to-house-an-inmate-in-a-provincial-institution-per-diem; Statistics Canada, Operating expenditures for adult correctional services, https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3510001301; Blacklock's Reporter, "Prison Cost Now $436 Daily," https://www.blacklocks.ca/prison-cost-now-436-daily/. Accessed 2026-07-14.