FROZEN EDITION — July 2026. This page is part of the v1 Toronto homelessness research library, a complete, sealed project frozen at its July 2026 figures (92% of its 770 claims individually verified against primary sources). It is a product of its July-2026 campaign context, published as a citable historical artifact. Where our newer research disagrees, a dated margin note points at the living page — the frozen text is never silently rewritten. Why there are two editions.

REGENERATIVE TORONTO — FOLLOW THE MONEY

Toronto Shelter System: Public Contractors, Not Charities

The Campaign for a Regenerative Ontario | Regenerative Toronto | By The Unknown Soldier

Document ID: MAIN-DERIV-011 Version: 1.1 | Source data: CRA T3010 filings via Charity Intelligence Canada; Ontario Sunshine List (Public Sector Salary Disclosure) for named individual compensation Legal note: All data from public CRA filings and public Sunshine List statutory disclosure. No allegations of wrongdoing. All claims are about taxpayer value, not fraud. (Note added AUDIT-002, July 2026: this note is extended to explicitly cover the Sunshine List section below, added after this note was first written — see HMK-005 for the source profile this document draws from.) Use: Deputation material, journalist backgrounder, candidate questionnaire, social media


THE HEADLINE

Toronto's shelter system is not a charity safety net. It is a $154M-per-year contractor network — four non-profit organizations that receive 83-95% of their revenue from government, operate on 29% average overhead (corrected July 3, 2026 — see note in THE NUMBERS below), and collectively hold $86.8M in reserves during a period when the City claims a $165M funding cliff.


THE NUMBERS

Overhead correction, July 3, 2026: Homes First's overhead was previously listed as 14.4% — the administrative-cost component only. Charity Intelligence's own published figure for Homes First's total overhead spending (admin + fundraising, the same combined metric independently confirmed for all three other operators below) is 21%. Corrected below; this also corrects the average, which the 14.4% figure was quietly pulling down.

Operator Gov Funding Gov % Overhead Reserves
Homes First Society $60.8M 94.6% 21% $13.2M
Fred Victor Centre $61.1M 82.8% 31.3% $14.8M
Dixon Hall Services $28.3M 86.0% 31.3% $10.3M
Covenant House Toronto $4.3M 9.4% 33.1% $48.5M
Total (4 operators) $154.5M ~29% avg (corrected — see note above) $86.8M

Source: CRA T3010 annual information returns, Charity Intelligence Canada (charityintelligence.ca) ✅ Fiscal years: Homes First Dec 2023; Fred Victor Mar 2025; Covenant House Jun 2024; Dixon Hall Mar 2024

A real, still-open gap, flagged consistently with HMK-037 (added June 30, 2026): this table has revenue and overhead percentages for these four operators but no corresponding headcount — CRA T3010 filings don't require it. HMK-037's sector-wide employment analysis hit the same wall: a precise, sourced headcount across all ~81 POS shelter operators does not exist anywhere this library has found (see HMK-037 Part 4, and Tier 1 Gap #9). This is the same data-access gap surfacing from two different angles, not two separate problems.


THE SUNSHINE LIST: A DIFFERENT LENS, ADDED JULY 2026

CRA T3010 filings give aggregate overhead percentages, not individual compensation. Ontario's Public Sector Salary Disclosure Act gives the reverse — named individuals earning $100,000+, but no organization-wide headcount. Shelter operators, as recipients of substantial provincial funding, are covered employers under the Act. This section adds that second lens; it does not close the headcount gap above, which remains open.

Fred Victor — the one operator with a complete, verifiable 2025 breakdown: 24 employees appeared on the 2025 Sunshine List, combined salaries totaling $3,114,911. ✅ CEO Keith Hambly: $206,371. Below him, a genuinely wide senior-management layer: VP Philanthropy & Communications ($178,459), VP Programs & Services ($159,246), VP People & Culture ($156,432), Director Finance & Technology ($154,467), VP Finance & Technology ($147,415), and at least eight more director-level roles between $119,000 and $137,000. [Ontario Sunshine List 2025: https://www.ontario.ca/public-sector-salary-disclosure/2025/all-sectors-and-seconded-employees/] This is a genuinely useful data point for the "how many layers of management sit above frontline staff" question this library has asked elsewhere (HMK-015) — at Fred Victor specifically, the answer is: at least five VP-level roles and roughly a dozen director-level roles, all disclosed by name.

