Across the country, transit ridership is still down significantly from pre-covid numbers. After three years, it’s safe to say that commuting patterns have changed permanently, and transit agencies are still working on adjusting.
As can be seen in the chart, some transit agencies have done better than others at getting ridership back to pre-pandemic levels - and the best major North American agency isn’t in America at all. It’s in Canada - where Vancouver has seen Skytrain ridership at over 80% of pre-pandemic levels, and rising - it hit 88% of pre-pandemic levels this summer. The Vancouver metro area has approximately 2.5 million people and only 50 miles in its Skytrain network which in many ways makes it a more replicable example than a megacity of 20 million like New York. For example, the Portland area is also in the Pacific Northwest, also has a metro population of ~2.5 million, and also has significant transit in the form of a 60-mile light rail that allows the cities to serve as natural comparison points. Vancouver is currently averaging over 400,000 Skytrain passengers per day while Portland’s MAX light rail is averaging only 70,000.
I was recently in Vancouver and was able to take note of all the things the city is doing right to bring transit ridership back, which can serve as a roadmap for other American transit agencies that are still trying to find their footing after the pandemic.
1. Frequency, Frequency, Frequency
In Vancouver, the Skytrain comes all the time. In Boston, Portland, or Los Angeles, I would frequently worry about missing a train and being stuck waiting for more than 15 minutes. Not the case in Vancouver, where the Skytrain comes every 3-4 minutes during the day and every 5-10 minutes late at night. This means that you don’t have to worry about “missing” a train because the next one is right behind it. Additionally, it softens the time penalty of transferring lines. In a city like LA where you might have to wait 15 minutes at a transfer, taking transit becomes much less appealing. But when that worst-case scenario transfer wait time is shortened to 5 minutes, it becomes far more competitive with driving.
Additionally, where Vancouver excels is off-peak frequency, with frequent service in the middle of the day, at night, and on weekends. Pre-covid, many transit agencies were built around the rush hour peak where everyone was trying to get to work from 7-9AM and leave work at 5-7PM. However, with the rise of remote and hybrid work, this peak has flattened, with mid-day transit ridership matching (and in the case of New York City, actually exceeding) pre-pandemic levels, even as peak commute hour ridership is down significantly. This actually increases transit flexibility. Pre-covid, it was pretty inefficient for transit agencies to have to build a system around these peaks, with extra train cars and capacity unnecessary for most of the day.
With much of the shift to remote/hybrid work being permanent for many former commuters, transit agencies should (if they haven’t already) be adjusting to these changing commuting patterns in all aspects of their work, from procurement of new trainsets to railyard size to hiring and labor practices. Additionally, “commuter rail” networks currently built around the 9-to-5 commute should be restructured as “regional rail” with frequent all-day service.
2. Better Bus Service
Frequent train service is the baseline but many people in Vancouver (and most American cities) don’t live a short walk from a train - so how do they get to the train? Buses. But buses aren’t an effective transportation option if they only come every 30 minutes. Even though Canada suffers from the same sprawly suburban-centric urban planning that America does, they have far higher bus ridership - because the buses come more frequently (their most frequent bus line comes every 3 minutes!) Additionally, this reduces the transfer penalty for switching from a train to a bus and provides a compelling alternative to driving.
Vancouver (and other Canadian cities) punch far above their weight class when it comes to bus ridership, and there are additional lessons we can learn beyond having the buses run more frequently. Specifically:
We should redesign our bus networks to account for changing travel patterns. Many cities are still using the same exact bus networks that were drawn up pre-covid. That’s obviously silly given how much travel patterns have changed, and re-drawing bus networks based on recent population growth and travel patterns is a best practice.
