uniteTOlove.ca

Housing First — Playbook

Give people housing before demanding they fix other problems first — what the strongest Canadian trial actually found.

DRAFTThe playbookThe evidence file

What Toronto can actually do on the Housing First model for homelessness — each move with its costs, its beneficiaries, and its receipts.

v2.0 · 2026-08-11

---

The honest bottom line

Toronto ran part of the world's largest randomized controlled trial of Housing First — give people permanent housing immediately and unconditionally, then wrap voluntary support around them, instead of making housing a reward for sobriety or treatment progress first. Among the trial's highest-needs participants, 73% ended up stably housed versus 31% under the old system — more than double, and this is trial evidence, not a hopeful pilot (the 73%/31% figures carry their own confirm-against-primary-study flag from the source material). Toronto's own Auditor General separately found that emergency shelters cost 3 to 10 times more than supportive or subsidized housing, and the City has adopted Housing First as its stated philosophy — but not as the system's funded default: 78% of shelter users meet the City's own chronic-homelessness threshold, yet the system moved fewer than 4,000 people a year into housing against a nightly count above 12,000. The same approach has worked at city and national scale elsewhere: Houston cut its homeless population 63% since 2011 without new municipal spending, Medicine Hat became the first Canadian city to reach "functional zero" chronic homelessness in June 2021 using a real-time by-name list and weekly cross-provider case conferencing, and Finland cut shelter use 76% between 2008 and 2017 after taking the model national — and Toronto already belongs to the same national network (Built for Zero Canada) the Canadian results come from, with the assessment tool that network's coordinated-access work is built on already in place.

Unlock TCHC's own empty units first. (a recommendation card) Toronto's Auditor General already found, in a 2020 audit ("Opening Doors to Stable Housing"), that a 50% improvement in Toronto Community Housing Corporation vacancy management would house about 2,200 more people and recover roughly $7 million a year now being wasted — with no new construction, no new program, and no other government's permission required. This is the cheapest, least contestable first move in the entire file.

Fund Housing First as the default, not the exception. (a recommendation card) Shift Toronto Shelter and Support Services spending toward rent supplements and supportive housing, prioritizing expiring hotel-lease conversion and modular construction over costlier shelter-bed builds where the two can substitute. Toronto's own capital strategy for new shelters (HSCIS) costs $421,875 per partitioned shelter sleeping space versus $309,000 per private supportive unit with kitchen and bathroom under the City's Modular Housing Initiative — 36% more per space for a worse, partitioned-not-private result, a real City-sourced number. This depends on the Province not repeating its 2024 suspension of the Canada-Ontario Housing Benefit, which coincided with a 26.7% drop in housing exits that year (5,927 to 4,344).

Build the by-name list at Toronto's scale. (a recommendation card) Medicine Hat and Houston both show that knowing every homeless individual by name, tracked in real time, and reviewing every case weekly across every provider, is itself part of what makes housing placement work — not just having units available. Toronto already has the pieces (Built for Zero Canada membership, the STARS assessment tool) but has not run this at the scale a city its size requires; the action starts with the highest-cost, highest-need individuals, not the whole system at once.

Both Finland and Medicine Hat later slipped: Finland's homelessness rose for two straight years — up 20% in 2025 alone, the largest single-year increase since it started tracking the number — after cuts to housing-advice budgets and income supports, even as the underlying model kept running; Medicine Hat lost its functional-zero status within a year, as rising rents pushed already-housed people back into homelessness faster than the system could re-house them. Neither case reads as Housing First failing — both are read, by the sources themselves, as proof the model needs continuous housing and income investment to hold, not a one-time fix. Toronto's own October 2025 count showed a 21% drop, real but driven almost entirely by fewer refugee claimants in the shelter system, not by improvement in chronic, long-term homelessness, which has not moved on the same measure; the City's own November 2025 release says plainly that even this year's gains are "at risk without stable funding agreements between all three orders of government." The exact size and cost of the supportive-housing pipeline Toronto needs is not answered anywhere in this file, and none of the three moves above is fully costed or funding-secured — each names what is and isn't confirmed rather than presenting an aspirational path as settled.

