Would an ALTO Stop Help Kingston?
Kingston has one of the busiest stations on the network. The question that matters is not whether it gets a stop, but whether a stop would leave more people riding the train, or fewer.
ALTO’s chief executive has said Kingston will probably get a station, and that most trains would pass through without stopping. Neither the timetable nor the location of the station has been published.
Those two missing facts are exactly the ones that decide the outcome. This brief therefore tests the range: today’s railway, a faster conventional railway using the existing station, and ALTO with a station either inside the city or a twenty‑seven‑minute drive north of it, at normal fares and at fares 25 per cent higher. Every number that goes into the model is listed, so any of them can be argued with.
Of the options tested, only one leaves Kingston with more rail trips than it has today: a faster conventional railway serving the existing station, at about 12 per cent more. The best ALTO case — a station inside the city, at normal fares — roughly matches today. Every other ALTO case comes out below today’s service, by 8 to 17 per cent.
The reason is simple. Speed is only one part of what makes a train trip worth taking. ALTO’s faster run to Toronto is worth about 10 per cent more trips on its own. But cutting the number of daily stops from eighteen to eight gives that back. Charging 25 per cent more gives it back again. Moving the station twenty‑seven minutes north of the city costs another 6 to 8 points on top.
Running more trains cannot rescue it by itself. Even at eighteen stops a day, matching what Kingston has now, an out‑of‑town station at a premium fare still comes out around 9 per cent below today. And about 8 per cent of Kingston’s trips — Belleville, Brockville, Cobourg, Napanee, Oshawa — have no ALTO equivalent at any frequency, because high‑speed trains do not stop at those places.
Six versions of Kingston’s railway
The table below is the whole brief in one place. The first row is what Kingston has today. The second is a faster conventional railway from the same station. The last four are ALTO, differing only in where the station sits and what the ticket costs.
Table 1 · Headline comparison
| Option | To Toronto | Stops a day | Fare premium | Annual trips | Change |
|---|---|---|---|---|---|
| Today’s service | 135 min | 18 | none | 450,000 | — |
| Faster conventional railway, existing station | 95 min | 18 | none | 502,000 | +12% |
| ALTO, station in town | 80 min | 8 | none | 449,000 | 0% |
| ALTO, station in town | 80 min | 8 | +25% | 403,000 | −10% |
| ALTO, 27 min north | 80 min | 8 | none | 416,000 | −8% |
| ALTO, 27 min north | 80 min | 8 | +25% | 376,000 | −17% |
All four ALTO rows assume eight stops a day and that today’s conventional service is withdrawn. They differ only in where the station is and what the ticket costs. No fare structure for intermediate stations has been published, so both possibilities are shown rather than assumed. The faster conventional railway is the 240 km/h new‑build line proposed under the High Performance Rail framework, serving the existing station.
Why Kingston already rides the train
Kingston’s place among the busiest stations on the network gets cited as the reason it should have a high‑speed stop. But what produces that ridership decides whether a different kind of station would reproduce it. Four things do most of the work, and a high‑speed alignment north of the city removes two of them.
It gets two sets of trains, not one
Kingston sits halfway along the Toronto–Montréal mainline, and the Toronto–Ottawa trains use the same track as far as Brockville. So Kingston collects two timetables instead of one, and ends up with a level of service beaten only by the three biggest cities on the corridor. Frequency matters to ridership on its own, quite apart from speed: in intercity rail, a 10 per cent increase in service typically brings 4 to 7 per cent more trips.
The station serves a region, not a city
Napanee, Gananoque, Amherstview and the western Thousand Islands have no intercity rail of their own, so people drive to Kingston to catch the train. Ridership credited to a city of 132,000 is actually generated by an area several times larger.
The population is unusually inclined to take the train
Roughly forty thousand post‑secondary students live in a city of about 132,000, one of the highest ratios in the country. Many come from the Toronto and Ottawa regions and travel without a car. Kingston also has a large retired population, for whom avoiding the highway is the point of the trip, and an unusually high share of hospital, university, military and public‑sector jobs where travel is expensed and defaults to rail.
