Loading...
Structural Analysis
Governance & AccountabilityData sourced from:
The Ontario Autism Program's institutional architecture — the funding model, the procurement choices, the accountability gaps — maps precisely onto a known organizational pattern: reducing public accountability for public spending while maintaining political control.
I. The Question
Almost every story written about the Ontario Autism Program (OAP) asks the same question: why isn't it working? The waitlist is treated as a failure — a problem to be solved, a gap to be closed, a scandal to be fixed. That framing is wrong. Or, more precisely, it is incomplete.
The question a business school analyst would ask first is different. Not "why is it broken?" but "what is it designed to do?" Public programs are organizations. Organizations have structures. Structures produce predictable outcomes. If the outcome of a program is that public money flows through it without being auditable, and the public cannot easily see who received what or why — that is not a bug in the system. That is the system.
This article does not allege wrongdoing. It does not accuse any individual, minister, vendor, or public servant of any impropriety. It does something more useful for journalists, opposition staff, and parents: it takes the OAP apart as an org chart and examines what the org chart is for. The thesis is narrow and structural. The OAP's institutional architecture — the funding model, the procurement choices, the information flows, the lobbying activity — maps precisely onto the organizational playbook used when governments want to reduce public accountability for public spending while maintaining political control. This is not conspiracy theory. It is organizational design. Governance scholars call it the accountability-deflection model: a structure that shifts statutory oversight away from public bodies and toward arm's-length intermediaries that are exempt from freedom-of-information law, independent audit, and direct legislative scrutiny.
II. The Standard Model
In the standard public-program accountability model taught at every policy school — the one that underpins the Canadian Auditor General system, the UK National Audit Office, the US GAO — four structural features appear without exception.
Each of these features exists because, in its absence, a predictable set of failures follows. The governance literature calls this the principal-agent problem: when the agent controls the information the principal needs to hold the agent accountable, the agent will act in the agent's interest. Every one of the four features above exists to solve that information asymmetry. Remove any one, and the agent gets discretion. Remove all four, and the agent gets silence.
III. The OAP Model
The OAP departs from the standard accountability model on every one of the four structural features above. Not on one. On all four. That is the finding that should have been the lede in every news story since 2019.
Departure 1 — The Principal Cannot See
AccessOAP — the single-point-of-access intake layer through which every OAP family must pass — is operated by a third-party contractor and is not subject to Ontario's Freedom of Information and Protection of Privacy Act (FIPPA). That is not an oversight. FIPPA applies to "institutions" as defined by statute. When the province procures a service through a private entity structured outside that definition, the FIPPA lever disappears.
Departure 2 — The Spending Ratio Is Inverted
In 2023–24, the OAP budget was $691.2M. Of that, core clinical therapy reached $307.3M — 44.5%. The remaining 55.5% was absorbed by administration, intake, foundational services, and program overhead. A clinical program spending more on administration than on clinical services is, in a governance review, the headline finding. In the OAP it is the unexamined baseline.
Departure 3 — Accountability Is Fragmented
The OAP is not one thing. It is a ministry (MCCSS), an arm's-length intake contractor (AccessOAP), a designated clinical provider network, a foundational services network, and a family-directed funding stream. Each layer reports differently. Each has a different data system. Each publishes different metrics on different schedules. Consolidation of accountability — the single line item a public-accounts committee can interrogate — does not exist.
Departure 4 — No Independent Audit Has Occurred
AccessOAP, despite absorbing $57.9M per year in public funds, has never been audited by the Office of the Auditor General of Ontario. The FAO conducted a program-level review in June 2024, which is the only near-independent examination in the public record. No line-item audit of AccessOAP operations, vendor costs, staffing, or subcontracting has been tabled in the legislature.
IV. Principal-Agent Breakdown
In a functioning public program, the principal is the public. The elected government is an intermediary. The ministry is an agent. The contractor is a sub-agent. Information flows upward from contractor to ministry to cabinet to legislature to public, and accountability flows downward in the reverse direction. The principal at the top of the chain sees everything.
