Power Doesn’t Attract People With Diminished Empathy. It Produces It.

Between 2019 and 2024, provincial payments to Accerta Services Inc. — the private company that administers Ontario’s autism program, along with Ontario Works and the Ontario Disability Support Program — more than tripled, from $202.5 million to over $670 million.² Across the same period, the number of children waiting for funded therapy roughly quadrupled, and at no point did more than a quarter of them receive the core clinical services the program exists to deliver.³⁷ Both facts are drawn from public records. Only one of them appears in the government’s own reporting.

As of May 13, 2026, 91,974 children are registered in the Ontario Autism Program. Of those, 71,263 are waiting for a Core Funding Agreement; 20,711 have active funding.¹ A child registered today waits, on the Ontario Autism Coalition’s reading of the current invitation cohort, at minimum 5 years — entering the program at eight if they were diagnosed at three, long after the developmental window that makes early intervention most effective has closed.

The province describes this as a nearly $1 billion investment. The question this investigation asks is narrower and answerable from the record: of that money, how much reaches a child, how much is consumed administering the queue, and what does the province publish about the difference? The answers, in order, are less than half, a growing share, and — since 2024 — nothing that arrives without a freedom-of-information request.

I. The Architecture of Extraction

Privatization in publicly funded health systems does not arrive as a press conference. It arrives as a procurement decision: the creation of a third-party administrator, or TPA. In Ontario, the TPA for the autism program is Accerta Services Inc., contracted in 2021 to handle intake through what was branded AccessOAP.

The tripling is documented in provincial spending data,² and the arrangement is not unique to autism: an OPSEU analysis found that the top three private administrators of Ontario government programs collected over $1 billion in a single year.³

The autism-program share of that total is documented separately, and it is much smaller: MCCSS records obtained under freedom of information put AccessOAP intake and administration at $57.9M in 2023-24.²⁶ Both numbers matter, and conflating them concedes the argument to anyone holding the spreadsheet.

Meanwhile, the company fell well short of its targets. Accerta missed its target enrolment by approximately 5,000 children and youth in 2024.⁴ Parents described the annual re-interviews of children already receiving services as a “make-work project” for the private administrator — bureaucratic friction that generates billable administrative hours while consuming time families do not have.

“This is an example of how the Ford government has taken public money and put it into black boxes so it’s impossible to trace.”

— JP Hornick, President, Ontario Public Service Employees Union (OPSEU), June 2026

The Financial Accountability Office estimated in 2020 that clearing the waitlist entirely would cost approximately $1.35B in Year 1.⁵ The province chose to fund at $600 million. The gap is structural: the system was built to under-serve, and the administrator’s workload — and therefore its billing — scales with the volume of intake processing rather than with the number of children who receive an hour of therapy.

II. The Evidence the Province Does Not Cite

What the Cochrane Collaboration found

Applied Behaviour Analysis is routinely described in Ontario policy documents as the “gold standard” intervention for autism. The phrase has achieved the status of received wisdom — repeated so often that questioning it feels like questioning gravity. But the highest tier of evidence synthesis tells a different story.

The 2018 Cochrane systematic review of early intensive behavioural intervention (EIBI) — the most rigorous class of ABA program — found five relevant studies and concluded that “the quality of the evidence is weak” and that results “should be interpreted with caution.”⁶ Only one of those studies used random assignment, the minimum threshold for causal inference.

A 2022 review chapter in The Wiley-Blackwell Handbook of Childhood Social Development, as cited by Reframing Autism’s position statement on interventions, concluded that intervention effects did not generalise beyond the intervention environment and that there was inadequate collective evidence for change on any selected outcome.⁷ That chapter reaches us through the secondary citation and has not been independently verified against the original volume; it is reported here at that strength and no higher.

In a 2019 report to the United States Congress, the Department of Defense stated that after a year of ABA treatment through its Autism Care Demonstration program, 76% of participants showed “little to no change in symptom presentation,” and 9% showed worsening symptoms.⁸

This is the evidence base on which Ontario has staked nearly $1 billion per year in public funds. The literature that supports ABA as the dominant intervention is, by the field’s own highest standards of review, weak — and it is produced overwhelmingly by researchers who profit from the intervention’s continued market dominance.

