The widespread failure of American mass transit is usually blamed on cheap gas and suburban sprawl. But the full story of why other countries succeed is more complicated.
When it comes to the quality of public transit, comparisons between American cities and international counterparts are usually met with a simple response: “It’s different over there.”
These differences, the argument goes, are vast and fundamental: Europe is far more densely built, and its older cities—settled centuries before the automotive age—will always be innately transit-friendlier. In Asian cities, meanwhile, explosive urban growth has been accompanied (and accelerated) by massive government investments in urban rail networks. But the U.S. boomed in the 20th-century’s automobile age, and the private car is still king; most American cities and their suburbs are utterly dependent on them.
How did transit become such an afterthought in Americans’ transportation habits? I addressed that question in detail in an earlier CityLab piece. But to briefly summarize: Transit everywhere suffered serious declines in the postwar years, the cost of cars dropped and new expressways linked cities and fast-growing suburbs. That article pointed to a key problem: The limited transit service available in most American cities means that demand will never materialize—not without some fundamental changes.
The U.S. did see a resurgence of transit spending on big projects in the 1960s, ‘70s, and ‘80s. But there was a big difference between America’s approach to big urban metro projects and that of Canadian and European cities: Even when the United States built expensive rail systems, it never took care of the basics. Cities rarely provided good, frequent local connections that allowed people to actually get to rail stations without walking or driving. To this day, in most parts of American cities, it is all but impossible to get anywhere on a Sunday at 8 p.m. by transit, and if you miss the bus you might be waiting an hour or more for the next one. Such a situation is virtually unheard of in most other developed countries, where even many small villages have a relatively regular bus.
That’s the fundamental problem that makes transit useless for most people in most American cities. The key to great transit service is not about getting 100 percent of people to ride transit for 100 percent of trips. It’s about giving people a viable choice of getting around without needing to drive.
Figuring out how to improve transit isn’t like curing cancer or inventing a quantum computer, either. There are good, viable models of transit systems that already exist in cities that look a lot like U.S ones. They are successful both at attracting riders and at being financially viable, from places that have more in common with American cities than one might expect.
Like Americans, Europeans saw the car as the wave of the future after World War II. Automobiles became a symbol of postwar reconstruction, from the Citroën Deux Chevaux to the Volkswagen Beetle. Cities across Europe developed elaborate plans to accommodate cars, bulldozing many historic neighborhoods (if they hadn’t already been destroyed in war) to make way for wide roads and parking. Urban areas expanded dramatically with the construction of suburbs and new towns.
There was an important difference with what was happening across the Atlantic, however. Even in a country like Switzerland in 1960, which had a per-capita income higher than the United States, vehicle ownership per capita was barely a quarter as high as in the U.S. What happened? Unlike their American counterparts, European planners designed new suburbs in ways that made transit use still viable. Many new towns were built around train and metro stations.
Early U.S. suburbs like Levittown, New York, on the other hand, were built along highways and had virtually no transit service at all. They’re almost all still built on the model developed in the 1940s: single-family homes on isolated streets, with stores surrounded by parking lots a decent drive away.
In contrast, places like Vällingby, a Swedish suburb outside Stockholm built in the 1950s, were sited around a new Metro station. Building rail infrastructure through built-up areas is extremely expensive, but building it through farmland, before new neighborhoods are built, is comparatively cheap.
This is the model that cities like Copenhagen, Paris, Munich, and Amsterdam have used. Instead of building highways first, which tends to make neighborhoods auto-centric and de-prioritizes transit, European cities tended to put transit first when they built new neighborhoods.
Many, though not all, major cities in the U.S. have a number of rail lines radiating out of their centers. Most of them are only used by freight or a few commuter train trips a day. It’s a huge, untapped resource. There’s no reason why those railway lines can’t be turned into what are effectively subway lines—high-capacity routes that allow people to get across the city quickly—without the immense cost of tunneling. In Europe, what we usually call “commuter rail” operates frequently, all day, and cost the same fare as other local transit. That’s the difference between regional rail and commuter rail. A transit system with service that is only useful to 9-to-5 commuters to downtown will never be a useful one for most people.