In the OAP, that chain is inverted at two structural points. First, AccessOAP's FIPPA exemption severs the information link at the lowest tier — the public cannot see what the contractor is doing without the contractor's consent. Second, the MCCSS annual reports aggregate OAP data at a level of abstraction that does not permit reconstruction of the underlying ledger. The public sees totals. The public does not see flows.
The practical result is that the effective principal of the OAP is not the public. It is the minister. And the agents — the intake layer, the designated providers, the foundational services network — report to the minister on terms the public cannot verify. This is a well-documented governance failure mode. In the corporate-governance literature, it is called information asymmetry between principal and monitor: the person who is supposed to verify the program's performance cannot see the numbers that would let them do so.
The structural point. When a principal cannot verify an agent's work, the agent's discretion expands. The agent's discretion is not, in itself, proof of wrongdoing. It is proof that wrongdoing, if it occurred, would be undetectable. That is the governance risk. It does not require malice. It only requires opportunity, and the opportunity is architectural.
V. The AccessOAP Design
AccessOAP is the operational hinge of the entire OAP. Every family that enters the program passes through it. Every funding determination depends on its intake assessment. Every data point the ministry publishes about wait times, service volumes, and registrations originates in AccessOAP's systems. Whoever controls AccessOAP controls the program's factual record.
In 2020 and again in 2022, the province procured AccessOAP operations through a contracted third party rather than bringing intake inside the ministry. That is a procurement choice, not a necessity. Other Canadian jurisdictions — Alberta, British Columbia — run equivalent intake functions inside government or through public agencies that remain subject to freedom-of-information law. Ontario's choice was different. Ontario's choice placed intake outside FIPPA's reach.
The governance parallel is precise. In the academic literature on "agency capture" and "accountability offshoring," this structure appears repeatedly: a government creates an arm's-length entity that performs a core public function, writes the contract in a way that places the entity outside statutory transparency regimes, and then points to the arm's-length relationship when asked for records. The structure is lawful. The structure is also, predictably, opaque.
What the contract structure permits
Each of those is individually defensible. Each has a rationale offered in the contract. Stacked together, they describe an information environment in which the Office of the Auditor General cannot initiate a financial audit of intake operations without ministerial cooperation, and in which opposition MPPs cannot obtain the operational detail that would let them ask informed questions in committee.
VI. Spending Without Auditing
$57.9M is not a rounding error. In Ontario's public-sector financial reporting, any standalone program running at that scale would ordinarily be subject to at least one value-for-money audit within its first five years of operation. The Office of the Auditor General of Ontario has a statutory mandate to conduct such audits. The Auditor General has, to the public record, not yet conducted one of AccessOAP.
What does $57.9M per year purchase? The public record tells us it operates a web portal, a call centre, intake coordination, a determination-of-needs process, and family-facing case management. What the public record does not tell us is: how many of those dollars are staff, how many are vendor fees, how many are software licences, how many are subcontracted to other firms, and at what margin the contractor operates. Those are ordinary questions a public-accounts committee would ask. They are not answerable from publicly available records.
The structural point here is subtle but important. The issue is not that we know the money is wasted. The issue is that we cannot know whether it is or not. A governance system that makes that question unanswerable has, in effect, answered it — by removing the mechanism that would force an answer.
VII. The Information Asymmetry
A functioning oversight ecosystem needs three groups to be able to see program data: the families the program serves, the legislators who fund it, and the journalists who report on it. In the OAP, each of these groups faces a different structural barrier.
Families can see their own case. They cannot see comparable cases. They cannot see average wait times by region. They cannot see the distribution of funding amounts by diagnosis or geography. When a family asks "is my child being treated fairly relative to others?" the program does not publish the data that would answer the question. 88,175 registered children, approximately 20,666 receiving funded services, approximately 67,509 waiting an average of four-plus years — these are aggregate numbers. The distributions beneath them are not public.
