Carolyn Maloney, the Congresswoman representing Manhattan’s East Side, gave an interview to the Globe and Mail in which she called for high-speed rail between New York and Canadian cities. She did not specify which cities, but presumably those are Montreal and Toronto. The article quoted Andrew Cuomo as saying that connecting New York to Montreal and Toronto would be “transformative,” though it did not mention that Cuomo killed plans for HSR from New York to Buffalo. It is unclear to me whether Maloney is serious, or merely as serious as Cuomo; for the purposes of this post, let us assume that she is serious. Is it justifiable to build HSR from New York to Montreal and Toronto?
Long-time readers will know that I am skeptical of international HSR lines. But let me explain why I think New York-Toronto could be successful, while New York-Montreal could not.
First, perhaps because of the common language, the travel markets from the US to Montreal underperform those to Toronto. According to Statscan data, Toronto has about three times as many travelers to New York, Chicago, Los Angeles, and the other top metro areas in the US as Montreal does. The two cities’ metro area population ratio is only about 1.5:1; this is indeed the ratio of their travel markets to leisure destinations such as Las Vegas and Miami. US data generally points to higher numbers, sometimes by a substantial margin; it also points to a ratio of about 2.5-3:1 between Toronto and Montreal travel, this time even to Las Vegas and Miami. (US data excludes planes with up to 60 seats, but these are only about 20% of New York-Toronto departures, and of course a smaller proportion of seats.)
In addition, New York-Toronto may be in a similar situation to New York-London, in which the two cities’ common industry (finance) leads to more business travel. For some evidence of this effect, the Canadian data shows that Calgary and Houston, the two countries’ respective oil capitals, are each other’s top air market on the other side of the border. The same is of course true of financial capitals New York and Toronto, though as the largest cities in their respective countries, this is less surprising. But we should not overinterpret this effect: the New York-Toronto air market is still just 900,000 people a year (according to Canada) or 1.5 million (according to the US), though it far beats New York-Montreal’s 300,000 or 600,000.
Even 1.5 million times an induced demand factor is not enough to build HSR by itself. We could add existing travel volumes from New York to Niagara Falls and from Toronto to Buffalo, but most likely they are not enough by themselves.
The main reason New York-Toronto could be defensible is that a large majority of the New York-Toronto construction would not be done just for New York-Toronto travel. HSR on the Empire Corridor, up to Buffalo, is justifiable entirely based on domestic traffic. At the other end, the Lakeshore West corridor, which already can sustain medium speeds (GO’s top speed is 150 km/h), should be electrified and retrofitted with passing sidings based entirely on local commuter traffic. There are about 100 km between Buffalo and Hamilton, and 160 between Buffalo and Toronto, compared with 850 between New York and Toronto. Since HSR fares and operating profits roughly scale with distance traveled, the operating revenue of the lower-trafficked 100 km between Buffalo and Hamilton should really be multiplied by 8.5. If New York-Toronto traffic is about 3.5 million a year, a similar multiple of preexisting air traffic as Eurostar, then we can expect the construction of the 100 km to add about 3 billion passenger-km a year; 30 million passenger-km of revenue per km of route to be constructed is very good, comparable to the Sanyo Shinkansen. If we need to use New York-Toronto traffic to justify even Toronto-Hamilton upgrades, then we’ll have 18.5 million passenger-km of revenue per km of construction, comparable to the JR East Shinkansen network.
Of course these passenger densities, and hence returns on investment, are not available to the full line; they’re only available to this last link completing New York-Toronto. To enjoy such favorable ratio the preexisting routes must already be in place. We cannot use the 30 million passenger-km/km figure to justify building New York-Buffalo as a first step toward New York-Toronto. If Maloney intends to do that, then she is setting the line up for failure; 3.5 million passenger-km/km is too little. Amtrak has about the same on the Northeast Corridor, from which it squeezes operating profits, but the capital construction was paid by private railroads between 1831 and 1917; building a greenfield line for this performance is unwarranted. At most, we can use it to add to domestic traffic in case the merits of a domestic line are close to good enough but not quite.
New York-Montreal does not have the same advantage as New York-Toronto. Not only is the travel volume much smaller to being with, but also it would require building about 360 km of route, in the rolling hills of Vermont, to create a link of 590 km. Very little of that 360 km is a reasonable commuter rail route by itself – on the line I sketched to measure distance, only 30. So at best this is 330 out of 590. If we attempt the same calculation as for New York-Toronto, we obtain just 2.7 million passenger-km/km. Moreover, the intermediate markets are much weaker than US-Niagara Falls or Buffalo-Toronto. For now, HSR between New York and Montreal should remain an unfulfilled dream of Montreal boosters.
Of course, it’s possible that Maloney just emphasized the possible connections to Canada, and her actual drive is going to be Empire Corridor HSR, which is a welcome change from Cuomo’s opposition. Canadians do not vote in US elections. In that case, a link to Toronto would become stronger, because of the piggybacking on preexisting New York-Buffalo HSR. The line would hinge entirely on constructibility over the river and border control issues then. International links underperform, but sometimes they are short enough relative to the possibility to be worth it.