---

a recommendation card — Unlock TCHC Vacancies as the First, Zero-New-Money Housing First Ask

Card id: a recommendation card · Issue: homelessness-housing-first-model · Backgrounder: our research file for that page · Trust: carried-forward

Problem

Toronto's own Auditor General found in a 2020 audit ("Opening Doors to Stable Housing") that a 50% improvement in Toronto Community Housing Corporation (TCHC) vacancy management would house approximately 2,200 more people and recover approximately $7 million per year in currently wasted vacancy losses, without building a single new unit [this library's homelessness master synthesis, REC 0]. This card addresses only that specific vacancy-turnover gap, not TCHC's broader capital or governance issues.

Action

Embed dedicated Housing Placement Workers, drawn from the By-Name List priority queue, inside TCHC to actively match arriving vacancies to chronically homeless individuals; fast-track eligibility review; require monthly public vacancy reporting to Council's Economic and Community Development Committee.

Jurisdiction split

Cost

Order-of-magnitude: low, in the range of the cost of a small dedicated staffing team (illustratively, single-digit millions annually for placement workers and fast-tracked eligibility review), offset by the AG's own identified ~$7M/year in currently wasted vacancy-loss recovery [this library's homelessness master synthesis]. Named comparator: the AG's 2020 audit finding is itself the anchor for both the housing-outcome scale (2,200 people) and the savings figure — the source's own reported figures, not an invented estimate.

Funding path

Existing TCHC operating budget reallocation plus the recovered ~$7M/year in vacancy losses identified by the AG; no new levy, grant, or external program required. This is the most fiscally conservative, non-ideological ask in this page’s evidence base for exactly that reason [this library's homelessness master synthesis, REC 0].

Who benefits, and how

Approximately 2,200 chronically homeless individuals on the By-Name List priority queue, via faster matching to existing TCHC vacancies that currently sit empty during turnover — a mechanism, not merely a target [this library's homelessness master synthesis].

Who bears the cost, and how

TCHC's existing operating budget absorbs the placement-worker cost; City taxpayers bear no net new cost given the AG's own ~$7M/year recovery estimate offsets it. Opportunity cost: TCHC administrative capacity redirected toward this priority-matching function.

Who benefits from the status quo

No beneficiary identified in the backing backgrounder's Cui Bono table specific to TCHC vacancy management itself — the backgrounder points to homelessness-political-economy.md's two rows (GardaWorld, unnamed hotel operators), both of which concern emergency-shelter/security spending rather than TCHC's own vacancy-turnover process.

Financial ROI

The AG's own audit is the source for both the ~$7M/year currently-wasted recovery and the 2,200-person housing outcome — closer to a sourced figure than a rough estimate, though it is a 2020 audit figure not independently re-verified against 2026 TCHC vacancy data this review [this library's homelessness master synthesis]. “still being checked”: 2020 figures not re-checked against current TCHC vacancy rates.

Economic ROI

Not yet estimable at a modelled figure: a staffing/placement-process change with no new construction has no local-growth or induced-spending effect of its own beyond the placement workers' own salaries (a small, city-internal reallocation, not a new spending injection); any broader economic effect would run through the 2,200 formerly-homeless individuals' own subsequent income/employment outcomes, not modelled in this page’s sources. No vacancy-management-specific economic-impact study was identified; the closest available literature is general Housing First employment-outcome research (Federal Reserve Bank of Kansas City's summary of Elior Cohen's Los Angeles County study finding Housing First assistance increases income and employment, https://www.kansascityfed.org/ten/2022-summer-ten-magazine/ask-an-economist-homelessness-and-the-effect-of-housing-first-programs/), named here as adjacent context rather than a direct comparator since it models a different mechanism in a different jurisdiction. Confidence: low — no on-point comparator found.

