This morning, the Parliamentary Budget Officer (PBO) released a cost analysis report on ALTO, the federal Liberal government’s proposed high-speed rail line between Toronto and Quebec City. According to the PBO, ALTO is likely to cost substantially more than the government’s public estimates – a conclusion shared by me and others who have examined the government’s case for ALTO.
The PBO report is available here: https://www.pbo-dpb.ca/en/publications/RP-2627-009-C–canada-eastern-corridor-high-speed-rail-cost-analysis-using-global-data–train-grande-vitesse-corridor-est-canada-analyse-couts-aide-donnees-mondiales
—–o0o—–
- Global cost of constructing ALTO is $75 billion – $113 billion (not $60 – $90 billion). The midpoint of this range ($94 billion) is higher than the highest-range estimate given by the Government. This equals a 25% increase in estimated costs, as the midpoint of the new estimate ($94 billion) to the midpoint of the Government’s earlier estimate ($75 billion is the midpoint between $60 and $90 billion).
—–o0o—–
- This report deals only with construction costs. Operating costs and ridership projections are separate, and relevant. The following statement is made on p. 8 of the PBO report (emphasis added by me):
The motivation for HSR projects typically includes the economic benefits to business and pleasure travellers of shorter, more predictable, and ideally cost-effective, travel as compared to travelling by road or air – as well as the environmental benefit of reducing fuel consumption and infrastructure depreciation accordingly. PBO is currently compiling information about regional travel volumes across these different modes of transport as the basis for this subsequent study.
And in the summary on p. 6:
This report is the first of a two-part assessment of the viability of the proposed HSR line; a subsequent report will examine whether projected ridership is sufficient to support the system’s operations.
—–o0o—–
- Total cost must include operating costs, which are likely to be between $64 – $69 billion. Given that only construction costs are taken into account, the huge operational costs that were leaked to the Globe and Mail (published on August 31, 2026) should continue to be added to the new, elevated construction-cost estimate. The leak to the Globe shows that the cost to maintain and operate the rail line would be somewhere between $64 and $69 billion. Added together, the operating costs and the construction costs lead to a range of $139 billion – $182 billion, with a midpoint of $160.5 billion.
—–o0o—–
- Diverting the rail line south to Kingston would add roughly $4.4 – $6.6 billion to construction costs, leading to a new project total of $143.4 – $186.6 billion.
The PBO refuses to assign a cost to the proposed diversion, but notes on p. 14 that
While this report does not estimate the cost of a Kingston route as announced on June 22, 2026, it should be noted that the more sensitive nature of the Kingston ecology as well as its denser population and later planning start vs the baseline route are all significant risks that may place the Eastern HSR project at greater risk of U.K. and U.S. style cost escalation.
I have based the additional cost on the PBO’s twin assertions that:
(a) the diversion would add 50 km to the train route (p. 15); and
(b) construction costs per km = $88 million – $133 million (p. 16).
Please note that the PBO is explicit that the $88 – $133 million per km is a base estimate; tunnels, bridges and viaducts are significantly more expensive, and the PBO clearly assumes that at least some bridges and viaducts would be needed along the 50 km expansion, meaning that the $133 million-per km estimate is probably unduly optimistic.
The $4.4 – $6.6 billion estimate is based on multiplying the per-km construction costs by 50 km.
—–o0o—–
- Factors based on local opposition could cause costs to go up significantly. The PBO’s report is vague on what exactly is meant by this, but the matter is clearly taken seriously. On p. 17, the PBO writes,
[In] scenarios where Bills C-5 and C-15 are not sufficient to contain costs related to process issues (e.g., environmental or stakeholder challenges, etc.) as was observed in the United States, and/or management-related delays extended the project, as was observed in the United Kingdom, costs could rise significantly beyond the PBO-estimated range. The difference between this adverse scenario and the reference scenario primarily reflects institutional and cultural factors, more so than geographic and technical factors, as the cost overruns observed in the U.K. and U.S. are seen to be largely unrelated to engineering challenges.
Elsewhere in the report, the following explanation is provided:
Both the fast-tracking of large projects through Bill C-5 (One Canadian Economy Act, June 26, 2025) and the High-Speed Rail Network Act, enacted through Bill C-15 on March 26, 2026, aim to mitigate some of these potential difficulties related to project delay. The legislation declared that the network is to work for the general advantage of Canada, deemed the required line-location approval as given, and consolidated review under a single Impact Assessment Act process. It also altered land acquisition, including a right of first refusal for Alto and amended the federal Expropriation Act.
