uniteTOlove.ca

Mental Health Crisis Response — Playbook

Who actually shows up to a mental-health crisis call in Toronto — police, or the newer non-police team — and at what cost.

DRAFTThe playbookThe evidence file

What Toronto can actually do on mental health crisis response — each move with its costs, its beneficiaries, and its receipts.

v2.0 · 2026-08-11

---

The honest bottom line

Toronto already paid for the answers here. The Toronto Community Crisis Service (TCCS) — mental-health workers instead of police — has been operating citywide since July 2024 and is still growing: in November 2025 it started piloting embedded teams on the busiest subway stations, and in 2026 its call-dispatch function is moving into Toronto Paramedic Services. Its reported diversion rate has climbed from 78 percent in 2024 toward a 90 percent target by 2027. Toronto Police Service's own budget documents back this up from the other side: mental-health calls fell 18.5 percent in 2025 versus 2024, while referrals from police to the crisis service rose 63 percent over the same stretch — two separate City documents, produced by two different divisions, independently telling a consistent story, though consistency isn't proof one caused the other.

There's a real, unresolved problem inside TCCS's own headline number. This project's own earlier research says the crisis service "diverted 7,800" calls out of 10,339 received, but the City's actual 2026 budget notes say something different: 7,800 is the total number of 911-sourced calls received, and 6,891 of those — not 7,800 — were the ones actually resolved without police. Worse, 6,891 divided by 7,800 is 88.3 percent, not the 78 percent the City's own sentence states right next to those two numbers — ten percentage points too big to be rounding, and all three numbers come from the same sentence in the same document. Until someone at the City reconciles it, treat the 78 percent diversion rate — the single most-quoted number in this file — as reported but not confirmed accurate. That's a reason to distrust quoting the program's own numbers without checking them, not a reason to distrust the program itself.

The two American programs cited constantly as proof this model works aren't what they used to be: CAHOOTS stopped serving Eugene, Oregon entirely in April 2025 — the city where it was invented for over three decades couldn't keep funding it, though it still runs in neighbouring Springfield — and Denver's STAR program is still operating five years in but never reached the round-the-clock coverage it set out to achieve, with reporting on its fifth anniversary calling that "unlikely given Denver's current budget woes." Toronto's own program, by contrast, is the one still expanding — a genuinely different story, but it means Toronto's case rests on Toronto's own evidence, not borrowed American success stories that have since developed real cracks.

Get an independent, non-self-reported look at what happens after the crisis call ends. (a recommendation card) Not just whether police were avoided, but whether people stay out of the ER, stay housed, stay stable, months later. The program's own numbers don't yet answer that, and don't even fully agree with themselves.

Use the City's own Auditor General's list — six specific, already-named call types that could go to a crisis team instead of police — and actually pilot routing some of them there. (a recommendation card) Well-being checks, noise complaints, and four others, instead of leaving a finding the City already produced sitting unused.

Neither requires inventing a new number or a new program — both require reading what the City already published, including the parts that don't reconcile cleanly, and the parts about other cities that don't flatter the case as much as the highlight reel usually does. This file doesn't confirm what TCCS itself actually costs as its own line item, separate from the broader $50.7 million community-safety budget category it sits inside; a $31.7-to-$35.3-million figure circulates but wasn't confirmed against a primary City document. It also doesn't update TCCS's own safety record — how often its own workers still need police or ambulance backup — past a 2022 pilot-era snapshot (2.5 percent police, 1.7 percent ambulance), and doesn't say whether any provincial mental-health dollars, as opposed to City tax dollars, actually fund the service.

