Canada Wants High-Speed Rail. Megaproject Reality Wants a Word.

Support CleanTechnica’s work through a Substack subscription or on Stripe.

Canada’s revived high-speed rail (HSR) proposal deserves a serious hearing. It also deserves an outside-view assessment grounded in the history of rail megaprojects, not just the aspirations of sponsors and advocates. The current Alto proposal for a high-speed line between Toronto and Québec City is the most concrete Canadian version in years, with a defined station set, a selected development partner, and an explicit early capital estimate of C$60 billion to C$90 billion in 2024 dollars. But it remains an early-stage project with an unfinished alignment, unresolved city approaches, and a first construction start that Alto itself places around 2029–2030 for the initial Ottawa–Montréal segment.

That distinction matters because Canada has been here before, just in different forms. For decades, the country has revisited variants of faster rail in the Québec City–Toronto or broader Québec City–Windsor corridor, moving from true high-speed ambitions to improved conventional rail and now back again. VIA’s High Frequency Rail proposal, submitted in 2016, was the immediate ancestor of the current effort, built around dedicated passenger infrastructure and much better reliability than today’s freight-constrained corridor. The 2025 decision was not the invention of a brand-new concept so much as the re-escalation of an existing corridor plan from higher-frequency conventional rail to genuine high-speed rail.

There is a real public-interest case for the project. The Toronto–Ottawa–Montréal–Québec City corridor contains a large share of Canada’s population, politically important urban centers, and a substantial portion of national economic activity. Existing intercity passenger rail performs badly by the standards of modern advanced economies. VIA Rail’s on-time performance has been poor, and the basic reason is well understood: passenger trains operate largely on infrastructure they do not control, subject to freight interference and network congestion. A mostly dedicated electrified passenger railway would be cleaner, more reliable, and more useful than the system Canada has today.

That improvement is the strongest argument for Alto. If the line is built as proposed, it would be a much better rail product than the current corridor service, and likely a much better product than the earlier high-frequency rail (HFR) concept as publicly understood. Alto is being framed around speeds up to about 300 km/h, with a seven-city network including Toronto, Peterborough, Ottawa, Montréal, Laval, Trois-Rivières, and Québec City. It is also being advanced in phases, with Ottawa–Montréal chosen first because Alto describes it as the shortest and technically simplest segment. From a service perspective, that is a meaningful step up from the long Canadian tradition of discussing better rail while remaining trapped in freight-owned rights-of-way and incremental upgrades.

One of the simplest reasons for caution is corridor density. In an earlier comparison I did with a roughly similar high-speed rail corridor in China, the Chinese route had about double the population potentially served. That does not make the Canadian project irrational, but it does change the economics and ridership expectations. High-speed rail works best where large numbers of people live, work, and travel along the spine of the line and can reach stations easily. A corridor with only half the population density has less natural demand to draw on, which means every question of station siting, feeder transit, frequency, fare levels, and competition from cars and short-haul flights matters more, not less.

One of Alto’s other risks is that Canada is not starting from the same behavioral baseline as the world’s strongest rail markets. Canadians live in a car-oriented country, fly heavily on longer domestic trips, and use intercity passenger rail only modestly outside the Québec City–Windsor Corridor. That makes automatic ridership assumptions dangerous. But the global evidence also suggests that culture is not destiny. Where high-speed rail is fast, reliable, central, frequent, and well connected, travelers shift. The real Canadian question is not whether Canadians are genetically incapable of becoming rail users. It is whether Alto will be good enough, door to door, to break entrenched habits of driving and flying.

The stronger concern about Alto is not that seven station cities across roughly 1,000 kilometers is, by itself, unusually stop-heavy. On a global reference-class basis, it is not. Japan’s Tokaido Shinkansen covers a much shorter corridor with far more stations, and Taiwan High Speed Rail also serves a dense chain of intermediate stations. The leading practice internationally is not to minimize stations at all costs, but to protect end-to-end competitiveness with layered operations: fast express services for the biggest city pairs, semi-fast services for secondary markets, and local stopping patterns where justified. JR Central’s Nozomi, Hikari, and Kodama service tiers are the classic example. That means Peterborough and Trois-Rivières are not automatically evidence of bad design, although they have small populations. They become a problem if Canada bakes them into too many mandatory stops, without the express operating pattern and station accessibility needed to keep high-speed rail from turning into fast regional rail.