What the WHO actually recommends

Canada is a signatory to the WHO’s Comprehensive Mental Health Action Plan 2013–2030, which calls on member states to address gaps in “early detection, care, treatment and rehabilitation for mental and neurodevelopmental conditions, which include autism.”⁹

The WHO does not recommend ABA as a standalone gold standard. Its flagship autism intervention is the Caregiver Skills Training (CST) program, developed with Autism Speaks, which is free, deliverable by non-specialists, based on naturalistic developmental behavioural interventions (NDBIs), and designed to empower parents rather than create institutional dependency.¹⁰ The CST program consists of nine group sessions and three home visits, teaching caregivers to embed learning into daily routines.

The WHO’s 2023 fact sheet on autism states that “evidence-based psychosocial interventions can improve communication and social skills,” and emphasizes that “care for people with autism needs to be accompanied by actions at community and societal levels for greater accessibility, inclusivity and support.”¹¹ This is a fundamentally different philosophy from the 20–40-hours-per-week clinical model that ABA providers sell and private equity firms invest in.

III. The Global Profit Machine

Ontario’s autism landscape does not exist in isolation. It is one node in a continental industry whose growth metrics would be the envy of any tech startup.

The U.S. ABA therapy market is estimated at $25–35 billion, growing at 10–13% annually.¹² Private equity firms completed 85% of all mergers and acquisitions in the autism healthcare segment between 2017 and 2022 — a rate not found in any other segment of healthcare.¹³ A JAMA Pediatrics study identified 574 autism therapy centres acquired by PE firms across 42 states, with acquisitions peaking between 2018 and 2022.¹⁴

The Wall Street Journal’s 2025–2026 investigation documented billing fraud, phantom services, and price inflation across the industry. Insurer Aetna reported a 300% increase in investigations of likely fraud or abuse by autism-therapy providers between 2024 and 2025.¹⁵

The most instructive cautionary tale is the Center for Autism and Related Disorders (CARD), once the world’s largest autism therapy provider with 265 locations, acquired by Blackstone in 2018. By 2023, CARD had closed 100 centres and filed for bankruptcy — leaving families without services and demonstrating the human cost of the leveraged buyout model applied to vulnerable populations.¹⁶

“They’d have to be doing these services 24 hours a day, seven days a week.”

— Pam Kehaly, Blue Cross Blue Shield of Arizona, on the therapy hours some providers bill per child

IV. The Ontario Arithmetic

The numbers tell a story of structural triage. Autism prevalence among Canadian children has risen substantially over the past two decades, consistent with global trends and with the Public Health Agency of Canada’s own surveillance reporting.¹⁷ Ontario’s working prevalence figure is 1 in 50.

The waitlist has grown in direct proportion. Under the Ford government it has roughly quadrupled, from approximately 23,000 registered at the 2019 redesign to 91,974 today, with 71,263 of them waiting for core clinical funding.¹⁸ The province increased OAP funding to $965 million for 2026-27. But funding has not kept pace with demand, and the structure of the program ensures it cannot: the FAO estimated that $600 million would serve only 10,142 children, while even $723 million reaches only 12,629.¹⁹

The journey of a typical child

1. Age 2–3

   Parents observe developmental differences. Pediatrician referral begins. Diagnostic waitlist: 12–18 months in many regions.

2. Age 3–4

   Diagnosis received. Family registers with the Ontario Autism Program. Clock starts on the Core Clinical Services waitlist.

3. Age 4–7

   Wait continues. Foundational Family Services available but limited. Critical early intervention window narrows. Families with means pay out of pocket. Others wait.

4. Age 8+

   Invitation to Core Clinical Services arrives — 5 years after registration. Invitations are currently reaching August 2021 registrants (Ontario Autism Coalition, July 2026 (reported from community intake)). The most neuroplastic developmental window has largely closed. Funding ranges $6,600–$65,000/year based on needs assessment.

Some families, unable to wait, have relinquished custody of their children to access services through the child welfare system — a measure of desperation that no policy framework should produce. The Ontario Autism Coalition’s 2025 community survey describes a “system in freefall.”²⁰

V. The Children Are the Back End of the Business

Follow the money past the waitlist and you arrive at the system’s most expensive terminus: emergency residential placement. This is where Ontario’s autism architecture stops being merely under-resourced and becomes structurally extractive — where the children who fall through every gap become, paradoxically, the highest-value line items in the ministry’s budget.