Social ROI

Directional and strongly supported by the backgrounder's broader Housing First evidence: stable permanent housing for a high-needs population is the same outcome At Home/Chez Soi demonstrates at RCT strength (73% vs. 31% housing stability for high-needs participants) [backgrounder, "The decisive Canadian trial"], applied here to a specific, already-existing Toronto housing stock rather than new construction.

Environmental ROI

Genuinely environmentally neutral: this action reallocates existing, already-built TCHC units faster, building nothing new and changing no land use, so it carries no incremental emissions, land-use, water, or waste footprint of its own (distinct from a recommendation card below, which does involve new construction and is scored accordingly). Confidence: high — the mechanism is explicitly a turnover/placement-speed change to existing stock, not a build.

Evidence

Confidence & uncertainties

High confidence on the mechanism (existing audit finding, existing City-owned asset, no new program design required). Medium confidence on the exact 2,200/$7M figures holding six years after the underlying 2020 audit — this card takes the position that the mechanism is sound even if the exact figures need refreshing, and says so rather than presenting 2020 numbers as current.

Status

DRAFT — blocked on: a refreshed TCHC vacancy-rate figure (2020 audit is six years old), fairness and legal review.

---

a recommendation card — Fund Housing First and the Supportive-Housing Pipeline as the Default, at Scale

Card id: a recommendation card · Issue: homelessness-housing-first-model · Backgrounder: our research file for that page · Trust: earlier research plus this review

Problem

Toronto has adopted Housing First as stated philosophy but not as its funded operating model: 78% of shelter users meet TSSS's own chronic-homelessness threshold, and the system produced 3,808 housing exits in 11 months of 2024 against a nightly census of 12,000+ [backgrounder, "Toronto's own adoption record"]. The binding constraint, per this page’s own sources, is that Housing First "cannot" function without an actual supply of supportive and non-market housing units to place people into [backgrounder, "Key tensions / tradeoffs"].

Action

Redirect a minimum share of the Toronto Shelter and Support Services (TSSS) budget from indefinite shelter/hotel operation toward Housing First rent supplements and supportive-housing acquisition, prioritizing conversion of expiring hotel leases and modular/rapid-build construction over the costlier Homelessness Services Capital Infrastructure Strategy (HSCIS) shelter-bed model where the two are substitutable [this library's homelessness master synthesis, REC 4; ].

Jurisdiction split

Cost

Order-of-magnitude: the City's own HSCIS program is a directly named comparator at $421,875 per partitioned shelter sleeping space versus $309,000 per private supportive unit with kitchen and bathroom under the City's Modular Housing Initiative — a real, City-sourced per-unit cost delta, not an invented range [this library's homelessness master synthesis]. A second named comparator: HSCIS's own stated savings claim is up to $127/bed-night versus current shelter-hotel costs, City-projected at $74 million in operating savings over 20 years across its 20-shelter, ~1,600-bed program [City of Toronto, 2024-12-11 release]. Precision is limited to what these two named City programs report; this card does not manufacture a single Toronto-wide supportive-housing-pipeline cost figure beyond these named comparators, consistent with the backgrounder's own flagged open question on total supply-gap cost.

Funding path

Blended: TSSS operating-budget reallocation (existing $786.068M-order budget per 2026 program documents, referenced in this page’s spine as $897.957M gross for 2025 [this library's homelessness master synthesis] — ⚠️ note added 2026-07-13: the City's own 2026 Budget Notes cite a different, unreconciled $912.032M comparator for the same "2025 Budget" line; see our research file for that page for the reconciliation note); COHB restoration (federal-provincial cost-shared, Province-administered); federal capital transfers via existing housing-investment programs; reallocation of encampment-clearance and warming-centre spending, itself independently documented at $22.2M for 2023/24 warming centres and winter respite alone [this library's homelessness master synthesis]. This is a contested, not-yet-confirmed funding path precisely because the COHB piece depends on provincial cooperation that has already lapsed once, in 2024 [this library's homelessness master synthesis].

Who benefits, and how

The approximately 78% of Toronto shelter users meeting the chronic-homelessness threshold [backgrounder, "Toronto's own adoption record"], via a shift from indefinite shelter stays to permanent supportive housing with voluntary wraparound support — the same mechanism At Home/Chez Soi demonstrates at RCT strength.

