Love-Hate Relationships with High-Speed Rail, Waymo, and Electric Cars

Off-Ramps for 19 June 2025

Welcome to Off-Ramps! Today I’ll highlight four interesting pieces that I think you will enjoy reading. Please enjoy all of these on your morning commute, or save them for your weekend.

After two weeks of playing with the format for Off-Ramps, today we’re going back to basics with pieces emphasizing recent developments in mobility and transport innovation.

Transport Megaprojects as Fashion Choices

The concept of 'high frequency' faces strong opposition. There's widespread disinterest [sic] and dissatisfaction associated with the term, hindering any meaningful discussions and support...

Unsigned briefing note, quoted in Maura Forrest’s “Via Rail subsidiary paid Quebec marketing firm $330K as it pivoted to high-speed rail”

I argued in February that Canada Shouldn't Build High-Speed Rail; specifically, that the Alto project, which aims to build a high-speed rail (HSR) connection between Toronto and Quebec City, was doomed to failure because it neglected transportation fundamentals. It gives me no pleasure to learn that the latest evidence supports my case.

Documents obtained by the Canadian Press reveal that VIA Rail's subsidiary responsible for the originally-planned high-frequency rail (HFR) link, VIA HFR, decided as early as fall 2023 that the project needed to change. You’ll recall that the original plan was to build or upgrade track linking Toronto, Ottawa, Montreal, and Quebec City so as to offer faster service of 200kph and trains at least once an hour. It would have been a real upgrade for rail passengers: faster, more frequent, and more reliable service, segregated from the freight traffic that so often impedes passenger rail in the corridor.

It was a wonderful idea, but it polled poorly. So VIA HFR officials turned to an external PR firm to fix it, and what they were told was that high-frequency is boring, and what the public wants is speed. So that’s what the public is supposedly getting: a rebrand from VIA HFR to Alto, and promises of 300kph service.

This validates my February critique about Alto's “wrong process”, namely that the project emphasized looking good rather than solving problems. When faced with public skepticism, the Alto team could have attempted to explain better what they were doing to the public, or make the case to policymakers behind closed doors of the wisdom of their approach. Instead they simply opted to make the project sound more exciting, irrespective of the cost or value of the changes.

I also argued that Alto suffered from the “wrong timing”, a critique which seems sounder than ever. If the project was struggling with public support by late 2023, and the decision to shift to HSR was made, it seems, in early 2024, why was that not announced until February 2025, during then-Prime-Minister Trudeau’s final weeks in the office? The answer seems clear: electoral desperation trumped transportation planning.

The thrust of the news article is that it’s scandalous that VIA HFR spend $330,000 on this rebranding. That framing is obnoxious. The problem here is not that they spent money, it’s what they spent the money on, which was to design their project around headlines rather than solving problems. High-frequency rail would have solved VIA's fundamental problem of passenger trains yielding to freight traffic on shared tracks. Building dedicated trackage, regardless of speed, removes that bottleneck. But “dedicated tracks for reliable service” does not, it seem, buy votes.

My February analysis concluded that Alto suffered from wrong structure, wrong people, wrong process, and wrong timing. I should have added wrong priorities. A transportation strategy that emerges from focus groups and marketing research rather than ridership analysis and operational needs is an exercise in political theatre, not responsible governance.

Canada needs better intercity rail, but this evidence reinforces that Alto isn't the answer.

Some People Don’t Like Waymo…

While I’m on the subject of my older articles, let me quote myself from last fall:

…surveillance of public spaces, empowered with AI and/or facial-recognition technology, is a problem, but only because we haven’t made it clear how, and under what circumstances, it should be done. We should set those rules. Having done that, we should go ahead and deploy this technology widely, because… it’s a valuable and useful tool that would, on margin, make us all better off.

This is on my mind because of recent reporting from Wired.

During the recent pro-immigration demonstrations in Los Angeles, protesters torched five Waymo robotaxis. In a recent conversation with

Jeff Fong

about this, he suggested that this was because the protestors see Waymos as “collaborationists” and as such are fair game. I can see why the arsonists reached this conclusion; as the article states, each burned vehicle carried 29 external cameras plus more inside. As such, Wired refers to Waymos as “roving data collection nodes” that turn city streets into a “mobile panopticon”. And that data is used by the state: as per the article, Waymo has shared footage with police since 2016, responding to search warrants for everything from road rage to murder investigations.

Wired sees this as inherently problematic. I disagree.

As a reminder of the context in which we live: every Tesla records everything around it continuously; smartphones capture everything; fixed CCTV networks blanket city centres with hundreds of cameras per square kilometre. Police can, and do, use footage from public recordings in their investigations, and can issue warrants for access to private recordings (as they did in this case; Waymo didn’t give any data until a judge issued just such a warrant).

In other words, the surveillance state is all around us. Yet Wired’s concern is for the Waymos.