While the passage of those bills can be expected to have diminished legal risks, the prospect of court challenges or other litigation remain risks to construction timelines. Moreover, even if the bills survive legal challenges, it cannot be assumed that they have correctly addressed the potential drivers of cost escalation.
Since the objective of project opponents is to force the government to abandon ALTO, it is reasonable to assume that they will investigate how best to use court challenges to tie up the expropriation process and to undermine the streamlining efforts laid out in C-5 and C-15.
—–o0o—–
- The report assumes that construction is becoming less efficient and more costly every year, and that this will drive up construction costs if there are any delays.
Frankly, I was astonished by this claim—and by the casual assumption that there is no point in addressing the underlying problem, which goes against every normal economic process (where costs continue to fall over time). All I can do is to quote this extraordinary claim verbatim:
The inclusion of a start date in the estimation also merits additional comment. For each later year that the project starts construction, $1.5 billion ($1.1 to $2.0 billion, in real 2024 CAD) is added to the overall cost. While productivity improvements tend to lower the cost of goods and services, some sectors of the economy (in most countries) do not benefit as much. Illustrative is the Highway, Street and Bridge Construction in the U.S. Productivity has been trending downward since 2009, and by 2024 was 25 per cent lower than in 2002.17 The construction industry in general has not been able to achieve productivity growth similar to that in the rest of the economy. This signals that the start date of construction can be a factor in how much the project will reflect these competing pressures.
Surely a 25 percent decline in productivity merits further investigation and remediation, rather than a race to construct quickly, against a trend that is casually assumed to inevitably project out ad infinitum? Surely, as well, is it not likely that regulatory factors lie at the heart of this decline in productivity? So wouldn’t it make more sense to address these first, and to start construction on ALTO-style megaprojects after the problem has been corrected?
Moreover, is it the case that the productivity decline has been even more precipitous in Canada? (It seems odd that an American statistic is being cited for a project that is planned for Canada.)
—–o0o—–
- The PBO projects a worst-case scenario of over $240 billion.
The worst-case scenario is not the likeliest scenario, but it should nevertheless be included in the discussion, so that Canadians embark on this infrastructure project with eyes wide open. On p. 21, the PBO notes:
Uncertainty in the estimation can be gauged by bootstrap replication. Using a bootstrap methodology, an upper-bound cost with 16 per cent tail probability would give a total cost $115 billion (versus $113 billion using the parameters of Table A1). Moving to an upper-bound cost with 2.5 per cent tail probability gives $243 billion. This implies that based on these international experiences, there is a 1 in 40 chance that costs would reach or surpass a total of $243 billion.
—–o0o—–
- The worst-case scenario is for construction costs only. To this must be added the operating costs leaked to the Globe and Mail.
The operating costs leaked to the Globe and Mail were $64 – $69 billion. Therefore the total cost of ALTO, in a worst-case scenario that has a 1-in-40 chance of happening, is between $307 and $312 billion.
—–o0o—–
- The per-citizen and per-family costs of Alto, as measured by the PBO, are outlined below.
There are 40 million Canadians, and therefore, there are, notionally, 10 million Canadian families. The per-citizen and per-family costs of Alto are therefore:
- PBO’s lowest-cost reference-range estimate ($75 billion):
Per citizen: $1,875 Per family of four: $7,500
- PBO’s high-cost reference-range construction cost estimate ($113 billion):
Per citizen: $2,825 Per family of four: $11,300
- Total construction cost plus operating cost ($160.5 billion). This is PBO’s mid-range construction cost ($94 billion) + mid-range operating cost ($66.5 billion, as leaked to the Globe and Mail):
Per citizen: $4,012 Per family of four: $16,050
- Additional cost of diverting the rail line south to have a stop in Kingston (mid-range estimate is $5.5 billion):
Per citizen: $137.50 Per family of four: $550
- Worst-case cost scenario—4% probability, according to PBO ($240 billion in construction costs)
Per citizen: $6,000 Per family of four: $24,000
- Worst-case construction cost scenario, plus mid-range projected operating costs ($309.5 billion)
Per citizen: $7,737.50 Per family of four: $30,950
—–o0o—–
Scott Reid