In the 1970s and ‘80s, Paris built its RER network, which included new tunnels to connect its disconnected Paris terminals. The system now moves more people per year than all American commuter railways combined. Many German cities built similar systems, and London is now following them with Crossrail.
Look at any American commuter rail system, and imagine that instead of high fares and a few trips a day designed to take people to jobs downtown, it had trains every few minutes, all day, and the same fare as other local transit. Los Angeles’ Metrolink, Chicago’s Metra, and Boston’s MBTA commuter rail all have extensive networks that blanket their metropolitan areas, but their trains are too infrequent and too expensive to be used as a real local transit service. Why do we only run decent service on expensive subways that were built from scratch?
A good feeder bus system can save huge capital costs, because it can bring people to the existing rail line, thus eliminating the need to bring a rail line through existing development to the riders. Converting existing rail lines to run real transit service can be shockingly cheap: Ottawa converted a lightly-used freight route to a five-stop rail transit line with trains every 15 minutes for only $16 million. By comparison, one station on New York’s Second Avenue Subway cost $740 million; the 2.2-mile-long D.C. Streetcar cost $200 million. Even on lines with heavy freight traffic, adding extra tracks for passenger service costs a fraction of the cost of subway tunnels.
For another example, take Munich, a German city with a population roughly comparable to that of Denver or St. Louis. It has 95 kilometers of subway—only about half the length of Washington’s Metro system. But on top of that, it has 434 kilometers of S-Bahn, which is like American commuter rail, except that it uses the same fare as the local bus and metro service and its trains come every 20 minutes, all day. Through the downtown core, all the S-Bahn routes combine into one underground line that acts as a kind of super-subway, with trains every two minutes.
If you wanted to make a trip from your house in an outer, single-family residential district to your friend’s house on the other side of the city, you can do that at any time. You catch a bus near your house and ride it to the nearest S-Bahn station. On one of those quick trains, you can get across the city in 30 minutes. Since Munich worked to locate development around rail stations, there’s a good chance that your friend’s house is within walking distance of one of those stations. Because there’s frequent, all-day local bus service at my house, and quick rail service to get across town, you can make a trip whenever you want, and it’s not too much slower than driving.
It would never have been affordable for Munich to build that many miles of rail if it had to be in tunnels, like a subway. But most of these lines were already built a century before and had previously been used just for freight, long-distance, and conventional commuter trains. So it was comparatively cheap to adapt them to also provide local transit service. With that many kilometers of rail transit, the idea of centering most development around rail stations is actually realistic.
The 1950s, ’60s, and ’70s were a time of almost unimaginable highway construction, in both the the U.S. and Europe. Many European metros are encircled by expressway networks that would impress anyone from L.A. or Houston.
In Germany, for example, high-speed Autobahnen go just about everywhere. The land of BMW and Mercedes-Benz boasts a strong car culture, and its plans for a national network of expressways were first formed in the 1920s; indeed, these highways helped inspire America’s interstate build-out. But Germany never stopped building rail systems.
The U.S. did stop building rail, despite much talk among American planners of “balanced” transportation plans that included both highway and public transit improvements. There were nearly no significant rail projects between the New Deal era in the 1930s and early ‘40s, and the Great Society era of the 1960s. During that long gap, the expertise of U.S. builders atrophied, which helped drive up construction costs.
Take the Northeast Corridor Improvement Project of the 1970s, for example, which was Amtrak’s first major project after its formation in 1971. Amtrak had to rebuild its engineering expertise, at great expense, and the project ended up going way over budget. It cost about as much as the Paris-Lyon TGV line, an entirely new dedicated track for high-speed trains. But it still failed to meet its goal of offering genuine high-speed rail in the United States’ densest corridor.
Fares need to be low enough that people can afford to take transit. New York City will soon join other cities like Tucson and Ann Arbor in having discounted fares for low-income people. That is important to make transit accessible to everyone. But fair fares isn’t just about keeping fares low. It’s also about eliminating arbitrary inequities. People shouldn’t have to pay a transfer penalty or a double fare just because they switch from bus to rail, transfer between agencies, or travel across the city limits. A transfer is an inconvenience—you shouldn’t have to pay extra for it.