But that ridership is hard to charge a premium for
This travel is not spread evenly. It piles up at term boundaries, Thursday and Sunday afternoons, reading weeks and holidays, and it creates a matching flow of families travelling to Kingston. These are the travellers most sensitive to how often trains run and how far the station is from where they are going, and the least able to just drive instead. They are also the least profitable: peaked, price‑sensitive, and largely outside the weekday business hours a high‑speed operation’s revenue depends on.
Two things worth being clear about
The ridership figure itself is not published. Kingston’s standing as one of the busiest stations rests on statements by the operator and the Minister, not on released station‑level data. That is the first item on the list of things that should be published, at the end of this brief.
Existing demand is not the same as new demand. Busy today proves Kingston already travels by train. It does not prove that a different station would generate additional trips. Only new trips add ridership to the corridor.
There is also no flight from Kingston to Toronto. Elsewhere, high‑speed rail wins its premium passengers off aircraft. In Kingston those passengers are already on the train, so there is nobody to convert. Extra trips can only come out of cars, or be created from nothing.
Both of the things that built Kingston’s ridership — frequent trains, and a station within the city, roughly ten minutes from the core and the university — are the two things a high‑speed alignment north of the city takes away. That is what the model is built to test.
How the numbers were worked out
Every trip is priced in minutes. Add up the time on the train, the time getting to and from the station at each end, the waiting created by having fewer trains, and the fare converted into minutes using what an hour is worth to that kind of traveller. Time spent driving to a station or standing on a platform counts for more than time sitting on a moving train, because people dislike it more. Journeys that involve changing trains carry an allowance for the change. That matters for one market in particular: ALTO reaches Montréal from Kingston by way of Ottawa, so some of those journeys involve a change, where a direct lakeshore railway does not.
That total is the real cost of the trip. If it goes up, fewer people travel. If it goes down, more do. The response used here is roughly one for one: make the total 10 per cent better and you get about 10 per cent more trips.
Travellers are split into four destinations and four types, each divided by whether they have a car available: thirty‑two groups, each worked out separately and then added up. That matters because a student without a car and an expensed public‑sector traveller react to a distant station in completely different ways.
Table 2 · Everything the model assumes
| Input | Value used |
|---|---|
| Trips today | 450,000 a year through the station (tested from 400,000 to 550,000) |
| Where people go | Toronto 58%, Ottawa 22%, Montréal 12%, other corridor stations 8% |
| Who travels | Students 30%, seniors and leisure 25%, public sector 20%, other 25% |
| Share without a car | Students 85%, seniors and leisure 50%, public sector 15%, other 20% |
| Worth of an hour | $14, $20, $48 and $24 respectively, in the same order |
| Time on the train | Today 135 / 120 / 160 min; ALTO 80 / 45 / 105 min (Toronto / Ottawa / Montréal) |
| Getting to the station | Existing station 10 min by car, 20 by transit; ALTO 27 by car, 35 by shuttle |
| How that time is weighted | 1.5 times if a car is available, 2.0 times if not |
| Waiting | Half the gap between trains, weighted at 0.5, across a fifteen‑hour day |
| ALTO fare premium | 25% in the central case; 0% and 40% also tested |
| Sensitivity of demand | One for one in the central case (tested from 0.8 to 1.2) |
Far‑end access time is held identical in every scenario, which is a conservative choice: it gives ALTO the benefit of the doubt at the Toronto and Ottawa ends.
Two possible futures for today’s trains
Every service level is tested twice, because the answer depends less on ALTO than on what happens to the service Kingston already has.
Replacement
ALTO becomes Kingston’s rail service to Toronto, Ottawa and Montréal, and conventional service is withdrawn or cut below a useful level. Trips to Belleville, Brockville, Cobourg, Napanee and Oshawa lose their train altogether.
Both together
Today’s service keeps running at present frequency and ALTO is added on top. Travellers pick whichever is cheaper in total, and only the improvement over the better of the two creates new trips.
What is assumed rather than known
Four inputs are estimates, not published data: the number of trips today, where those trips go, ALTO’s journey times (the alignment for this stretch has not been published), and where the station would be. All four appear on the list at the end of this brief. The model also applies a constant response to a very large change in trip cost, which is at the outer edge of where this method behaves well. The direction of the results is solid. The exact sizes are indicative.