Who bears the cost, and how

City taxpayers via TSSS budget reallocation (not a new levy, but opportunity cost against other City priorities if reallocation is insufficient and new funds are required); provincial and federal general revenue via the demanded COHB restoration and capital transfers.

Who benefits from the status quo

Two entities named in homelessness-political-economy.md's own Cui Bono table: two hotel operators ("Hotel C"/"Hotel D"), left anonymized in the Toronto Auditor General's 2022 hotel-shelter audit itself (so no entity can be registered by name here), which found these operators overcharged the City for vacant emergency-shelter rooms — this card's action (redirecting TSSS budget away from indefinite shelter/hotel operation toward Housing First) directly reduces the City's reliance on the hotel-based capacity this finding concerns. Garda Canada Security Corporation/GardaWorld (ENT-0009), per ACL-0009, holds an $11.9 million encampment-support contract under the current emergency-response-heavy system this card's reallocation is designed to shrink.

Financial ROI

At Home/Chez Soi's own net-cost figure (~$6,311/person/year after ~69% cost offset) is the strongest available financial-ROI anchor, though it is trial-population evidence (high-needs chronic homelessness), not a Toronto-specific programmatic ROI model [backgrounder, "Cost-effectiveness"]. The HSCIS-vs-modular cost delta above is a second, narrower, but City-specific ROI signal. Confidence: medium — the trial ROI figure is strong evidence the model works; it is not a costed Toronto-scale pipeline budget, which the backgrounder itself flags as a genuine open question.

Economic ROI

Order-of-magnitude only: applying Statistics Canada's general construction GDP multiplier (approximately 1.8x — each $1 of construction output generates roughly $1.80 in total output, from StatCan's "National and Provincial Multipliers" input-output series, https://www150.statcan.gc.ca/n1/en/catalogue/15F0046X; current detail-level tables at https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3610059401) to the modular/supportive-housing construction share of this card's reallocated TSSS spend implies a real, non-trivial local/regional induced-output effect on top of the fiscal-offset case already stated in Financial ROI above — but this is a general construction-sector multiplier applied by analogy (StatCan's dedicated multiplier report series was discontinued and is now maintained only as linked data tables), not a Housing-First-pipeline-specific modelled figure, and no total reallocation dollar amount is specified in this page’s sources to apply it to. Confidence: low — a credible official multiplier exists but neither the construction-specific dollar amount nor a Housing-First-specific economic-impact study was found to anchor a program-level range.

Social ROI

Strongly supported: housing stability (73% vs. 31% in the RCT's high-needs arm), reduced mortality and emergency-service reliance are the same evidence base as the backgrounder's "well-supported" findings [backgrounder, "What the evidence does and doesn't support"]. Also carries the Finland/Medicine Hat caution that gains reverse without sustained investment — this card's own confidence section names that risk rather than presenting scale-up as a one-time fix.

Environmental ROI

Real but modest and design-dependent: modular/rapid-build supportive housing construction (the mechanism this card prioritizes over costlier HSCIS shelter-bed builds) carries a genuine embodied-emissions footprint per unit, on the same order as other Ontario low-rise multi-unit housing, with design choice mattering more than the modular-vs-conventional distinction itself. The closest genuinely on-topic comparator is Rankin, Arceo, Isin & Saxe, "Embodied GHG of missing middle: Residential building form and strategies for more efficient housing," Journal of Industrial Ecology, 2024 (https://doi.org/10.1111/jiec.13461; summary at https://csbe.civmin.utoronto.ca/research/embodied-ghg-ofmissing-middle/), which finds embodied GHG of Ontario low-rise multi-unit housing ranges 5,540–39,600 kgCO2eq/bedroom, with the most-efficient design quartile cutting future Ontario residential embodied emissions by up to 46.7% — an on-point (Ontario, low-rise multi-unit) study, but not specific to modular construction or to this program's actual unit design. Confidence: low-medium.

