Alto High-Speed Rail: Expropriation Already Law, $60–90 Billion That Isn't a Budget, No Cost-Benefit Analysis

The figure that isn't a budget

Start with the number everyone is arguing about, because the most useful thing about it comes from the people building the project. Transport Canada calls $60 billion to $90 billion an "early capital costs estimate," stated in 2024 dollars. [1] Alto, the Crown corporation subsidiary developing the line, is more explicit about what that means. The estimate, it says, rests on "the scale of the project and available research, not on final designs or construction contracts," and is "for planning purposes only and should not be considered as a project budget." [2]

That is not a critic talking. It is the proponent, in writing, telling readers that the headline figure is a planning range, not a costed commitment. Alto's chief executive put it more plainly in committee testimony: to have a budget, he said, you need engineering; to have engineering, you need an alignment; and the project is still at the stage of a corridor. [7] The distinction matters because the public debate treats the number as if it were firm enough to defend or attack to the billion, and it was never built to carry that weight.

There is one boundary around the number that is firm. The $3.9 billion announced for the project's first phase covers design and co-development, not construction. [3] It sits outside the $60–90 billion estimate. That clean separation is worth holding onto, because it establishes the principle that runs through everything else: the headline figure does not contain everything the project will cost. The open question is how much it leaves out.

What the number cannot contain

An estimate made before a route exists cannot price the things a route determines. That is not a criticism of Alto's arithmetic; it is a limit on what any figure can mean at this stage. The corridor under study is about 10 kilometres wide, and no alignment has been chosen. [5] Until it is, several of the largest cost drivers in any high-speed rail project remain open.

Rolling stock, the trains themselves, is not itemized in any public costing document. Stations are described as part of "the network," but no station-by-station figure has been published, and Alto's chief executive has described the downtown Toronto leg as the costlier final segment. [7] The approaches into Toronto and Montréal, where land is dense and tunnelling or elevated track is most likely, cannot be priced before they are designed. Land acquisition is described by the CEO in interviews ("we will need to buy significant pieces of land, so we're talking about thousands of properties"), but no costing document confirms the $60–90 billion figure encompasses it, and no contingency percentage or financing cost has been published. [7]

What unites these gaps is their direction. None of them would push the estimate down. Every category that is undetermined, whether trains, urban stations, tunnelled approaches, land, contingency, or financing, can only add. The international record on comparable lines bears this out: dedicated high-speed track has been built for well under C$45 million per kilometre in Spain and on France's recent greenfield lines, but dense urban approaches and tunnelling are what carried Britain's HS2 past C$700 million per kilometre. Which of those worlds Alto lands in depends almost entirely on the decisions that have not been made. So when the estimate moves, the question worth asking is which direction the unpriced items can take it.

The powers came first

While the route and its costs remain open, one piece of the project is settled and difficult to reverse. The High-Speed Rail Network Act became law on March 26, 2026, enacted not as a standalone bill but as one division of Bill C-15, the 603-page Budget Implementation Act that amended 49 separate statutes. [4] The Act deems the rail lines to have been approved under section 98 of the Canada Transportation Act and rewrites how land along the eventual route can be taken. Under the general Expropriation Act, the Crown must first attempt to negotiate a purchase; the new framework creates an exception for high-speed rail land, allowing the government to proceed directly to expropriation and replacing the Act's standard notice obligations with a simplified process. [8]

The scale of what those powers reach is not small. Alto's CEO has said the Montréal–Ottawa segment alone is expected to cross roughly 1,700 properties, including at least 500 agricultural parcels, with the acquisition process likely beginning in 2027. [7] The legal machinery to take that land now exists. The map of which land will be taken does not. The enabling power has arrived ahead of the thing it enables, the sequence the bridge pointed to, now on the statute books.

Enabling legislation that precedes a final alignment is, on its own, ordinary practice for large infrastructure. What is worth noticing is the combination: the most irreversible legal step is complete while the route, the cost, and the economic case all remain unsettled.

The analysis that isn't there

For a commitment of this size, the document a reader would expect to find is a cost-benefit analysis, a full accounting of whether the benefits exceed the costs. No federal body has published one, and none has committed to publishing one before the go/no-go decision targeted for 2029. [6] The Parliamentary Budget Officer, who routinely costs major federal undertakings, has produced no report on the project. During the period Bill C-15 passed the House, the office did not have a permanent officer in place: the interim Parliamentary Budget Officer's term was ending in early March 2026 with no successor named. [12]

The only mandatory federal review in the sequence is environmental. Each segment is to undergo an impact assessment under the Impact Assessment Act, with an Initial Project Description expected in early 2027. [6] That review examines environmental effects, not whether the project is worth its price. The comparative studies that produced the decision to pursue high-speed rail over the cheaper high-frequency option have not been released; Quebec's farm union has formally demanded them. [9] A written question on the order paper, Q-923, asked the government directly whether it had conducted a financial analysis of the project and what methodologies it used. [13] The question is on the record. A published cost-benefit analysis is not.

So the order is visible. The expropriation powers are law. The environmental review comes in 2027. Construction is targeted for 2029. The economic case that would tell Canadians whether the benefits justify the cost is nowhere in that line.

Both sides are arguing with unsourced numbers

The gap in the analysis does not only afflict the project's supporters. It shapes the opposition too. The Bloc Québécois has put the total cost as high as $200 billion; when asked how it reached that figure, the party did not provide a published methodology. [10] The Parti Québécois leader's statement that Quebec's share could reach $40 billion is, on its own terms, conditional. It is roughly a 20 percent share of the Bloc's $200 billion, not an independently costed estimate of the Quebec segment. [10]

The proponent's headline benefit figure is no better anchored. Alto has cited an annual economic impact of roughly $24.5 billion, equal to 1.1 percent of national GDP. [2] An independent study from the C.D. Howe Institute, modelling selected benefits over a 60-year horizon, found total benefits of $15 billion to $27 billion in present-value terms, and stated plainly that it does not assess whether total benefits exceed the project's costs. [11] Reporting on the two figures put the proponent's annual benefit at roughly 54 times the annualized figure in the independent study, a gap the study's author attributed to different methodologies and scopes. [11] The proponent's own GDP claim has itself been stated inconsistently, framed as an annual effect at launch and as a one-time level effect in later departmental material. [1]

The shape of the debate, then, is two camps trading figures that neither can fully source. The $200 billion warning has no published derivation. Which leaves it as unanchored as the proponent's own numbers.