MCCSS spent an estimated $19.4 billion in 2023–24 across all programs.²¹ Within that, the “Children and Youth at Risk” vote-item funds child protection services delivered by children’s aid societies, youth justice, and residential placements for the province’s most complex cases. The children who land here — many of them autistic, many with co-occurring mental health conditions — are not being served by the autism program. They have aged out of it, been denied by it, or their families have collapsed under the weight of waiting for it.

The economics of abandonment

The costs are staggering. A Toronto Star investigation found that emergency placements for children with complex needs — in hotels, Airbnbs, rental apartments, even camping trailers and children’s aid society offices — can cost upward of $200,000 per child per year, and in some cases have approached $1 million.²² Often, the children in these placements are receiving no treatment at all. They are being warehoused at premium rates.

The number of children in these unlicensed emergency placements has nearly tripled since 2021. Between April 2023 and March 2024, 339 children and teens in the care of child welfare agencies were living in unlicensed settings — up from 124 just two years earlier.²² One ten-year-old was reported sleeping in a children’s aid society office. In other regions, teenagers were housed in budget hotels with rotating staff and no therapeutic programming.

“No parent should have to choose between earning a living and caring for their child with special needs. No parent should have to relinquish custody to access essential treatment.”

— Letter to Ontario government on behalf of Northern Ontario families, 2025

In Northern Ontario, local news outlets documented the case of an autistic child separated from their family and placed in a treatment facility hours from home because no local residential services existed.²³ The family described being told that relinquishing custody to a children’s aid society was effectively the only path to funded residential care.

The perverse incentive

Here is the arithmetic that reveals the system’s true priorities. A single child in emergency residential placement at $500,000 per year costs the province as much as providing OAP core clinical services to 75 children at the minimum band, or 15 children at the moderate band, or roughly 8 children at the maximum. Three hundred and thirty-nine children in unlicensed placements, at a conservative average of $300,000 each, represents approximately $100 million per year — enough to fully fund more than 15,000 children through the OAP’s minimum band.

But emergency placements are legally mandated spending. Once a child enters child-welfare care in crisis, the province must pay. There is no waitlist for a child who has already fallen off the cliff — only for the children standing at the edge. This is the perverse logic of the system: it is cheaper, in political terms, to let 71,263 children wait and then pay catastrophic rates for the 339 who crash through the bottom than to fund the preventive services that would keep them from falling.

The FAO has identified the structural underfunding that drives this cycle. Its 2025 spending review found the province’s budget plan is $1.5 billion below what is needed in 2025–26, growing to $3.1 billion by 2027–28, just to maintain current service levels.²⁴ The province is not merely under-investing. It is budgeting for decline — projecting that MCCSS spending will decrease at an average annual rate of −0.5% while the FAO says it needs to grow at 4.2%.²⁴

One independent analysis of MCCSS spending puts roughly 32% of OAP funding in expenditure other than core services.²⁵ The government’s own records reach the same place by a different route: less than half of the $691.2 million spent on the autism program in 2023-24 went to core services — the therapy children actually need.²⁶ The rest was absorbed by administration, legacy programs, capacity-building initiatives, and AccessOAP’s $57.9M intake operation.²⁶

This is the structural narcissism of a system that has learned to monetize its own failures. Every child who waits 5 years and enters crisis is worth more to the system’s operators than a child who receives $6,600 in early intervention and never needs residential care. The revenue model does not depend on the 71,263. It depends on the queue continuing to exist.

VI. Narcissism Is a Measurable Variable

“Narcissism” is used loosely enough in political writing that it is worth being precise about what the research actually shows — because it is not a metaphor, and the findings are more damning than the metaphor would be.

Management scholars solved the obvious problem — you cannot administer a personality inventory to a sitting chief executive — by building unobtrusive indicators validated against clinical scales: the prominence of an executive’s photograph in the annual report, their pay relative to the second-highest-paid officer, first-person pronoun use, and the physical size of their signature.²⁷²⁸ These measures predict overinvestment, weaker returns, and a significantly higher probability of corporate fraud.²⁹

Two findings from that literature bear directly on Ontario. The first is that markets already price this trait as a risk: auditors charge measurably higher fees when they detect narcissism in a chief executive.³⁰ The second is the reason such structures persist — narcissistic executives extract higher compensation and larger pay gaps over their own teams.³¹ The trait is rewarded before it is punished, so selection pressure runs toward it.