TSSS itself, for direct comparison: General Manager Gordon Tanner's disclosed salary was $238,899 (2024). ✅ [Ontario Sunshine List: https://www.ontario.ca/public-sector-salary-disclosure/2025/all-sectors-and-seconded-employees/] The City of Toronto as a whole disclosed 13,035 employees on the 2025 Sunshine List, combined salaries of $1.76 billion — useful only as scale context, since this figure spans every City division, not TSSS specifically.

What this section could not do, stated plainly, and now precisely diagnosed (July 2026, confirmed via direct inspection of Dixon Hall's page). Homes First, Covenant House, and Dixon Hall all appear on the Sunshine List (confirmed), but this library could not retrieve their complete individual dollar figures — and the exact mechanism is now identified rather than merely suspected. A direct fetch of Dixon Hall's employer page returned real, confirmed metadata (67 total records, active years 2008-2011 and 2015-2025 with a gap 2012-2014, categorized as "Other Public Sector Employers") but the "Top Earners" and "Most Common Positions" sections — where individual names and dollar figures would appear — rendered as empty headers. These specific tables are populated by client-side JavaScript after the page loads; the initial HTML this library's tools can retrieve does not contain them. A person viewing this page in an actual browser sees populated tables; automated fetch retrieval sees the pre-JavaScript shell. This is a real, demonstrated tooling boundary, not a research shortfall or a sign the data doesn't exist — it visibly does, one JavaScript execution away from what this library's tools can reach.

RESOLVED, July 2026 — a second tool with different capabilities closed most of this gap. An independent LLM extraction (different tooling than this library's own) successfully retrieved complete, individual-level 2025 breakdowns for Dixon Hall (8 employees, CEO Nermin MinaMawani $306,000) and YWCA Toronto (12 employees, CEO Heather McGregor $222,058) — both independently cross-validated against this library's pre-existing HMK-005 compensation-band estimates and found accurate, not stale. Three entirely new operators were also profiled for the first time: WoodGreen Community Services (27 employees, CEO Anne Babcock $358,492 — the highest executive compensation of any operator checked [independent spot-check July 4, 2026: multiple Sunshine List aggregators consistently confirm $335,576.88 for 2024, following a clear upward trend (2022: $241,603; 2023: $265,842; 2024: $335,577); the specific 2025 figure of $358,492 could not be independently re-confirmed in this pass, though it is plausible given the trend and not contradicted by anything found — flagged for re-confirmation once 2025-specific data is better indexed by third-party trackers]), The Neighbourhood Group Community Services (18 employees, CEO William Sinclair $232,022), and Sistering (3 employees, the smallest operator profiled, ED Volletta Peters $151,125). ✅ See HMK-005 for the complete, updated profiles.

Further expanded, July 2026 — three more operators, plus a ten-year trend layer. LOFT Community Services (Ontario's largest mental health/homelessness provider outside the operators already covered, $71.6M gov funding) was added with a genuinely valuable multi-year Sunshine List trajectory — CEO Heather McDonald's documented salary across three consecutive disclosure years ($185,273→$239,890→$247,087). Good Shepherd Ministries (Toronto) and Eva's Initiatives for Homeless Youth (Toronto's self-described second-largest youth shelter provider after Covenant House) were added with full T3010-derived financial profiles. Three more operators — Na-Me-Res, Sojourn House, Houselink & Mainstay — were confirmed real and scale-documented but no financial profile was located, honestly reported as a gap rather than papered over.

A ten-year historical trend layer, added July 2026, with two specific data-quality issues caught and disclosed rather than smoothed over: cross-referencing a second historical (2016/2021) data delivery against this library's own prior verification found a real ~$56,000 unexplained discrepancy in Dixon Hall's 2025 total between two deliveries from the same tool (flagged, unresolved), and a specific cited 2016 CEO salary figure that could not be independently confirmed (the confirmed figure is for 2017 instead). The genuinely valuable underlying pattern — Fred Victor's six-figure staff count growing from 1 (2016) to 24 (2025), a structural expansion of the management layer, not primarily executive pay inflation — was integrated with full attribution. See HMK-005's dedicated "Ten-Year Compensation and Headcount Trends" section for the complete picture, caveats included.