Order-paper questions to the minister return aggregate answers. FIPPA requests to AccessOAP return jurisdictional refusals. Estimates committee can interrogate the minister but cannot compel the contractor. The result is that the parliamentary tools designed for exactly this kind of oversight return the same information the government already chooses to publish.
Investigative reporters are left with the three tools they always have: leaked documents, whistleblowers, and the FAO's periodic reviews. None of these scale. None of these produces the line-item detail that would let a reporter compare, for example, AccessOAP's staffing ratios to analogous intake functions in other provinces. The reporting that does exist is excellent. The reporting that would exist if the data were public is absent because the data is absent.
That configuration is not random. It is the configuration that results when the program is designed around political manageability rather than public verifiability. In governance theory, this is called asymmetric transparency: the principal who controls the program has full information; every other principal has the information the controlling principal chooses to release. The OAP is a textbook case.
VIII. The School Asset Parallel
If the OAP were the only place this structure appeared in Ontario governance, it could be written off as a one-program anomaly. It is not. The 2025 amendments to the Education Act place eight Ontario school boards under provincial supervision and introduce a centralized framework under which approximately $123.3B in school assets — land, buildings, equipment, consolidated operating budgets — become centrally manageable through a mechanism that, like AccessOAP, operates outside the ordinary municipal-accountability regime that governed school boards previously.
The parallel is structural, not rhetorical. In both cases the move is the same: a public function previously lodged in an institution subject to local democratic control and statutory transparency is re-lodged in a centralized or arm's-length layer with fewer statutory oversight hooks. The governance pattern literature calls this institutional isomorphism in accountability avoidance — the same structural response appears across different policy domains when the underlying incentive is the same: control without scrutiny.
Noting the parallel does not require attributing motive. It requires only noting that the same architecture, once built in one policy area and shown to be politically durable, becomes the template for the next. That is how organizational design propagates in government. It is how it has always propagated.
IX. The Counterfactual
The structural critique above implies its own remedy. If the OAP's accountability architecture departs from the standard model on four features, the reform is to restore those four features. None of this is radical. None of this requires new law — most of it requires only a procurement decision or a ministerial directive.
Each of these is a well-established public-administration practice. Collectively, they would not "fix the waitlist." They would do something narrower and more consequential: they would make it possible for the public to know whether the waitlist is being managed in good faith. That is a precondition for fixing it.
X. The Organizational Verdict
The framing that has dominated OAP coverage for seven years — "the program is broken, the government should fix it" — has not worked. It has not worked because it accepts a premise that the structural evidence contradicts. A broken program is one whose design is intended to produce an outcome it is not producing. The OAP's design is producing precisely the outcome its design predicts.
The budget has grown from roughly $300M in 2019 to $965M in the 2026–27 expenditure estimates. Over that same period, the number of children receiving clinical services as a share of registered children has not meaningfully improved. The number of children waiting has grown. The share of spending reaching clinical therapy has, according to the FAO, remained below half. And the accountability layer through which all of this flows has been moved further from statutory transparency, not closer to it.
A system that spends more, reaches fewer, and is audited less is a system performing to specification. The specification may not be the one stated in press releases. But it is a specification, and it is visible in the org chart.
The World Health Organization's guidance on autism intervention identifies ages 0–6 as the critical developmental window. For every year a child waits, the structural architecture of the OAP is doing exactly what it is built to do: absorbing spending, producing aggregate numbers, and remaining beyond the reach of the tools that would compel it to do otherwise. The child, meanwhile, ages out of the window in which the intervention works.
The remedy is not louder outrage. The remedy is a procurement amendment, an Auditor General referral, a standing committee motion, and a mandatory reporting regime. Each of those is achievable. None of them requires anyone to behave badly to be exposed. They require only that the principal — the public — recover the tools the current architecture has put out of reach.
That is the organizational verdict. The OAP is not a failure of execution. It is a success of design. The question for the legislature, for journalists, and for families is no longer why isn't it working? The question is: whose design is this, and are we willing to leave it in place?
Verified Facts