The political-psychology evidence is stronger still, because power can be manipulated in a laboratory in a way that executive personality cannot. Randomly assigning people to a high-power condition causes measurable reductions in perspective-taking — the capacity to model another person’s viewpoint.³²³³ Across seven studies, higher social class predicted more unethical behaviour, mediated by favourable attitudes toward greed.³⁴ These are causal findings, not correlations. Institutional design sits upstream of individual character.

The structural claim, stated carefully

Here the argument must be made honestly, because the strong version is also the defensible one. Nothing in this record establishes that any particular official is a narcissist, and no study reports a validated prevalence rate for narcissism among executives, physicians, or politicians. Anyone citing such a figure is inventing it.

The finding that survives scrutiny is structural, and it does not depend on anyone’s character. Industry-sponsored medical research is roughly four times more likely to reach conclusions favourable to its sponsor.³⁵ Private-equity acquisition of hospitals is associated with a 25.4% increase in hospital-acquired adverse events.³⁶ Neither study measured a personality. Both measured what incentive architecture does to outcomes at scale — and both found harm.

Applied to Ontario, the three features map without needing to diagnose anyone. Grandiosity: a $965 million budget presented as historic investment, serving fewer than a quarter of registered children. Exploitation: an administrator whose workload rises with the volume of the waitlist it processes rather than with the rate at which the waitlist shrinks — payments tripled while fewer than one in four waitlisted children received core clinical services.³⁷ Impaired empathy: a government that stopped publishing waitlist figures for nearly two years,³⁸ leaving advocates to obtain through freedom-of-information requests data the public is entitled to have.

The suppression of outcome data is the tell. A system confident that its numbers vindicated it would publish them.

VII. The Broader Pattern

Ontario is not an outlier. It is a leading indicator. Insurance mandates create revenue streams; private equity consolidates providers; conflicted researchers supply the evidence base; families absorb the gap between policy rhetoric and service delivery. When that evidence is scrutinised, it does not always hold — an article defending private-equity involvement in autism services was retracted from the International Electronic Journal of Elementary Education in 2023 after its sole author acknowledged concerns about scientific integrity.³⁹

The alternative is already articulated, by autistic adults, parents, independent researchers, and the WHO itself: earlier diagnosis, parent-mediated naturalistic intervention, community-based support, and outcome measures built on quality of life rather than behavioural compliance. It does not require $35 billion in market capitalisation. It requires political will.

VIII. What Would Change Look Like?

If Ontario took its obligations under the WHO framework seriously — the framework Canada upholds — the following reforms would be immediate:

1. Decouple administrative payments from intake volume. Pay for outcomes — children served, developmental milestones met — not for processing applications.
2. Implement the WHO Caregiver Skills Training program at scale across Ontario’s public health units — a free, evidence-based intervention deliverable by non-specialists that can begin during the waitlist period rather than after it.
3. Require disclosure of conflicts of interest in any research used to justify provincial spending on autism interventions.
4. Publish waitlist data monthly, automatically, on a public dashboard — as the province once did before it became inconvenient.
5. Invoke the Canada Health Act. The federal Minister of Health has the statutory authority to withhold transfers to provinces that permit privatization in contravention of the Act’s principles. That authority exists. It has not been used.

The WHO Fact Sheet on Autism Spectrum Disorders states that timely access to early evidence-based psychosocial interventions can improve the ability of autistic children to communicate effectively and interact socially.¹¹ The word “timely” is not decorative. A 5-year wait is not timely. A system in which less than half of program spending reaches core clinical services while 71,263 children wait is not a health system. It is an extraction system wearing the costume of one.

A system in which less than half of program spending reaches core clinical services while 71,263 children wait is not a health system. It is an extraction system wearing the costume of one.

The narcissism of the current system is not a metaphor. It is a structural description of an arrangement in which the institution’s self-image — its budget announcements, its language of investment, its claims of historic spending — has become entirely decoupled from the experience of the people it exists to serve. The children do not appear in the budget lines. They appear on the waitlist. And the waitlist is the product, not the problem. It is the inventory that the system monetizes.

71,263 children. Standing a metre apart, they form a line more than 71.3 kilometres long — Toronto to Hamilton, and past it. 71.3 kilometres of children waiting for a system that has forgotten they are there.