Evidence

Confidence & uncertainties

Medium-high confidence that reallocation toward Housing First is directionally correct per the evidence; medium confidence on the specific reallocation share and total pipeline cost, both explicitly unresolved in the backgrounder's own open questions. This card takes a position on direction, not on an unsupported total dollar figure.

Status

DRAFT — blocked on: a real Toronto-specific supportive-housing pipeline cost estimate (not yet in the claims register or spine), provincial COHB-restoration commitment, fairness and legal review.

---

a recommendation card — By-Name List / Coordinated Access as the Toronto-Scale Built for Zero Mechanism

Card id: a recommendation card · Issue: homelessness-housing-first-model · Backgrounder: our research file for that page · Trust: carried-forward

Problem

Toronto is a Built for Zero Canada member community, and its STARS Common Assessment Tool is already referenced in BFZ Canada's coordinated-access resources, but Toronto has not achieved functional zero for any homelessness sub-population, while Medicine Hat (population ~63,300) achieved it in 2021 and London, Ontario has achieved it for veteran homelessness specifically [backgrounder, "City-scale precedent: Medicine Hat"]. The mechanism Medicine Hat used — a real-time By-Name List updated at every service contact, plus weekly case conferencing on every individual across all providers — is documented but not established at Toronto's much larger scale [this library's homelessness master synthesis].

Action

Fund and mandate a real-time, cross-provider By-Name List with weekly case conferencing for Toronto's highest-acuity, highest-cost chronically homeless individuals (an initial cohort, not the full 78% chronic population at once), modeled on the Medicine Hat/Built for Zero Canada framework and Houston's Coordinated Entry model prioritizing highest-vulnerability cases first [this library's homelessness master synthesis].

Jurisdiction split

Cost

Order-of-magnitude: low relative to the housing-pipeline card above — this is a coordination/data-system and staffing cost (case conferencing staff time, a shared data system), not a housing-construction cost. Named comparator: Medicine Hat implemented Coordinated Entry in 2010 (first in Alberta) at city scale for a population of ~63,300 [this library's homelessness master synthesis]; Houston's Coordinated Entry/HMIS real-time cross-agency tracking scaled to a much larger metro population without additional municipal spending beyond existing federal allocations [this library's homelessness master synthesis]. Toronto's own master briefing flags a longitudinal client-data system (REC 3) as requiring a privacy impact assessment before implementation — a real, named precondition, not a formality [this library's homelessness master synthesis, REC 3].

Funding path

Existing TSSS operating budget for the coordination function; provincial mental-health/health-system data-sharing agreements for cross-system inputs (hospital, corrections, OW/ODSP). No new levy identified; funding path is largely an administrative/data-governance lift rather than a capital one.

Who benefits, and how

The highest-acuity, highest-cost chronically homeless individuals first (mirroring Houston's vulnerability-prioritized Coordinated Entry and the master briefing's own REC 13 "High-Need Housing Initiative" targeting the 200 highest-cost individuals across shelter/health/justice data) [this library's homelessness master synthesis REC 13], via faster identification and continuous tracking rather than a slower, uncoordinated intake process.

Who bears the cost, and how

City taxpayers via TSSS operating-budget allocation for coordination staffing; no identified new payer class; opportunity cost is administrative capacity redirected from other TSSS functions.

Who benefits from the status quo

No beneficiary identified in the backing backgrounder's Cui Bono table specific to By-Name List coordination itself — the backgrounder points to homelessness-political-economy.md's two rows (GardaWorld, unnamed hotel operators), neither of which bears on a cross-provider data-coordination mechanism specifically.

Financial ROI

Not separately quantified for Toronto in this page’s sources; Houston achieved its results "without additional municipal spending beyond existing federal allocations" [this library's homelessness master synthesis], suggesting the coordination mechanism itself is not necessarily costly relative to its outcomes, but this is not a Toronto-specific costed estimate. Flagged: no direct financial ROI figure for this specific action at Toronto scale; inferred from Houston's own "no new spending" framing, not independently modeled.

