uniteTOlove.ca

Full Cost of Homelessness — Playbook

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

DRAFTThe playbookThe evidence file

What Toronto can actually do on the full fiscal cost of homelessness — each move with its costs, its beneficiaries, and its receipts.

v2.0 · 2026-08-11

---

The honest bottom line

Every shelter-budget debate in Toronto happens as if the shelter budget were the whole cost of homelessness. It isn't. Toronto currently carries $109 million in private-security contracts running through 2026 and 2027, concentrated almost entirely on shelters and encampments — $71.7 million of it specifically tied to shelter respite centres, shelter core sites, and one named encampment-support contractor, with $35.2 million already spent this year: a standing, multi-year commitment, not a one-time crisis cost. Canada's own federal Value of a Statistical Life methodology — the same tool used to price a new safety regulation — puts a dollar figure of approximately $9.8 million on a human life; applied to the 59 shelter-resident deaths Toronto's own system recorded in a single year, that is $578 million, the government's own number applied to the government's own count. Canada's national health data agency separately found a hospital stay for someone experiencing homelessness costs more than double the average — $16,785 versus $7,803 — driven by 93% of these patients arriving through the emergency room rather than a family doctor's office. But two of the figures this kind of case usually leans on hardest turned out to be shakier than expected once checked directly: Ontario's own correctional per-diem cost data is currently withheld by the province's own open-data portal ("not available right now"), so the widely-used $326/day figure is actually a Canada-wide Statistics Canada average, not an Ontario number; and the $16,785 hospitalization figure is a national figure with no published Ontario- or Toronto-specific version. Neither gap weakens the case — the confirmed numbers alone ($109 million security portfolio, $578 million mortality cost, a $3.1 billion province-wide social-services funding gap the Financial Accountability Office itself projects by 2027-28) are already large enough — but padding them with unconfirmed Ontario-specific figures wearing a national label gives a skeptical reader a free win.

Get a real, government-published, jurisdictionally-precise cost picture. (a recommendation card) Homelessness generates documented costs across at least five government levels and ministries — municipal shelter, legal, and security costs; provincial health, corrections, and social-services costs; federal program costs; and mortality costed via Canada's own official VSL methodology — but no single published document brings these together using government-owned methodologies. This card closes that gap using tools that already exist and already belong to government bodies (Treasury Board's VSL methodology, CIHI's hospitalization-cost coding, the FAO's cost-driver projections, Statistics Canada's correctional-expenditure series), explicitly resolving the Ontario-specific-vs-Canada-wide ambiguity this research ran into directly rather than leaving it unresolved.

Do the mortality math properly. (a recommendation card) The $578 million mortality figure applies a single present-day VSL rate retroactively across many years — an honest, disclosed simplification, not a hidden one, but it means the strongest number in this whole leaf still carries an asterisk. This card replaces the illustrative single-rate calculation with a year-appropriate, discounted one, using Treasury Board's own historical VSL guidance, so the number stops needing a caveat.

Neither move produces a single "true cost of homelessness" grand total — that would manufacture false precision the source research itself declines to invent — and neither sums in the real but undocumented costs (a live $50 million class action over a 2021 encampment clearance, completely unquantified education and child-welfare costs tied to family homelessness) that remain outside what any current source can support. Both moves require finishing math the government's own tools already make possible, not inventing a new argument.

---

a recommendation card — Publish a Consolidated, Government-Standard-Methodology Full-Cost Statement

Card id: a recommendation card · Issue: homelessness-full-cost-accounting · Backgrounder: our research file for that page · Trust: earlier research plus new load-bearing findings (backgrounder Part 4, NEW-2026-1)

Problem

Homelessness generates real, documented costs across at least five distinct government levels/ministries (municipal shelter and legal/security costs, provincial health/corrections/social-services costs, federal program costs, and mortality costed via Canada's own official Value of a Statistical Life methodology), but no single published document brings these together using government-owned methodologies and figures. The backgrounder itself documents this fragmentation directly — several figures (Ontario's correctional per-diem cost, an Ontario-specific hospitalization-cost differential) could not be confirmed as genuinely Ontario-specific rather than Canada-wide averages, precisely because no consolidated, jurisdictionally-precise government accounting exists.

