Structure, Not Capital
Canada’s GDP per capita fell two percent over the five years to 2024, the worst sustained decline since the Great Depression. The 2025 per-capita number was essentially flat after two consecutive years of decline. The OECD’s long-run outlook projects Canada will rank last among all thirty-eight member countries in real per-capita GDP growth through 2060.
The country’s response is to build its way out. Ottawa has announced a Major Projects Office to accelerate sanctioning of nationally significant infrastructure, an expanded mandate for the Canada Infrastructure Bank, and a regulatory framework reorganised under the One Canadian Economy initiative to move projects faster. The Spring Economic Update 2026 leads with a “second-fastest G7 growth” headline and an infrastructure pipeline measured in hundreds of billions of dollars.
Speed has not been the problem. The procurement that produced the failed PPPs of the last twenty-five years is still in place, along with the demand modelling that supported them, the capital matching that funnelled short-horizon private equity into long-ramp assets, and the missing phase-gate accountability that other mature project industries treat as basic. Adding capital to an unchanged structure produces the structure’s failures at scale.
The demographic foundation
Every merchant risk infrastructure project is underwritten on demand assumptions. The assumptions depend on population growth, working-age cohort size, household income trajectory, and willingness to pay user fees. The reality of the country those assumptions are supposed to track has shifted in ways most of the public-facing forecasting has not accounted for.
Canada’s total fertility rate sits at approximately 1.26, well below the 2.1 replacement level and at a historical low. The 65-and-over share of the population now exceeds the under-15 share and the gap is widening. The federal government has revised immigration targets sharply lower for 2025-27 from earlier trajectories, removing one of the principal supports under aggregate GDP growth and exposing the per-capita reality more directly. Productivity growth has been weak for over a decade. Household financial stress is elevated, with shelter costs consuming more than half of median household income in major urban markets and youth unemployment running above the post-2016 average.
These conditions matter for infrastructure because the demand cases that supported Canadian toll roads sanctioned in the 1990s do not match the country of 2026, and the demand cases being constructed in 2026 will not match the country of 2040. A forecast that assumes 2008-era population growth, 2008-era working-age employment, and 2008-era willingness to pay user fees, denominated in 2026 dollars, is a marketing document with footnotes.
The infrastructure community knows this. The procurement frameworks have not yet been required to.
What the Canadian PPP record actually shows
The 407 ETR is the standing example of how to give away the value of a public asset for a one-time fiscal payment. The Mike Harris government of Ontario privatised the toll road in 1999 for approximately $3.1 billion under a 99-year concession. The concession holders have generated multiples of that figure in dividends over the following twenty-five years, with the asset itself valued at substantially more than the original sale price. The Auditor General of Ontario subsequently found that the concession structure limited the province’s ability to regulate tolls, which now sit among the highest in North America. The structure transferred long-run upside to private holders for a payment that has been retrospectively understood as a significant discount to fair value. The mechanism that produced this was a procurement structure that prioritised closing the deal over pricing the asset.
The Champlain Bridge replacement was originally structured as a PPP and was progressively pulled back into direct federal procurement after concerns about lifecycle cost, tolling, and risk allocation. The completed bridge opened in 2019 under federal ownership with no toll. The decision to abandon the PPP structure mid-process was the correct one. It would have been a less costly correction to make before the procurement started rather than during it.
The Ottawa Confederation Line LRT was procured as a PPP through a consortium led by SNC-Lavalin. The system opened in 2019 and experienced sustained operational and reliability failures, including a major derailment in September 2021. The public inquiry led by Justice William Hourigan concluded that procurement and governance failures had compromised both delivery and operational integrity. The City of Ottawa has continued to absorb operating and capital costs the original PPP was structured to insulate it from. The asset is now in active service with a maintenance and reliability profile the original commercial case did not anticipate.
The Réseau express métropolitain in Montreal, structured by CDPQ Infra as a hybrid public-private project with Quebec’s pension manager as the principal sponsor, has been delivered substantially over its original budget and behind its original schedule. The hybrid model produced both Quebec pension exposure and a project subject to ongoing criticism over routing, station design, and integration with the existing transit network. The structure shielded the political principals from direct accountability for the overruns by placing the financial risk inside a pension institution whose oversight model is not designed for project delivery accountability.