The Homes First/Covenant House Sunshine List absence is now confirmed across two independent tools and two different historical windows (current data and 2016/2021 data) — strengthening, though still not fully resolving, this as a genuine open question rather than a one-off retrieval failure. See TRACK_001_Master_Issues gap #49.

A genuinely surprising, NOT-yet-confirmed finding: the same extraction reported finding no current Sunshine List records for Homes First Society (last record 2011) or Covenant House Toronto (no records at all) — despite both being confirmed via T3010 to have substantial staff above $100k. This is flagged as requiring independent confirmation, not accepted as fact — it may reflect a legal entity name mismatch, a genuine reporting gap, or the same tooling pattern affecting these two specific organizations across multiple independent attempts by different tools. This is now the single highest-priority open item in this line of research.

Also resolved definitively: the underlying disclosure data structure contains only seven fields (sector, name, salary, taxable benefits, employer, job title, year) — no department, division, or region field exists. Salvation Army's Toronto-specific staff and TSSS-specific City of Toronto staff genuinely cannot be isolated from the raw disclosure data itself. This is a structural limitation of the Act's disclosure format, not a gap in this library's research.

What a second pass did add: a fuller City of Toronto executive breakdown (13,035 employees, 1,514 distinct positions, top earners including the Medical Officer of Health at $475,421 and multiple Deputy City Managers between $392,000-$427,000) — useful scale context, not TSSS-specific. And a second, more precise (though still third-party, non-primary) headcount estimate for Homes First: 275 employees per RocketReach (https://rocketreach.co/homes-first-profile_b5f7a7bbf42d2aaa), versus the earlier much wider "500-1,000" range from ZoomInfo. These two aggregator estimates disagree by a factor of nearly 2-4x with each other — which is itself the finding. Neither should be treated as authoritative; this is presented as evidence that third-party headcount aggregators for this sector are unreliable, not as a resolved number.

The Salvation Army entry requires a specific caveat: its Sunshine List filing (Governing Council of the Salvation Army in Canada) is a national entity covering all of Canada, not Toronto shelter operations specifically — 190 staff, $26.3M combined (2023), topped by a Chief Financial Officer at $357,234, none in Toronto-shelter-specific roles among the top five. Do not cite this figure as if it represents Toronto shelter spending — this is exactly the caution this document's own "How to Use This Safely" section already gives, now confirmed directly rather than assumed. This is not a restriction on access — the data is fully public — it is a scope limitation: the Salvation Army reports as one combined national entity, with no way to isolate Toronto-specific figures from the disclosure itself.

The headcount gap remains genuinely unresolved. A direct, sustained attempt across two research passes to find primary-sourced full-time/part-time employee counts for Homes First and Covenant House specifically — beyond CRA T3010's revenue/overhead fields, which don't capture this — found only conflicting third-party estimates with no organization-published breakdown located. This confirms, rather than closes, the gap HMK-037 already logged. Consistent with this library's standing discipline (see HMK-037 Part 4): this document will not manufacture a precise headcount figure where none was found, and will not treat one imprecise aggregator estimate as more authoritative than another merely because it sounds more specific.


THE THREE FACTS THAT MATTER

Fact 1: The contractor layer costs money the City doesn't count. The City's $136/night shelter accounting cost (Toronto AG 2025) is what the City pays per bed per night. Fred Victor's implied cost per night, when you include their overhead and capital servicing, is approximately $271/night — twice the City's accounting figure. Housing First (Latimer 2020: https://psychiatryonline.org/doi/10.1176/appi.ps.202000029) costs $17.29/night net. The City is paying for three pricing tiers simultaneously and calling only one of them the "cost of shelter."