Action

The Province of Ontario or the City of Toronto (jointly, given the cross-jurisdictional nature of the costs) commissions and publishes a consolidated full-cost statement using exclusively government-owned methodologies and datasets already in use elsewhere (Treasury Board's VSL methodology, CIHI's hospitalization-cost coding, FAO's cost-driver projection methodology, Statistics Canada's correctional-expenditure series), explicitly resolving the jurisdictional-precision gaps this backgrounder documents (Ontario-specific vs. Canada-wide figures) rather than leaving them ambiguous.

Jurisdiction split

Cost

Order-of-magnitude: low — a consolidation and methodology-alignment exercise using data multiple bodies (CIHI, FAO, Statistics Canada, the City) already collect. Comparator: the carried-forward sources document itself, produced without dedicated government funding, as evidence the analytical work is achievable at modest cost. The harder, costlier piece is the currently-missing Ontario-specific correctional per-diem publication, whose cost this card cannot estimate without knowing why that dataset is currently withheld.

Funding path

Existing FAO, CIHI, and municipal budget-office analytical capacity; no new funding mechanism required for the consolidation itself. Publishing the currently-withheld Ontario correctional per-diem dataset is a data-release decision, not a spending decision, and this card does not assume it requires new funding.

Who benefits, and how

Policy-makers and budget offices at all three government levels, via a single, government-standard-methodology reference point for a cost picture currently scattered across at least six different bodies' separate publications; researchers and oversight bodies, via a jurisdictionally-precise baseline that would prevent the kind of national-vs-Ontario-specific ambiguity this backgrounder had to flag explicitly in two separate places.

Who bears the cost, and how

Taxpayers, via existing analytical-body operating budgets; no new payer class identified for the consolidation itself.

Who benefits from the status quo

No beneficiary identified in the backing backgrounder's Cui Bono table — the backgrounder's own Cui Bono section explains why it is empty (this page’s scope is cost-accounting, not a profit inquiry) and points to homelessness-political-economy's backgrounder for the Cui Bono findings that emerged from adjacent research. Stated explicitly per this template's guidance rather than manufacturing a beneficiary the backgrounder didn't establish.

Financial ROI

Not directly estimated — this is an analytical/publication exercise, not a spending program; its value is in enabling better-informed future spending decisions, not in producing a direct saving of its own. No comparator identified beyond the carried-forward sources document itself (produced without dedicated funding), which anchors the scale of effort involved. Confidence: low — the action's value is instrumental, not directly financial.

Economic ROI

Not yet estimable and not directly modeled — a consolidated cost statement has no direct construction, employment, or local-spending effect of its own, and no comparator source was identified. Confidence: low — genuinely not yet estimable.

Social ROI

Directional: a jurisdictionally-precise, government-standard-methodology cost picture plausibly improves the quality of future policy debate and resource-allocation decisions across all three government levels, though no source quantifies this effect for a comparable consolidation exercise elsewhere, and no comparator specific to this mechanism was identified. Confidence: low — directional only.

Environmental ROI

Genuinely environmentally neutral — a consolidated cost statement has no construction, land-use, or physical-infrastructure component; no comparator is needed to support a neutral finding. Confidence: high on the neutrality of this specific action's own footprint.

Evidence

Confidence & uncertainties

Medium confidence on the problem statement (the fragmentation and jurisdictional-precision gaps are directly documented in this review's own research); low confidence on cost/timeline for the Province releasing the currently-withheld correctional dataset, since no source explains why it is currently unavailable.

