What High-Speed Rail
Really Costs
The federal government is currently inviting public comments on the Alto high-speed rail project, with a deadline of April 24, 2026. Alto’s published capital cost estimate ranges from $60 to $120 billion — already the largest infrastructure project in Canadian history by a wide margin.
Before that consultation closes, communities in the corridor deserve to understand what international experience tells us about how these numbers actually evolve — and where the real costs tend to hide.
Based on a database of more than 16,000 megaprojects, nine out of ten run over budget. Rail projects average a 44.7% cost overrun. EU high-speed rail specifically averaged a 78% overrun across the lines audited in 2018. Applied to Alto’s $60–90 billion estimate, historical averages suggest a realistic final cost of $87–130 billion or more — before a single track has been laid.
This is not pessimism. It is the consistent, documented, peer-reviewed pattern from comparable projects on every continent where high-speed rail has been built.
Nine out of ten megaprojects run over budget — and rail is among the worst
Oxford professor Bent Flyvbjerg has spent decades studying megaproject performance. His database — the world’s largest, covering more than 16,000 projects across 136 countries — reveals a stark finding: nine out of ten megaprojects run over budget. Rail projects are among the worst offenders, with an average cost overrun of 44.7% and ridership shortfalls averaging 51.4%. These patterns hold across countries, decades, and political systems. 1 2
The European Court of Auditors reached similar conclusions in a landmark 2018 audit of EU-funded high-speed rail. Aggregate cost overruns across the lines examined were €25.1 billion — a 78% overrun at the line level. Construction delays of more than a decade affected half the lines studied. Four of the ten lines cost more than €100 million per minute of travel time saved. 3
The Iron Law of Megaprojects
Professor Flyvbjerg calls it the “Iron Law”: megaprojects are delivered over budget, over time, and under benefits, over and over again. Fewer than 1% of megaprojects are completed on time, on budget, and deliver the promised benefits. Project managers competing for funding tend to present costs at the floor of what is plausible — a dynamic Flyvbjerg calls “strategic misrepresentation.” 1 4
Every major HSR project has exceeded its initial estimate
The table below compares initial budget estimates with actual or current projected costs for major high-speed rail projects around the world. Every single project on this list exceeded its original estimate — most by enormous margins.
| Project | Length | Initial Estimate | Latest / Final Cost | Overrun |
|---|---|---|---|---|
| California HSR San Francisco–Los Angeles |
~800 km | US$33B (2008) | US$106–128B (2024) | +220–290% |
| UK HS2 London–Birmingham, Phase 1 only |
~225 km | £37.5B (2009, whole network) | £81–100B+ (2025, Phase 1 only) | +134–170% |
| Stuttgart–Munich HSR Germany |
~156 km | €2.6B (initial) | €14.8B (2022) | +469% |
| Stuttgart 21 Germany, station + tunnel |
— | €2.5B (1995) | €8B+ (2025) | +220% |
| Japan Shinkansen Tokyo–Osaka |
515 km | ¥200B (1958) | ¥380B (1964) | +90% |
| Channel Tunnel UK–France |
50 km | £4.65B (1985) | £9.5B (1994) | +80% |
| EU HSR average 2018 Court of Auditors audit |
Various | €32.1B (combined) | €57.2B (combined) | +78% |
| Jakarta–Bandung HSR Indonesia |
~140 km | US$5.5B (2016) | US$7.3B (2023) | +33% |
| Canada — Alto HSR Toronto–Québec City |
~1,000 km | C$6–12B (2021 HFR) C$60–90B (2024 HSR) |
C$80–120B (2025, various) | ? |
Overrun percentages are approximate, calculated in real terms where possible. 1 3 5 6 7 8
Canadian precedent: Trans Mountain and Ontario Line
Canada’s most recent comparable experience is the Trans Mountain Pipeline Expansion, estimated at $5.4 billion in 2013 and delivered at $34 billion in 2024 — a 530% overrun. Ontario’s Metrolinx has seen the Ontario Line nearly double from its original $10.9 billion estimate.
If Alto follows the historical pattern for rail megaprojects — an average 45% overrun — the $60–90 billion estimate would land at $87–130 billion. At the California or HS2 rate, the figure could exceed $150 billion. 10 14
Already Canada’s most expensive project — before overruns begin
Alto’s current public estimate of $60–90 billion (2024 dollars) already makes it the most expensive infrastructure project in Canadian history by a wide margin.
The original concept cost far less — and the difference has never been explained
Transport Action Canada has pointed out that the original “high-frequency rail” concept studied in 2016 — 170 km/h trains on dedicated tracks — was estimated at less than $5 billion (under $10 billion in today’s dollars). The escalation to $60–120 billion reflects a fundamental change in project scope that has never been subjected to a publicly released cost-benefit comparison. 16 17
The per-kilometre cost is out of step with global norms
Jerome Gessaroli, a senior fellow at the Macdonald-Laurier Institute, calculated that Alto implies capital costs of $250–375 million per minute of travel time saved. The EU average — itself considered excessive by the European Court of Auditors — is roughly $146 million per minute saved. Alto’s per-kilometre cost also substantially exceeds the European average of roughly $40 million per kilometre. 15
Five specific, auditable requests for disclosure and review
The following requests are submitted to ALTO and Transport Canada as part of the public consultation record. Each is a specific, auditable action required before any final route selection or construction decision on the proposed corridor.
Apply reference-class forecasting to ALTO’s cost and ridership projections
ALTO must publish a reference-class forecast of total project cost and ridership, applying Flyvbjerg’s methodology (44.7% average rail overrun; 51.4% average ridership shortfall) to its current estimates. This is the standard required for major infrastructure projects in the UK under HM Treasury Green Book guidance. A promoter-class estimate alone is insufficient for public accountability.
Break out grade separation costs as a separate line item
ALTO must publish grade separation costs as a separate and explicit line item in its capital cost estimate, including: the number of crossings proposed on each corridor option; the crossing treatment hierarchy (which roads receive structures vs. which are closed); and the cost per structure by type. This $3–8 billion cost category has never been publicly disclosed.
Publish the cost of environmental mitigation for the Frontenac Arch and Napanee Plain
ALTO must publish a preliminary estimate of environmental mitigation costs for proposed corridors through the Frontenac Arch UNESCO Biosphere Reserve, Napanee Limestone Plain karst aquifer, Rideau Canal World Heritage Site, and Eastern Ontario wetland complexes. These costs are absent from all published materials.
Commission an independent value-for-money review
An independent value-for-money review of the ALTO project, applying Flyvbjerg’s reference-class methodology, must be commissioned and publicly released before any route selection is finalised. The review must compare the realistic probability-weighted cost range against the realistic probability-weighted ridership and revenue range.
Publish the cost-benefit comparison between ALTO and the original HFR concept
Transport Action Canada has noted that the escalation from under $5 billion (HFR, 2016) to $60–120 billion (HSR, 2024) reflects a fundamental change in scope. ALTO must publish a formal cost-benefit comparison between the full HSR specification and the original HFR specification, showing what additional transportation benefit is produced per additional dollar of capital cost.
What communities along the corridor deserve to know
“Communities being asked to accept a rail line through their region deserve to know not just the stated budget, but the realistic all-in cost based on how every comparable project in history has actually unfolded.”