Where the speed gain goes
Start with today’s service and change one thing at a time. This is the clearest way to see why a faster train can still end up with fewer passengers.
Table 3 · One change at a time
| Step | Annual trips | Change | Effect of this step |
|---|---|---|---|
| Today’s service, as it runs | 450,000 | — | — |
| Cut the Toronto run to 80 minutes, change nothing else | 496,000 | +10% | +10 pts |
| Cut stops from 18 a day to 8 | 449,000 | 0% | −10 pts |
| Add a 25 per cent fare premium | 403,000 | −10% | −10 pts |
| Move the station 27 minutes north | 376,000 | −17% | −6 pts |
The second row is the entire value of high‑speed running time at Kingston: about 10 per cent. Each of the three things that come with it takes back as much or more. This calculation already leaves out trips to other corridor stations, which a high‑speed line cannot serve at any frequency.
More trains cannot fix it on its own
Suppose the number of stops is the thing that gets negotiated. Hold the station twenty‑seven minutes north and the fare 25 per cent higher, and vary how often ALTO calls.
Table 4 · ALTO at a station 27 minutes north
| Stops a day | Annual trips | Change | Range | If today’s trains stay |
|---|---|---|---|---|
| 6 | 359,000 | −20% | −17% to −24% | 0% |
| 8 | 376,000 | −17% | −13% to −21% | 0% |
| 10 | 387,000 | −14% | −10% to −18% | +0.2% |
| 18 | 408,000 | −9% | −4% to −14% | +1.6% |
| 12 (six each way) | 394,000 | −12% | −9% to −17% | +0.5% |
| 16 (eight each way) | 404,000 | −10% | −6% to −15% | +1.3% |
| 20 (ten each way) | 411,000 | −9% | −3% to −14% | +1.9% |
The range covers the whole plausible span of the model’s assumptions, at a 25 per cent fare premium. The bottom three rows read six, eight and ten as stops each way, which is the most generous reading available. It improves the result without changing the sign. The last column is the “both together” case, where today’s service survives: ALTO then adds almost nothing, because travellers only switch when it is genuinely better for them.
The fare premium matters more than the timetable
Table 5 · What moves the answer
| Stops a day | Normal fares | Fares +25% | Fares +40% | Station 25 min out | Station 45 min out |
|---|---|---|---|---|---|
| 6 | −12% | −20% | −24% | −17% | −23% |
| 8 | −8% | −17% | −21% | −13% | −20% |
| 10 | −5% | −14% | −19% | −10% | −17% |
The last two columns hold the fare premium at 25 per cent and vary the drive from the station to downtown; the central case is 27 minutes. Notice that going from normal fares to a 25 per cent premium costs more than doubling the distance to the station.
How many trains would it actually take?
The more useful question is what it would take for an out‑of‑town ALTO station to be no worse for Kingston than the service it already has. At normal fares the answer is nine stops a day for everyone. At a 25 per cent premium, the answer falls apart.
Table 6 · Daily stops needed just to match today, Toronto trips
| Who is travelling | At normal fares | At a 25% premium |
|---|---|---|
| Public sector and institutional (expensed) | 9 | 13 |
| Other business and leisure | 9 | 27 |
| Seniors, leisure, visiting family | 9 | 48 |
| Students and young adults | 9 | no number works |
Forty‑eight stops a day is a train every twenty minutes all day. For students, no frequency at all makes up for a distant station plus a premium ticket, because their time is worth less than the fare increase costs them.
One case runs the other way and should be said plainly: expensed public‑sector travel between Kingston and Ottawa is better off under ALTO in every scenario tested, because today’s service on that pair is slow and indirect. It is a real gain, and it is a small share of the total.
Under the friendliest assumptions available — normal fares, a station twenty‑five minutes out, today’s trains kept running alongside, ten stops a day — the best figure the model will produce for an out‑of‑town Kingston station is about +8 per cent. Getting there means giving up the premium pricing the revenue case depends on everywhere else.