Status

DRAFT — blocked on: fairness and legal review; confirmation of why Ontario's correctional per-diem dataset is currently withheld (not established in this review).

---

a recommendation card — Extend Mortality Monetization to a Full, Year-Appropriate-Discounted Model

Card id: a recommendation card · Issue: homelessness-full-cost-accounting · Backgrounder: our research file for that page · Trust: carried-forward (backgrounder Part 4, its own stated caveat)

Problem

The backgrounder's mortality-monetization figures (the cumulative multi-year total in particular) explicitly carry the carried-forward sources's own caveat: they apply a single present-day Value of a Statistical Life rate retroactively across many years, a simplification the source itself states should be read as an illustrative order-of-magnitude rather than a precise total. This is an honest limitation, not a hidden one — but it means the strongest, most defensible version of this figure (a properly year-appropriate-VSL, discounted calculation) does not yet exist.

Action

A future research pass builds the more rigorous version of the mortality-monetization calculation the backgrounder's own caveat already calls for: applying the historically-correct VSL figure for each year in the cumulative count (rather than a single present-day rate), with appropriate discounting for a multi-year cumulative total, using Treasury Board's own historical VSL guidance figures where available.

Jurisdiction split

Cost

Order-of-magnitude: low, an analytical refinement using an already-established methodology (Treasury Board's own VSL guidance across its historical revisions) — no comparator needed beyond the existing calculation's own scale.

Funding path

Existing research capacity within this project or a partner analytical body; no new funding mechanism required.

Who benefits, and how

Anyone citing this project's mortality-monetization figures in a policy or advocacy context, via a more defensible, harder-to-dispute calculation than the current explicitly-illustrative one; this project's own credibility, via closing a limitation it has already flagged rather than leaving it open indefinitely.

Who bears the cost, and how

This project's own analytical capacity (or a research partner's), via the time cost of the more rigorous calculation; no external payer identified.

Who benefits from the status quo

No beneficiary identified in the backing backgrounder's Cui Bono table — this card addresses a methodological refinement, not a contested resource allocation with an identifiable status-quo beneficiary.

Financial ROI

Not applicable in the direct sense — this is a methodological refinement to an analytical figure, not a spending or savings action. No comparator applies. Confidence: not applicable.

Economic ROI

Not applicable — no direct economic effect, and no comparator applies. Confidence: not applicable.

Social ROI

Directional: a more rigorous, harder-to-dispute mortality-monetization figure strengthens the evidentiary basis for any future policy argument built on it, without itself changing any social outcome directly; no comparator identified. Confidence: low — directional only, and modest in scale.

Environmental ROI

Genuinely environmentally neutral; no comparator needed. Confidence: high on the neutrality of this specific action's own footprint.

Evidence

Confidence & uncertainties

High confidence this is a genuine, already-flagged limitation (the backgrounder inherits the carried-forward sources's own explicit caveat rather than discovering a new one); low confidence on how much the refined figure would differ from the current illustrative one, since that depends on historical VSL figures this review did not attempt to collect.

Status

DRAFT — blocked on: collection of Treasury Board's historical VSL guidance figures across the relevant year range (not attempted in this review); fairness and legal review.

---

Production record

Drafting record

Version: v1.0 (cards content, tightened into v2.0 playbook shape 2026-08-11) · Original date: 2026-07-14 · Status: DRAFT · · Backgrounder: our research file for that page. Written per this library's standard page structure — every factual premise traces to a claim in the backgrounder; no figure here is invented. Both cards draw on the same backgrounder; provenance class and confidence noted per card. Author voice: The Unknown Soldier.

v2.0 restructure (2026-08-11, a recorded standing decision/PLAYBOOK conversion, Lane L3b): opened with "The honest bottom line," adapted from archive/dayone/homelessness-full-cost-accounting.md (retired day-one memo, a recorded standing decision); each card tightened, verbose ROI blocks collapsed to flowing prose; all citation tokens preserved verbatim.

About this file