The International Union of Railways (UIC) identifies access to city centres as one of HSR’s core structural advantages, while international station-planning work for California found that non-central HSR stations only perform well when they are designed from the outset as strongly multimodal nodes with credible local and regional transit links. The operational side matters just as much. Successful systems do not simply count stations and hope for the best. They preserve end-to-end competitiveness with layered service patterns, as on Japan’s Tokaido Shinkansen, where Nozomi, Hikari, and Kodama trains serve different market needs without forcing every train into an all-stops pattern. That is the benchmark Alto should be judged against: central or very well connected stations, strong feeder networks, and an operating plan that protects journey times for the biggest city pairs.

Against that reference class, Toronto and Montréal clearly look like natural HSR anchors, with very large transit systems and strong regional feeder networks already in place, and Laval is viable because it sits inside the wider Montréal transit ecosystem. Ottawa and Québec City look workable but less powerful, with meaningful urban transit but thinner regional feeder depth. The already somewhat problematic Peterborough and especially Trois-Rivières are where the concern becomes much sharper. Alto’s own consultation material says Peterborough’s station could be near major roads with bus connections and that scenarios north of Trois-Rivières are being studied because of downtown constraints, which is exactly the sort of edge-of-city compromise that weakens HSR’s door-to-door advantage. Alto does acknowledge the issue in principle, repeatedly describing stations as mobility hubs that must connect well to other modes, but that acknowledgment is still much stronger as a design aspiration than as a published feeder strategy. If the project proceeds, fixing that should be a first-rank priority. Otherwise Canada risks building a line that is fast between cities on paper, but less competitive in real trips because the weaker intermediate stops lack the kind of integrated catchment that successful global HSR systems rely on.

One of the emerging political liabilities of Alto is that it increasingly looks less like a neutral national corridor optimized from scratch and more like an eastern megaproject with a distinctly Québec center of gravity. The corporate and institutional backing is telling. The winning Cadence consortium is led by CDPQ Infra, part of Québec’s pension-fund ecosystem, includes Montréal-headquartered AtkinsRéalis, and relies heavily on French rail expertise through SNCF Voyageurs, Keolis, and SYSTRA. The official station map reinforces the same impression. The live project runs from Toronto to Québec City with stops in Toronto, Peterborough, Ottawa, Montréal, Laval, Trois-Rivières, and Québec City, giving Québec multiple beneficiary stations beyond Montréal itself, including one in Trois-Rivières, a metropolitan area of only about 161,000 people. That does not prove some sort of insider capture. It does suggest that the project’s institutional DNA and beneficiary geography are much more Québec-weighted than its national branding implies.

That matters because the political burden is landing unevenly. Ontario has formally partnered with Ottawa on Alto, but the provincial reaction has been much more conditional than the national rhetoric suggests. Doug Ford has pushed publicly for a Highway 401 alignment and a Kingston stop, while the Eastern Ontario Wardens’ Caucus and several municipalities have opposed the project in its current form, arguing that too much of rural Ontario absorbs corridor impacts without enough direct benefit. In practice, support is strongest where the train clearly stops, such as Peterborough, and opposition is strongest where the line cuts through land and communities without a station. Framed that way, Alto’s Québec-heavy character is not just a matter of optics. It is becoming a real coalition problem, because a project with multiple visible Québec wins and a thinner Ontario beneficiary map is naturally harder to sell as a balanced nation-building investment.