---

a recommendation card — Publish an Independent Outcomes Framework for TCCS, Beyond Diversion Rate and Satisfaction

Card id: a recommendation card · Issue: mental-health-crisis-response · Backgrounder: our research file for that page · Trust: New load-bearing findings (backgrounder NEW-2026-2, NEW-2026-3, NEW-2026-16, NEW-2026-18)

Problem

TCCS's own published metrics are, so far, almost entirely diversion-rate and satisfaction-based: a 78%-to-90% diversion-rate trend and a rising call-volume target through 2027 [NEW-2026-2], plus a 2022 pilot-era finding that 25%+ of service users were enrolled in post-crisis case management and 799 cases received a referral [NEW-2026-16]. No source located in this review reports outcomes past the point of referral — repeat-crisis rate, ER admission avoidance, or housing/health stability at 3, 6, or 12 months. This page’s own backgrounder also documents a real, unresolved arithmetic discrepancy in TCCS's headline call/diversion figures across its own inherited claim and this review's directly-quoted primary source [NEW-2026-3] — including an internal arithmetic error in the City's own primary source (6,891/7,800 = 88.3%, not the "78%" the same sentence states) that this backgrounder flags explicitly rather than silently resolving — a second, independent reason an outside outcomes framework matters: the program's own self-reported numbers do not yet fully reconcile, and the single most-quoted diversion figure has not been confirmed accurate even on its own terms. This card addresses only the outcomes-measurement gap, not TCCS's operating model or staffing.

Action

The City commissions and publishes an independent (third-party, not TCCS-self-reported) outcomes evaluation tracking post-referral results — repeat-crisis-call rate, ER visit rate, and housing/treatment-engagement status at fixed follow-up intervals — modelled on the same third-party-evaluator structure the City already used for TCCS's original 2022-2023 pilot evaluation (Provincial System Support Program and Shkaabe Makwa with CAMH, per this page’s own inherited pilot-era record), and reconciles the call/diversion figure discrepancy this backgrounder documents as part of the same review.

Jurisdiction split

Cost

Order-of-magnitude: low — a program-evaluation contract comparable in scale to the City's own original 2022-2023 third-party TCCS pilot evaluation (a defined-scope academic/health-system partnership, not a large capital or ongoing program cost) [NEW-2026-16]. No specific dollar figure for that original evaluation was located in this review; this card names the existing evaluation as its cost comparator rather than inventing a figure.

Funding path

Existing Social Development / TCCS program-evaluation budget capacity, following the same model already used once; no new funding mechanism is proposed.

Who benefits, and how

People who use TCCS, via a program that can demonstrate (or correct course on) what happens to them after the crisis call ends, not only whether police were avoided at the moment of contact; City Council and budget staff, via a real basis for the 2027+ scale-up targets already set in the current budget's call-volume trajectory [NEW-2026-2], rather than targets set on diversion-rate momentum alone.

Who bears the cost, and how

City taxpayers, via existing Social Development evaluation-contracting capacity — no new payer class identified; the cost is small relative to TCCS's existing budget footprint inside the $50.7 million Community Safety and Wellbeing service line [NEW-2026-9].

Who benefits from the status quo

No beneficiary identified. The backgrounder's Cui Bono table (added 2026-07-14) is empty after checking this library's internal records and a targeted live-discovery search — TCCS's own delivery structure runs through nonprofit anchor partners under a City contract, and no ESTABLISHED or REPORTED finding was located naming any entity profiting from the current absence of an independent outcomes framework specifically.

Financial ROI

Not separately estimated. A real outcomes framework is a precondition for making any financial-ROI case for further TCCS scale-up (e.g., avoided ER or repeat-crisis costs) that is not simply an inference from the diversion rate — this card produces the missing evidence rather than assuming the answer.

Economic ROI

Not yet estimable as a societal/economic-growth figure specific to this action (an evaluation contract). The closest available literature concerns the crisis-response program itself (TCCS), not the evaluation this card proposes, and is Financial/cost-avoidance in character rather than broader economic impact: Denver's STAR program (Urban Institute, "Understanding Denver's STAR Program," 2023) found a $151-per-offense-reduced cost against a $646 average direct criminal-justice cost per minor offense — a co-response/diversion program's offense-reduction roughly four times cheaper than the criminal-justice-system alternative. This is offered only as context for why a rigorous outcomes evaluation of a comparable program is a reasonable comparator class, not as a dollar estimate for this card's own action, which has no comparator of its own in the sources located. Confidence: low — the Denver figure is borrowed from a different city/program and measures the crisis-response program's own cost-effectiveness, not the evaluation contract this card actually proposes.

Social ROI

Directional: this page’s inherited trauma-informed-care findings show that structured, trauma-informed intervention has measured, cite-able downstream effects elsewhere in the same care system (a 50% EMS-call reduction following trauma-informed shelter staff training [CL-371]) — a comparable, independently-measured outcomes case for TCCS itself does not yet exist and this card exists to build it.