We should build fewer, nicer bus stops. Designing a bus network requires constant evaluation of trade-offs. A key one is: how far apart should bus stops be? Too close together and the bus will get very slow, stopping every single block to pick someone up or drop them off. Too far apart and the stops are less convenient to walk to. Thankfully, people smarter than I am have crunched the numbers and it turns out the optimal stop spacing is ~400-500 meters. But most American bus networks err on the side of too many stops, too close together - so consolidating bus stops can frequently result in significant speed gains (a recent analysis found that stop consolidation would speed up buses by 40% in New York City). The same thing can be said for bus lines as well - two parallel streets very close to each other don’t need duplicative bus lines. Instead, just have one bus line with more frequent buses. A final bonus is that fewer, more thoughtfully-planned bus stops means that it’s easier to upgrade each stop with features like benches, shelters and countdown clocks.
We should build more bus lanes. Buses full of dozens of people shouldn’t be stuck in traffic behind cars carrying one or two people. It’s common sense. Bus lanes fix this issue by allowing buses to bypass cars, drastically increasing the passenger capacity on a single road. Additionally, bus lanes increase bus frequency (a bus going faster covers more ground and can run the same route more frequently), prevent delays, and result in all-around more reliable service. For this to work, though, bus lanes must have physical infrastructure and automated enforcement to ensure that cars don’t just flaunt the rules and clog up the bus lanes anyway.
Smaller tweaks to improve reliability and speeds. There are a lot of little things transit agencies can be doing to improve bus service. Some of these include having all door boarding + proof of payment to prevent passengers from lining up to get in the bus (and slowing it down) along with better payment options such as mobile apps and contactless cards, adopting transit signal priority so that buses spend less time waiting at red lights, and finally using headway management rather than a fixed schedule to prevent bus bunching.
3. Turbocharged Transit-Oriented Development
One thing that was stunning about my time in Vancouver was seeing just how seriously the city embraced transit-oriented development. Again, it’s one of those ideas that seems like common sense: building apartments near transit so that more people can live near the train and conveniently take it. But seeing the scale of it in person was remarkable.
In the US, most suburban “transit-oriented development” takes the shape of 5-10 story buildings, but Canada is taking it to the next level. Almost every transit-connected suburb in the Vancouver area is building genuine skyscrapers of apartments immediately adjacent to their transit stations. Some examples in the Vancouver area include Burnaby (80 stories!), Surrey (67 stories!), and Coquitlam (50 stories!), where tens of thousands of new apartments are being built. The tens of thousands of people who live in these new high-rises will have incredible access to transit, boosting ridership and fare revenue. Additionally, many cities own under-utilized land near transit stations that can be developed into dense housing, with profits used to help fund the transit - it’s truly a win-win. This serves as a sharp contrast from, for example, Long Island where after spending $12 billion on a direct LIRR connection to Grand Central, most suburbs are refusing to build any housing. A few single-family homes and a surface parking lot near the train station generate much less ridership than dense, mixed-use apartment buildings. We shouldn’t let rich NIMBYs on Long Island prevent more people from utilizing this (extremely expensive) transit infrastructure.
4. Build Skytrains, Not Streetcars
Vancouver’s Skytrain is an extremely nifty train that works as an Automated Light Metro. Basically, at it’s core, it’s like a souped-up airport people mover and has a series of benefits that align quite well with our post-Covid reality. Recently, I have been able to ride similar automated light metros in Copenhagen and Taipei as well - and it’s incredible. Some key benefits include:
It’s entirely automated: Post-COVID transit worker shortages are preventing agencies from running frequent service. In Chicago, train frequency has slipped significantly due to worker shortages. However, thanks to the wonders of automation, Vancouver and other cities with automated light metros like Copenhagen and Taipei can easily run frequent trains around the clock.