The Confederation Bridge has performed competently for nearly thirty years on an availability-payment structure that did not depend on demand forecasts. The Anthony Henday Drive components of Edmonton’s ring road were delivered through PPPs that the Alberta auditors have generally treated as successful. The Vancouver Canada Line opened ahead of schedule in 2009 and has operated reliably since. BC Hydro’s John Hart Generating Station replacement was delivered on schedule and within budget.
PPPs in Canada have worked when the underlying asset is technically straightforward, the demand profile is predictable, the structure is availability-payment rather than demand-risk, and the public counterparty has retained the regulatory levers it needs to govern operation. PPPs have failed when demand risk has been transferred to capital that cannot bear it, when the asset is operationally complex, when the political incentive favoured signing over getting it right, or when the regulatory framework has not kept pace with the structure of the deal.
The acceleration agenda does not address any of the variables in the second list.
Why more money makes it worse
Capital chasing badly structured projects produces more failures, not fewer.
The Canada Infrastructure Bank, established in 2017 with a $35 billion mandate, has had a deployment record variously described as cautious, slow, or strategic depending on the political vantage of the commentator. Whatever the framing, the CIB’s evolution has demonstrated that a capital institution operating inside a flawed structural framework is forced to participate in the framework’s failures. The CIB cannot, on its own, fix the procurement practices of provincial transit agencies, the demand modelling practices of the major consultancies, or the capital matching practices of the sponsor community. It can refuse to participate, which it has occasionally done, or it can participate on its own terms and accept the resulting risk profile.
The federal Major Projects Office is intended to accelerate the sanctioning and delivery of nationally significant infrastructure. The acceleration model assumes the binding constraint is regulatory friction. The actual binding constraint in most failed Canadian infrastructure projects of the last two decades has been structural rather than regulatory. Demand has been allocated to capital that could not bear it, capital horizons have not matched asset ramp profiles, operational accountability has been displaced from sponsor to operator, and procurement has rewarded financial close over delivered performance. None of these are frictions acceleration would fix. Acceleration delivers more of each, faster.
A federal infrastructure push that delivers $200 billion of capital through the existing project pipeline, the existing procurement frameworks, the existing demand modelling capacity, and the existing capital matching practices, will produce a higher absolute volume of failed projects at the existing failure rate. The taxpayer will pay for the failures the same way the taxpayer has paid for Ottawa LRT, the equity holders of Cross City Tunnel in Sydney, the lenders of the Indiana Toll Road, and the public balance sheets that absorbed Carillion’s UK PFI portfolio after its collapse. The cost will be socialised because the assets will be essential. The political incentive will favour quiet restructurings over public failures.
On a successful PPP the upside accrues to private holders. The downside is absorbed by the public when the asset is essential, which it usually is. The risk transfer that justified the structure on day one turns out to have been imaginary by the time it matters. Adding capital makes the bill higher.
What structural reform looks like
The discipline that would fix this is regulatory, unglamorous, and unpopular with everyone whose fees crystallise at financial close.
Consider what a merchant risk transport project would look like under gates that worked. Take a tolled bridge, an inland port, or a regional rail link. The procurement sequence might run as follows.
The development phase would close on a forecast that had reconciled, in writing, to Statistics Canada’s regional population projection, the Parliamentary Budget Officer’s productivity baseline, and the relevant provincial economic outlook. Variances between the consultancy forecast and the agreed federal-provincial baseline would be a procurement disclosure item. A forecast that depended on demand the demographic numbers could not support would not progress to underwriting.
The underwriting phase would close on a capital structure where lead equity matched the asset’s ramp profile. A project with a seven-to-ten-year ramp would not take short-horizon private equity as its lead equity, and the CIB and Major Projects Office sanction criteria would say so explicitly. Where demand risk could not be absorbed by the equity stack, the public contingent liability would be priced and disclosed in the procurement documents on day one, capped, with the workout playbook agreed before the asset opened.
The construction phase would close with an operations director named, in the room, and accountable for the case being handed over. Critical development personnel would carry compensation that did not vest until the asset reached steady state. The receiving team would have a contestation right at the gate, with real consequences for the handing-off team if the deliverable did not meet the acceptance criteria.