Environmental ROI

Genuinely environmentally neutral. Commissioning and publishing a third-party outcomes evaluation is a data/reporting exercise with no material emissions, land-use, water, waste, or resilience footprint. Confidence: high — the action type (a paper-based evaluation study) has no plausible environmental pathway.

Evidence

Confidence & uncertainties

Medium-high confidence on the problem (the outcomes-measurement gap and the arithmetic discrepancy are both directly documented from primary sources). Low confidence on cost, since no comparator dollar figure for the original 2022-2023 evaluation was located in this review. This card does not resolve whether the discrepancy in TCCS's own figures reflects a reporting artifact or a real double-counting; it proposes the review that would find out.

Status

DRAFT — blocked on: fairness and legal review; locating a cost figure for the City's own original TCCS pilot evaluation as a firmer cost anchor.

---

a recommendation card — Formalize 911/211 Dispatch-Routing Rules Using the Auditor General's Own Named Six Call Types

Card id: a recommendation card · Issue: mental-health-crisis-response · Backgrounder: our research file for that page · Trust: earlier research plus new load-bearing findings (CL-90662 + backgrounder NEW-2026-4, NEW-2026-5, NEW-2026-11)

Problem

This page’s own inherited finding states the City's own Auditor General has already identified six further 911 event types — beyond TCCS's current mental-health-crisis mandate — that could potentially be handled by a non-police responder: check address, check well-being, dispute, noisy parties, landlord and tenant dispute, and unwanted guest [CL-90662]. Separately, TCCS's own call-intake function is being restructured in 2026, moving dispatch from the 211/911 chain into Toronto Paramedic Services [NEW-2026-5], and TPS's own 2026 budget documents show a 63% year-over-year rise in TPS-to-TCCS call referrals over the same period mental-health-call volume fell 18.5% [NEW-2026-11] — evidence that call-routing changes already measurably shift volume between services. No source located in this review shows the City has formally expanded TCCS's dispatch-eligible call types to include the Auditor General's own named six, despite that finding already existing in the record. This card addresses only that specific, named, City-sourced expansion opportunity — not TCCS's funding level, staffing, or the mental-health-crisis mandate it already serves.

Action

The City directs Toronto Paramedic Services (as TCCS's incoming dispatch authority) and 911/211 dispatch operations to pilot routing a defined subset of the Auditor General's own named six non-mental-health-crisis call types — starting with the lowest-risk of the six (e.g., "check well-being" and "noisy parties") — to TCCS or an equivalent non-police response, with outcomes tracked using the same evaluation structure proposed in a recommendation card.

Jurisdiction split

Cost

Order-of-magnitude: low relative to a new-program launch, since this uses TCCS's existing (and, per NEW-2026-2, already-growing) capacity and its 2026 dispatch-integration transition already underway [NEW-2026-5, NEW-2026-6] — comparator: the $1.8 million TCCS-TTC pilot [NEW-2026-8] is the closest named, City-funded precedent for "expand TCCS into a new, defined call category as a funded pilot," and is used here as the cost-scale anchor rather than a specific new figure, since no dedicated budget line for this specific expansion was located.

Funding path

Existing TCCS/Toronto Paramedic Services dispatch-integration budget already underway in 2026 [NEW-2026-5, NEW-2026-6], expanded incrementally; if a dedicated pilot budget line proves necessary, the $1.8 million TCCS-TTC pilot [NEW-2026-8] is the named, existing-program precedent for how such a request has been structured and funded before.

Who benefits, and how

Toronto residents who currently generate a "check well-being," "noisy parties," or similar 911 call that does not require an armed response, via a non-police responder matched to the actual need; Toronto Police Service, via reduced call volume freeing capacity for calls that do require police, consistent with TPS's own reported 18.5% mental-health-call reduction and 63% referral increase trend already underway for the mental-health-crisis category specifically [NEW-2026-11].

Who bears the cost, and how

City taxpayers, via the existing TCCS/Toronto Paramedic Services operating budget inside the Community Safety and Wellbeing service line [NEW-2026-9] — no new payer class identified, since this expands an already-funded dispatch-integration process rather than creating a new funding stream.