The honest question is not whether the end-state service would be better. Of course it would. The question is whether the new elements in the current proposal materially improve the probability of successful delivery, or whether they mainly raise both the upside and the difficulty. On that front, the answer is more sobering. Some of the most helpful elements were already embedded in HFR, including the focus on dedicated passenger infrastructure, better reliability, and a modern project-development structure. What is actually new is mostly the jump from higher-quality conventional rail to true high-speed rail. That improves the service proposition. It also moves the project into a more difficult class of megaproject, requiring straighter alignment, more grade separation, tighter safety requirements, more land, more civil works, and a much larger capital envelope.

This is where the outside view matters. The inside view looks at a specific project and asks what the sponsors intend to do. The outside view looks at the reference class of comparable projects and asks what projects like this usually do. For rail, and especially for large new-build or quasi-greenfield passenger rail in developed democracies, the answer is not encouraging. The UK Department for Transport and Oxford Global Projects reference class work on international rail projects shows cost uplifts at early business-case stages that are substantial even at the 50th percentile, with much larger uplifts for more prudent confidence levels. Those same studies find schedule overruns that are not outliers but typical features of rail delivery. Flyvbjerg’s broader rail megaproject work also found average cost overruns around 45%, while the European Court of Auditors found audited European high-speed lines suffering major cost escalation and, in many cases, delays extending for years or more than a decade.

California’s high-speed rail experience is a useful warning for Canada because it shows how a compelling rail vision can be eroded by the normal pathologies of megaproject delivery. What voters were originally sold was a San Francisco to Los Angeles/Anaheim system for about $33 billion and opening around 2020, but official and federal documents now describe a much more expensive and much later project, with the California authority’s current full Phase 1 estimate at roughly $89 billion to $128 billion and the first operating segment reduced for the foreseeable future to the Central Valley between Merced and Bakersfield. That means the initial service is not reaching the downtown city pairs that gave the project its political and economic force in the first place, but instead terminating well short of the major urban centers it was meant to connect, illustrating how cost escalation, delay, and phased delivery can leave a high-speed rail project technically alive while substantially weakening its real-world value proposition.

Applied to Alto, the outside-view implication is a bit more encouraging than the usual megaproject story. The public C$60 billion to C$90 billion figure is still preliminary and based on high-level assumptions, and the route, land requirements, city approaches, and later construction phases remain unresolved. Normally that combination would be a classic recipe for optimism bias. But Alto’s own estimate is already for a network of up to 1,000 km, which means a baseline of roughly C$60 million to C$90 million per route-km before later refinement. On an international reference-class basis, that is not a bargain-basement number. It is already at the high end of typical high-speed rail costs and closer to the expensive democratic megaproject class than to the lower-cost French or Spanish build examples. In other words, while outside-view uplift still matters and later cost growth remains a real risk, the current range looks less like a politically convenient lowball designed to win approval and more like an early estimate that at least recognizes the project is structurally expensive from the start. That is actually a constructive sign, because it suggests the sponsors may be entering the process with a more realistic view of the capital intensity than is common in the first public round of major rail proposals.

Schedule deserves equal weight. Alto’s own materials say construction on the first Ottawa–Montréal segment is expected to begin around 2029–2030, and that once underway, the first segment would take about seven to eight years to build. That points to an inside-view opening in the 2036–2038 range for that segment. Rail reference-class evidence suggests a later reality is more likely, with a first-segment opening around 2039–2041 as a central expectation and a mid-2040s downside that would still be normal rather than extraordinary by megaproject standards. For the full Toronto–Québec City network, the outside view points to the late 2040s or early 2050s as a more realistic frame than anything closer to the early 2040s.

That timing is not a side issue. It changes the market Alto would enter. One of the biggest analytical gaps in many high-speed rail discussions is that they frame competition against today’s transport system, not the system that will exist when the railway opens. For Canada, that matters a great deal. The revived rail proposal is often discussed as though its core competition will be current car travel and current short-haul aviation. But a railway opening in the late 2030s or 2040s is much more likely to face a transport market shaped by advanced driver assistance moving toward highway autonomy, and by short-haul aviation that is increasingly hybrid-electric or otherwise partially electrified.