It’s easier to build: As I covered in my analysis of American transportation cost overruns, the most expensive part of building a subway is building the station itself. However, because automated light metros come so frequently, they can use smaller trains to achieve the same capacity. As a result, the stations can be smaller, which creates massive cost savings. And these cost savings result in more transit construction: Montreal is currently building a 67-KM automated light metro network for $6 billion, less money than the proposed 3-KM Second Avenue Subway extension. It’s literally 20x cheaper and as a result, Canada is getting 20x as much rail service for the same price
It’s better aligned with post-Covid travel patterns. The main benefit of a heavy subway in comparison to an automated light metro is the pure capacity. A massive 10-car train on the New York City subway can carry over 2,000 people while a Copenhagen metro train can only carry 300 people. However, with lower peak travel demands that are smoothed out over the course of the day, most cities won’t ever reach that level of peak demand again. As a result, an automated 300-person train coming every 3 minutes is a far better option than an old 2000-person subway train coming every 20 minutes. That said, traditional subways still make sense on extremely-high-demand routes, particularly in large cities like New York due to this increased capacity, but mid-sized cities will be served far better by an automated light metro.
Streetcars and light rail are the alternative. Most American transit investments are like what we see in Seattle - a sprawling light rail network that only comes every 10 minutes. Light rail has a significantly lower capacity than heavy rail like the subway and often has lower frequencies than an automated light metro like Skytrain. As a result, Seattle’s light rail is expected to suffer from overcrowding. Even so, Seattle’s light rail is better than most in America thanks to the fact that it’s largely separated from cars and can travel at high speeds. Where American light rail/streetcars really suffer is in cities where trains are forced to mix with automobile traffic which slows it down and reduces frequency and reliability. On the other hand, Skytrain is fully grade-separated and completely immune to any traffic.
We already have the land for it! American streets are extremely wide, largely thanks to the legacy of streetcars running down the middle of major roads. While these streetcars were mostly destroyed during the 1950’s-70’s, there are tons of excessively wide roads with large medians that would be the perfect place to construct elevated Skytrains that would be much cheaper than tunneling a subway or acquiring land for an entirely new right of way.
Interestingly enough, there already is one American city building an automated light metro: Honolulu. The geography and size of the city are perfect for an automated light metro and while its first phase has opened, the real benefits of the project will be when the second and third phases are completed to the airport and city center over the next decade.
5. Building Safer Roads
It’s not your imagination: people are driving worse and our roads are more dangerous after the pandemic. I wrote a longer piece about how America could make its (extremely dangerous) roads safer, and I saw many of these strategies utilized in Vancouver. They have automated speed and red light cameras at many intersections and an extended network of protected bike lanes. While Vancouver is still improving their Vision Zero program, they’ve had under 20 traffic fatalities every single year since 2008. Meanwhile, Portland’s traffic deaths are rising, reaching a new high of 63 deaths in 2022. To make things worse, Portland is planning on removing protected bike lanes, which will result in even more traffic injuries and fatalities.
To wrap things up, we don’t need to travel to Tokyo or Paris to learn how to adjust our transit systems post-covid - we can just look north and see what is going on in Vancouver and how they’ve been so successful at getting ridership to rebound. None of the policies here are impossible or even difficult to enact, we simply need leaders willing to honestly evaluate these trade-offs rather than just hoping things can magically return to business as usual.
I'm curious how Vancouver’s Skytrain compares to the monorail options currently being proposed for the Sepulveda line here in LA that have received overwhelming pushback from basically everyone except the wealthy homeowners in Bel Air. I'm not familiar with the details aside from it supposedly being lower capacity, but I wonder if there's a compromise* — rail type? train length? — that gets us a cheaper and faster project with automated service and higher overall capacity. I understand not wanting a half-measure solution but I'm worried that a subway would turn into a decade-long megaproject.
(*a compromise between the proposed solutions, not with the residents, who would obviously prefer nothing to be built at all)
How much of Canadian ridership can also be attributed to lower wages and higher cost of living especially in Vancouver and Toronto? In other words, if the average Canadian was as wealthy as the average American, would we see the same ridership numbers? Assuming wealthier people will opt for driving a car over taking transit.