Or rather I should say concerns, because they have two. One is that Waymos record data in public, which I find to be special pleading. But the other has merit, which is that Waymo doesn’t talk about what data it records, how it retains it, and under what circumstances others can view it. It is indeed problematic that Waymo has not yet published any transparency reports about these requests.

But to avoid special pleading, let’s remember that we need mandatory transparency standards for all public surveillance systems, public and private. Google publishes detailed reports about government data requests; Waymo should do the same; and so too should your police department, and everyone else.

I recognize the anger in California, but burning robotaxis is a misdirection of it. Waymo's cameras aren't the problem. Inadequate oversight of how any surveillance data gets used is.

…But Others Like Waymo a Lot

Lyft offered the lowest average price at $14.44. Uber was next at $15.58. Waymo’s average price across the month’s worth of data was $20.43.

TechCrunch: Waymo rides cost more than Uber or Lyft -- and people are paying anyway

A new analysis from ride aggregator Obi shows that Waymo rides cost 31% more than comparable Uber and Lyft trips. Yet customers keep paying, with Waymo now handling 250,000 paid trips weekly across four cities.

Image credit: Obi via Techcrunch

I have argued in the past that, at scale, robotaxis will be cheaper than ridehail is today, by perhaps as much as 40%. The basis for that conclusion, naturally, was that robotaxis would not need a driver, and eliminating driver wages would make the whole system much less expensive to operate.

But it seems that even without that price advantage, which has yet to emerge, robotaxis are winning market share, because passengers are willing to pay more for the privilege of riding in a robotaxi. Obi surveyed riders on this question, and found that nearly 70% of riders prefer driverless cars. Frustratingly, Obi doesn’t seem to have asked why that is, but I imagine we can guess, especially given that the riders all stated their primary concern about robotaxi use was safety. If they feared the vehicle might drive in a dangerous manner, but were still willing to pay a premium to use it, they must have a strong concern about how human-driven ridehail operators drive… or simply that human ridehail operators are also in the car.

To coin a phrase: hell is other people.

I’m grateful to Obi and Techcrunch for making this finding available, but I am still coming to grips with what it means. One thing it might mean is that we will be in the 50% trap for even longer than I expected. If Waymo and other robotaxis can make money running a luxury service, they’ll feel less compelled to dilute their own service offering by expanding fleets to the point where they could take market share from human ridehail. And that will mean that, even in a world where robotaxis flourish, we’ll have more of the harms I anticipate from a world of privately-held AVs.

I hope that is not the case. This is a subject I will pay close attention to in the future.

Electric Vehicles Are Growing Up

Finally, I note with interest London's decision to begin charging electric vehicles (EVs) for the privilege of driving in the city core. As a reminder, London implemented a congestion charge back in 2003, charging £5 per day for access. At that time, EVs were exempt from the charge, because there were no EVs then. That is no longer the case.

It was the right decision to exempt EVs then; and I think it’s the right decision to remove the exemption now.

When London first introduced the congestion charge, it also introduced an Alternative Fuel Discount, to reward the lowest-polluting vehicles of the day. It was a wise policy, to subsidize the brave pioneers willing to bet on unproven technology simply to limit noxious emissions. Free passage through central London was a good incentive to that behaviour, and—given the few drivers who qualified—a cheap form of magnanimity.

Fast-forward two decades, and Tesla's Model Y claimed the UK's top EV spot for the third consecutive year in 2024, with over 32,000 units sold. (I suspect it will not do nearly so well when the 2025 numbers are rolled up.) Electric vehicle sales hit record levels, making up one in five new cars sold in Britain. The exemption that once encouraged innovation now simply rewards mainstream consumer choices.

TfL's original plan to eliminate EV exemptions entirely showed policy maturity; a Tesla in traffic creates congestion, and its propulsion system has nothing to do with it. TfL backed off when confronted with analysis that suggested that removing incentives could spike nitrogen oxide levels by 11% across central London. I am not sure about this argument. If EVs are suddenly subject to the charge, some EV drivers in London on the margin will cease to drive in the city, but if the charge holds it shouldn’t lead to more gasoline-powered cars arriving. But TfL, it seems, did find the argument compelling, or at least was willing to change their policy more slowly.

The compromise—discounts on the charge for EVs of 25%, declining to 12.5% by 2030—is a reasonable approach. By 2035, as we have observed, gasoline cars will no longer be available for purchase in the UK, so the policy has had a clock attached to it for some time; delay of five years seems unlikely to matter much.

Here in Ontario, EV license plates (like mine) still grant solo drivers access to Toronto-area HOV lanes, a privilege that I enjoy, but feels increasingly anachronistic. The perk made sense when green vehicles were rare enough that adding them to HOV lanes didn't materially impact traffic flow. But as EV adoption accelerates, solo drivers in electric vehicles (like me) will increasingly defeat the purpose of HOV lanes; moving people efficiently, not just rewarding vehicle choices.

Still, the passing of those prerogatives is a good thing. For electric vehicles to become ‘ordinary’ is policy success.