That future competition is especially important in Canada because intercity travel is not purely station-to-station. It is door-to-door. The line-haul part of a Toronto–Montréal or Ottawa–Montréal trip matters, but so do first-mile and last-mile friction, baggage, family travel, local access, parking, and dispersed origins and destinations. In a corridor with weaker feeder rail than European or Japanese analogues and a stronger culture of private vehicle use, autonomous or semi-autonomous EVs are a more serious long-run competitor than many rail advocates admit. If families can leave in the evening, sleep through a highway trip, and arrive near their destination without transfers, that does not eliminate the market for rail, but it does reduce the assumption that rail automatically dominates once it reaches high speed.

Short-haul aviation also cannot be treated as a static incumbent. The case for rail against fossil jet operations on short sectors is stronger than the case against hybrid-electric or increasingly electrified regional aircraft using existing airport infrastructure. Toronto–Montréal is not just a rail corridor. It is also an air corridor, and aviation technology is not standing still while Canada debates rail. That does not make Alto irrational. It means the honest case for the line should be framed as public-interest infrastructure in a changing transport system, not as an obviously superior commercial product immune to modal competition.

This distinction between public-interest logic and commercial logic is the most important one in the debate. There is a strong argument that Canada should build better intercity passenger rail because the country needs more resilient, lower-carbon, higher-quality transportation options in its densest urban corridor. There is also a nation-building argument. Countries that cannot plan and deliver strategic mobility infrastructure lose optionality over time and remain dependent on inherited systems that are increasingly ill-suited to climate and economic goals. But that is not the same as saying the project will be a slam-dunk market winner on its own narrow financial terms. The outside view does not support that optimism.

Governance improvements and the project-development model help at the margin. A more sophisticated co-development structure with a selected private partner can improve incentives, stage decisions more effectively, and reduce some interface risks. Transport Canada’s own audit work suggests governance and procurement controls have matured. That is good. But governance does not repeal geometry, land acquisition, grade separation, municipal politics, Indigenous consultation, utility relocation, inflation, or the normal laws of rail megaproject delivery. Good governance can make a hard project somewhat more manageable. It cannot make it simple.

So where does that leave the project? Not in the category of foolish, and not in the category of inevitable success. Alto looks like a proposal whose benefits are real, whose public-interest rationale is respectable, and whose service proposition is better than what Canada has today or what HFR alone likely would have delivered. The more nuanced outside-view read is that the current public cost framing may be less unrealistically optimistic than is common for rail megaprojects at this stage, which is actually a constructive sign. The schedule risk still appears substantial, and the market it eventually enters is likely to be more competitive than many supporters imply. But on cost, the emerging picture is not simply of a project being lowballed to gain approval. It is of a project whose sponsors may already understand that this is an expensive, high-interface piece of infrastructure, even if later refinement and upward pressure remain likely.

That means the right way to discuss Canada’s high-speed rail revival is neither with boosterism nor cynicism. It is to say that Canada may well want this project, but should want it with clear eyes. If the country chooses to build it, it should do so on the basis that the first truly useful segment is more likely around 2040 than the mid-2030s, and that the full network is likely a late-2040s or early-2050s proposition. It should do it while simultaneously improving regional transit around the stations. It should do it while working to move the needle on Canadian assumptions of driving cars and flying as primary modes of transportation. It should also do so knowing that rail will be competing against not only roads and planes, but smarter electric cars and cleaner short-haul aviation.

Ambition is not the problem. Canada should be ambitious about intercity mobility, and Alto at least appears to be entering the discussion with a more realistic sense of capital intensity than many megaprojects do in their first public phase. But ambition still does not repeal the reference class. The revived high-speed rail proposal should be judged as a serious public-interest investment in a difficult megaproject category, with real benefits, real uncertainty, and a market context that will keep evolving while it is built. The honest case for it is not that the normal rules no longer apply. It is that Canada may decide the benefits are worth pursuing even knowing that they do.