What if you just made today’s trains faster?
Now reverse the test. Keep the existing station, keep eighteen stops a day, keep normal fares, and change nothing but speed on the existing route.
Table 7 · Speed alone, from the existing station
| Toronto journey time | Time saved | Annual trips | Change |
|---|---|---|---|
| 135 min, as it runs today | — | 450,000 | — |
| 118 min, reliable 160 to 177 km/h | 13% | 471,000 | +4.6% |
| 95 min, a 240 km/h conventional railway | 30% | 502,000 | +11.6% |
| 80 min, upper bound for this station | 41% | 525,000 | +16.6% |
The last row applies high‑speed running time to the existing station. It is there to separate speed from station location, frequency and fare, not as a proposal.
The comparison that matters
Eighty minutes to Toronto from the existing station, eighteen stops a day, normal fares: about +17 per cent. The same eighty minutes from a station twenty‑seven minutes out of town, eight stops a day, fares 25 per cent higher: about −17 per cent.
The time on the train is identical. The two outcomes are thirty‑four points apart, and every one of those points is station location, frequency and fare.
Speed gives diminishing returns
Roughly speaking, every 1 per cent cut in journey time buys about 0.4 per cent more trips. A 30 per cent time saving buys about 12 per cent more passengers. For most of Kingston’s travellers, time on the train is a minority of what the trip really costs them — fare, getting to the station and waiting make up the rest, and speed does nothing about any of those. The gain concentrates where an hour is worth most: on a 95‑minute conventional railway, public‑sector travel grows about 17 per cent, business and leisure 13, seniors and leisure 12, students 10.
A conservative figure, and a warning
These figures are cautious. The response to journey time implied here is weaker than the rail literature usually finds, because the fare term in the calculation dampens it. Using a more standard figure, the same 30 per cent time saving would give about +22 per cent rather than +12. Table 7 should be read as a floor, with the 95‑minute case plausibly worth anywhere from +10 to +25 per cent. The comparisons earlier in the brief are unaffected, because they compare like with like.
The warning is that fares erode the gain fast in either direction. Raising tickets 10 per cent to help pay for an upgrade cuts the benefit from about +12 per cent to about +7 — two‑fifths of the speed gain eaten by a 10 per cent fare rise. That is the same mechanism that sinks the high‑speed cases, working here on the alternative. It is an argument for funding an upgrade from capital rather than from the farebox.
A stop is not the same as service
Three things set whether Kingston gains or loses, and speed is not one of them: how far the station is from where people are actually going, what the ticket costs, and whether today’s trains survive. Frequency cannot rescue the result on its own. At eighteen stops a day, matching today, an out‑of‑town station at a premium fare still comes out around 9 per cent down.
A public debate about whether Kingston gets a station, and how many trains stop there, is a debate about the wrong variables.
The two things ALTO has said do not fit together
A station justified by strong ridership, but served by a minority of trains, has its timetable set by the express service rather than by the demand used to justify it. Table 6 shows why that is not a workable compromise: the frequency needed to make the station work at a premium fare is far above what an express pattern tolerates. The usual international answer is two tiers, express and semi‑fast, which needs somewhere for fast trains to overtake at the intermediate station. Whether the cost estimate includes that overtaking capacity is a question with two possible answers, and both are informative.
A conventional railway does better here
A new conventional railway built for 240 km/h, running typically at 200, serves Kingston without moving the station, without the fare premium high‑speed operation needs, and without cutting the number of trains that stop. It captures a smaller share of the theoretical time saving and a larger share of the ridership. That is the trade the tables above quantify.
Three commitments, and four documents
None of this is a prediction that a Kingston station must fail. The results turn on assumptions, and those assumptions are all things the project could settle.
Four things that should be published
Before any of these figures are treated as more than an order of magnitude, four inputs should be replaced with real data:
The first sits with the operator. The other three sit with the project and its joint project office, whose report and business case remain unpublished.
The finding that matters
It is not that a Kingston station would fail. It is that the service Kingston already has is the benchmark the project has never been asked to beat — and on the assumptions set out here, it does not beat it.