Who benefits from the status quo

No beneficiary identified. The backgrounder's Cui Bono table (added 2026-07-14) is empty after checking this library's internal records and a targeted live-discovery search — no ESTABLISHED or REPORTED finding was located naming any entity profiting from the six named 911 call types remaining routed to police rather than TCCS.

Financial ROI

Not separately estimated; no source reviewed quantifies a per-call cost difference between a police response and a TCCS response for these specific six call types. The existing TPS-referral-volume trend [NEW-2026-11] suggests the mechanism (routing calls to TCCS reduces measured TPS call volume) already works for the mental-health-crisis category; whether it would produce a comparable measured shift for the Auditor General's six named types is untested in the sources reviewed.

Economic ROI

Not yet estimable as a societal-growth/employment figure specific to routing these six call types. The nearest real literature is a cost-avoidance case for a comparable diversion pilot in a different city, offered as directional context only: Denver's STAR program (Urban Institute, "Understanding Denver's STAR Program," 2023) reported roughly 30,000 calls reassignable to an alternative responder over a six-month evaluation window and a per-offense-reduced cost about four times cheaper than the criminal-justice-system comparator ($151 vs. $646) — a Denver-specific, different-call-mix figure borrowed as an order-of-magnitude signal that call-diversion pilots can show measurable volume/cost effects, not a Toronto-specific or six-call-type-specific estimate. Confidence: low — figure is borrowed from a different city and a different (broader) call-type mix than the Auditor General's six named types; no Toronto-specific or call-type-specific economic modelling exists in the sources reviewed.

Social ROI

Directional: consistent with this page’s overall equity framing (inherited from the master briefing, cited there rather than re-derived here) that a health/social response matched to actual need reduces unnecessary police contact for situations that do not require it — no claim in the claims register yet quantifies this specifically for the six named call types.

Environmental ROI

Genuinely environmentally neutral. Re-routing dispatch protocol for existing call types between existing response services (police vs. TCCS) involves no new construction, fleet expansion, or land-use change identified in this review. Confidence: medium — high confidence the action itself is neutral; slightly reduced because a pilot could, in principle, involve incremental vehicle trips not assessed here.

Evidence

Confidence & uncertainties

Medium confidence. The underlying opportunity (the Auditor General's own named six call types, not yet formally routed to TCCS per any source located) is well-documented and City-sourced. The cost and financial-ROI sections are explicitly anchored to a comparator pilot rather than a dedicated costed estimate for this specific expansion, since none was located. Whether operational or legal barriers (beyond funding) block routing these specific call types to TCCS was not addressed by any source in this review and is flagged as a real unknown, not assumed away.

Status

DRAFT — blocked on: fairness and legal review; confirming whether the City has already begun any of the six call-type expansions internally without a public announcement (this review did not locate one, but cannot rule it out).

---

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. Written per this library's standard page structure. Every factual premise traces to a formally registered claim, a carried-forward claim, or a NEW-2026-# source quote in the backgrounder — no figure invented. Per the costing bar (Q-06), all costs are order-of-magnitude ranges anchored to named comparators, never fake-precise line items. Note on jurisdiction: unlike several other pages' cards, this library's issue index currently has no ratified row (and therefore no Owner column) for this issue — both cards state jurisdiction based on what this review's sources directly show TCCS's actual governance and funding structure to be, flagged explicitly wherever a funding-source question could not be confirmed. 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/mental-health-crisis-response.md (retired day-one memo, a recorded standing decision); each card tightened, verbose ROI Range/Comparator-source/Confidence blocks collapsed into flowing one-line-per-dimension form; repeated per-card header fields (File:, duplicate slug/Backgrounder/Provenance/Date lines) condensed to a single header line; all citation tokens (CL-371, CL-90662, carried-forward, NEW-2026-2, NEW-2026-3, NEW-2026-4, NEW-2026-5, NEW-2026-6, NEW-2026-8, NEW-2026-9, NEW-2026-11, NEW-2026-16, NEW-2026-18, NEW-2026-#, New load-bearing findings) preserved verbatim.

About this file