Fact 2: Covenant House holds $48.5M in reserves while youth are turned away. Covenant House operates at 98% occupancy. They are legally turning away youth who need shelter. Their endowment — $48.5M, covering 1.55 years of operations — is CRA-compliant and properly governed. But it sits in reserve during the same period the City says there is a $164.87M federal funding cliff. This is not fraud. It is a two-tier system that raises a legitimate question: are we structured for efficiency or accumulation?

Fact 3: The system is 83-95% taxpayer-funded, and nobody audits it for exits. The Toronto AG 2025 found $2.9M in unrecovered operator surpluses. No operator is required to publicly report their housing exit rates. No contract includes a housing-exit performance requirement. The City pays per bed, per night — not per person housed.


THE COMPARISON

What we pay Annual cost per person Source
Emergency shelter (City accounting) $49,640 Toronto AG 2025 ✅
Emergency shelter (implied operator cost) $68,000–$99,000 Fred Victor T3010 arithmetic
Housing First (net after 69% offset) $6,311 Latimer et al. 2020 ✅
Saving per person housed vs sheltered $43,329 Arithmetic

WHAT TO ASK

For City Council candidates: "Will you support requiring all shelter operators to publicly disclose housing exit rates, overhead ratios, and reserve levels as a condition of City contract renewal?"

For journalists: "Four organizations receive $154.5M in public funding to operate Toronto shelters. Their overhead averages 29%. Their combined reserves are $86.8M. Nobody requires them to report how many people they house. What are we getting for $154M?"

For deputations: "The Toronto Auditor General found $2.9M in unrecovered surpluses and $13.2M in improper hotel charges. Yet the City has no housing-exit performance requirement in any shelter contract. Why does the City count beds and not people housed?"


HOW TO USE THIS SAFELY


THE ASK

Housing First costs $17.29/night. Emergency shelter costs $136/night. The difference is $118.71/night per person. Toronto shelters approximately 9,700 people every night. If just 10% of them were in Housing First instead of shelter — 970 people — the system would save approximately $42M per year net.

The barrier is not evidence. The barrier is a procurement system that pays for beds, not exits — and three levels of government each waiting for the other to fund the solution.


MAIN-DERIV-011 | Version 1.0 | June 30, 2026 | The Campaign for a Regenerative Ontario All T3010 figures: public CRA filings via Charity Intelligence Canada Legal review recommended before publication of executive compensation figures


TIER 1 GAP PROGRESS (June 30, 2026): TOWARD A FULL POS CONTRACT REGISTRY

Confirmed scale: TSSS "works closely with more than 75 community service providers and sector partners" (2024 Program Summary) ✅ — this confirms the 4-5 operators profiled above represent a small fraction of the full contracted network, not its entirety.

A fifth large named contract, confirmed: Canadian Red Cross Society holds Non-Competitive Blanket Contract #47024284 for refugee/asylum-seeker lodging services. Value was increased via Council authority from $60,137,629 to $74,095,341 (net of HST) to maintain winter/early-spring service levels. ✅ [Council Item 2024.EC17.4: https://secure.toronto.ca/council/agenda-item.do?item=2024.EC17.4] This is comparable in scale to Fred Victor ($61.1M) and larger than Dixon Hall ($28.3M) among the operators already profiled — and notably, it is a non-competitive contract, awarded without the competitive procurement process used for most other shelter operator agreements, reflecting the emergency nature of the refugee response.

Workforce wage floor confirmed: New shelter operators (via the City's Expression of Interest process) must "commit to paying all front-line staff... an appropriate wage (minimum of $53,000 annual or higher, based on 2025 figures)." ✅ This is a useful benchmark against HMK-015's proposed Peer Navigator Corps wage ($55,000) — the City's own minimum for new shelter contracts is already close to that figure, suggesting the Peer Navigator proposal is consistent with, not far above, current sector wage floors.

Still open: A complete itemized list of all 75+ TSSS-contracted organizations with individual contract values. This remains genuinely unaddressed and is the right scope for a dedicated Toronto Open Data / FOI-based research pass — the "Daily Shelter & Overnight" dataset on the Open Data Portal may be the right starting point for site-level (not contract-value) data.

Frozen v1 edition (July 2026) · published 2026-08-12 · corrections to the living library are welcome — tell us where we’re wrong.