The operations phase would open with twelve months of customer pre-validation behind it: letters of intent from anchor users with volumes and indicative pricing, a documented pipeline of prospective customers, and independent industry surveys with sample sizes large enough to mean anything. The first board meeting of the operating company would receive a paper comparing actual ramp against the conservative case underwritten for debt service, with the marketing case shown separately. A covenant holiday would sit over the ramp years.
None of this is exotic. The aerospace primes have done versions of it for decades, the oil and gas majors run major projects under Front-End Loading frameworks, and the better regulated utilities use stage gates as standard. Canadian transport procurement does not.
The reforms do not produce ribbon-cuttings, and ministers who require them rather than sanctioning projects on quarterly timelines tend not to last in office. Each one, if adopted, would prevent more value destruction than the entire Major Projects Office is likely to add through acceleration.
The point about who pays
Canada has access to the capital, the political mandate to spend it, and the institutional capacity to deploy it. What is missing, in the part of the system that determines whether the spend produces assets that operate as promised, is the structural discipline.
Adding capital to the system without adding structure produces what the system has been producing. A pipeline of essential assets the public will eventually have to absorb, at the same time as the per-capita income that should have supported them is declining, and at the same time as the demographic base that should have driven their demand is shrinking.
The frame, of aggregate growth and infrastructure volume rather than per-capita performance and structural reform, is the same metric-capture move that produces the failed merchant risk projects this argument started with. Choose the number that supports the political case, defer the structural reckoning, and push the consequence forward to people who will not be in the room when it arrives.
The country deserves a debate about structure rather than a press release about capital.
Reading list
Ludovic Phalippou, “An Inconvenient Fact: Private Equity Returns and the Billionaire Factory” (Oxford Saïd Business School, 2020). The cleanest demolition of the PE outperformance claim. Fees as the actual product.
Eileen Appelbaum and Rosemary Batt, Private Equity at Work: When Wall Street Manages Main Street (Russell Sage Foundation, 2014). The standard sector-by-sector account of what happens to companies inside a PE hold period. Strong on employment and supplier effects.
Brendan Ballou, Plunder: Private Equity’s Plan to Pillage America (PublicAffairs, 2023). Written by a former DOJ counsel. Recent, accessible, anchored on the consumer harm transmission.
Eduardo Engel, Ronald Fischer and Alexander Galetovic, The Economics of Public-Private Partnerships: A Basic Guide (Cambridge University Press, 2014). The standard academic treatment of PPP economics. Strongest on risk allocation theory.
Aidan Vining and Anthony Boardman, various articles on Canadian PPP performance (Journal of Comparative Policy Analysis, Canadian Public Administration, 2008 onward). The Canadian empirical literature on what worked and what didn’t, by the two scholars who have spent the most time on it.
Justice William Hourigan, Report of the Ottawa Light Rail Transit Public Inquiry (Government of Ontario, 2022). The case study. Demonstrates concretely how procurement structure compromises operational delivery.
Auditor General of Ontario, reports on Highway 407 ETR (2003, 2009, and subsequent). The original sin of Canadian PPP privatisation, documented contemporaneously by the relevant officer.
National Audit Office (UK), Investigation into the government’s handling of the collapse of Carillion (HC 1002, 2018). The clearest official account of what happens when essential infrastructure delivery passes through a private balance sheet that fails.
Lord Justice Sheen, MV Herald of Free Enterprise: Report of Court No. 8074 (Department of Transport, UK, 1987). The canonical project-structure-as-safety-failure document. Required reading for anyone thinking seriously about commercial pressure in regulated transport.
James Reason, Managing the Risks of Organisational Accidents (Ashgate, 1997). The framework most modern safety thinking still relies on. The Swiss cheese model is Reason’s.
Charles Lammam et al., Standard of Living in Canada: Trends and Drivers (Fraser Institute, 2025). The per-capita GDP work that has set the public debate frame in Canada.
OECD, Economic Survey of Canada 2025. The international comparative perspective on Canada’s productivity and per-capita performance. The long-run projection that has Canada last in the OECD through 2060 comes from the same workstream.ream.