Economic ROI

Not yet estimable: a coordination/data-system and case-conferencing-staff cost has no direct local-growth, induced-spending, or employment effect of its own, and Houston's own "no additional municipal spending beyond existing federal allocations" framing [this library's homelessness master synthesis] suggests the mechanism itself is not a meaningful net new spending injection, cutting against a separate Economic ROI case distinct from Social ROI below. No comparator was found: a live search this review for coordinated-entry/By-Name-List economic-impact literature returned only general homelessness-system cost-effectiveness and employment-integration material (e.g. National Alliance to End Homelessness, "The Promise of Coordinated Entry," https://endhomelessness.org/blog/the-promise-of-coordinated-entry/), nothing modelling the coordination mechanism's own local/regional effect distinct from downstream housing outcomes, which this card already covers under Social ROI. Confidence: low — the footprint is plausibly negligible on its own, per Houston's no-new-spending framing, rather than unknown, but no source quantifies that.

Social ROI

Directionally supported by both Medicine Hat's and Houston's outcomes (real housing-stability gains attributable in part to the coordination mechanism, not construction alone) [this library's homelessness master synthesis], though Medicine Hat's own reversal is a caution that coordination without sufficient housing supply behind it is not sufficient on its own [backgrounder, "City-scale precedent: Medicine Hat"].

Environmental ROI

Genuinely environmentally neutral: a data-system and case-conferencing coordination function involves no construction, land-use change, or physical infrastructure; any environmental effect runs entirely through the housing placements it coordinates, scored separately under the housing-supply cards (a recommendation card, a recommendation card) rather than here. Confidence: high — the action is a coordination/data function with no physical footprint of its own.

Evidence

Confidence & uncertainties

Medium-high confidence the mechanism itself is sound (two independent precedents, one Canadian at comparable regulatory context). Lower confidence on cost and timeline at Toronto's much larger population scale relative to Medicine Hat specifically — this is named as an open scaling question, not resolved by analogy alone.

Status

DRAFT — blocked on: privacy impact assessment (per the master briefing's own REC 3 precondition), provincial data-sharing agreement scoping, fairness and legal review.

---

Tensions a serious critic will press

Steelman structure mined 2026-08-12 from the assembly reference stratum (FLAGSHIP FOR/AGAINST layer — argument shapes only; figures re-sourced to this card's own receipts or stated qualitatively; briefings never cited as evidence).

the model.** The model requires actual units of permanent, affordable, supportive housing; with Toronto's supportive-housing shortage and long social-housing waits, Housing First without a production pipeline is a proven engine with no fuel. These cards' asks stand or fall with the supply and non-market agendas, and should be argued as one program.

will test: clearances without genuine, prepared offers of permanent housing displace rather than house — a shelter mat is not housing, and displacement is not housing. Falling visible encampment counts are compatible with people becoming less visible, not housed.

into homelessness (eviction prevention, income supports, discharge planning), but the visible crisis pulls resources toward the already-homeless while the inflow continues. A serious critic will ask what share of the ask addresses inflow.

made homeless by poverty and unaffordable rent needs income and supply, not only supportive housing — different sub-populations, different responses. Overclaiming the model's scope is the fastest way to discredit it.

barrier to being housed, but a real scale-up must be resourced for acute support needs and a strained health workforce — the counterargument is capacity, not efficacy.

Housing First works for its target population; the honest debate is will, funding durability (Finland's reversal when prevention funding was cut), and housing supply.

Production record

Drafting record

Version: v1.1 (cards content, tightened into v2.0 playbook shape 2026-08-11) · Original date: 2026-07-13 · Status: DRAFT · · Backgrounder: our research file for that page. Provenance: cards draw only on the backgrounder's carried-forward and NEW-cited claims; no new factual claims are introduced in card prose beyond what the backgrounder already establishes, per the L6 template's firewall discipline. Voice note: "Toronto already paid for the answers — this work reads it all, integrates it, and shows the work, including the uncertainties." 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-housing-first-model.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