Fares should be set for the convenience of riders, not government agencies. A trip of a similar distance should have a similar fare, regardless of whether it’s on a bus or train, or if you have to cross city limits. Commuter rail shouldn’t be a “premium service” that only suburban professionals can afford.This is the kind of unfairness that infuriates people and drives them away from transit.
There are a number of good fare models in Europe. Paris has a single large agency, RATP, that oversees most transit in the region, like the MTA in New York, SEPTA in Philadelphia, or the MBTA in Boston. Unlike those American cities, however, every mode is part of the same fare structure. When you switch from bus to commuter rail, you don’t pay a transfer penalty. Within the city of Paris, there is a simple flat fare, even for commuter rail, while outside the city there is a zone fare system. This means that the city fare isn’t unreasonably high, and that people traveling from outside the city, including suburbs that are disproportionately low-income and minority, aren’t faced with punitive double fares when they switch from suburban buses or rail onto the city system.
In the German-speaking world, even when transit is run by many different agencies, they cooperate in a “traffic union” that coordinates fares and schedules. Riders don’t even realize they’re switching between the German equivalents of Amtrak, Metra, and the CTA “L” train.
All too often, transit planners—and even advocates—find themselves resigned to fatalism about the prospect of transit in American suburbs. They’re convinced that these spread-out and car-centric spaces are fundamentally irreconcilable with public transportation.
But transit can be successful even in typical North American postwar suburbs. The key? Providing good service, even in places that are typically deemed “not transit supportive.” It’s worth looking closely at how things happened differently for a nearby neighbor: Toronto, Ontario.
As in most American cities, Toronto’s suburbs grew rapidly in the postwar years. Its local transit system was relatively strong in the 1940s and ’50s—strong enough to fund construction of a subway line that opened in 1954—but service was mostly limited to the prewar city. Per capita ridership plummeted, as in other North American cities, as people bought cars and moved into bungalows with big backyards. But then something happened.
In 1954, a metro government—sort of like a very strong county government—was created by the province of Ontario and charged with overseeing many services for the both city and its suburbs, including transit. By then, suburban areas had a majority of the population within the new metro area, so they had a lot of clout. They demanded equal service to the prewar neighborhoods in the historic city limits, eventually along with a single flat fare for the entire city, thanks to provincial subsidy. Many experts felt it would bankrupt the transit system.
But, to the surprise of most, it worked. Many people who could afford cars chose not to buy them, and instead got around by transit. Unlike every major American city, Toronto reversed transit ridership decline in the early 1960s. A strong suburban bus network fed the older subway system and made transit a viable mode everywhere, which makes the subway system far busier per mile than its D.C. and San Francisco contemporaries to this day, which rely mostly on walk-in and park-and-ride traffic.
It’s not that Toronto suburbs are different from American ones—there are plenty of single-family homes on winding cul-de-sacs, and a large majority of Toronto households own a car. Not only does the city have highways, it boasts a 16-lane expressway that’s the busiest in the world. The difference, though, is that transit service is also good—so much so that lots of people choose to leave their car at home for many of their trips.
The suburban York Mills subway station, for example, is surrounded by surface parking, a golf course, and multi-million-dollar houses on large lots. It sounds like just about the worst environment imaginable for a subway station. And yet, it gets 10,890 riders a day--more than many stations in Manhattan, and most in Brooklyn.
York Mills also isn’t atypical. In fact, it’s one of the less busy suburban stations in Toronto. That’s possible because TTC subway stations in the suburbs don’t merely rely on walk-in traffic. Instead, the TTC uses frequent local buses to bring people to the subway, on which they can make longer-distance trips around the city. Nearly every Torontonian is within a 15-minute walk of a 24-hour bus route. Virtually every one of the major roads on the city’s grid has a bus route that comes at least every 10 minutes, all day long. People making long trips across town usually transfer to the subway for a quicker ride, but it is also convenient to make cross-suburban journeys by transferring between buses—they come frequently enough that there is little risk of standing for an hour at a forlorn suburban bus stop waiting for the connection.
These bus routes are remarkably well-used. Though it runs through a postwar suburban sprawl of strip malls and shopping centers, Finch East—a typical arterial route—is one of the busiest bus routes in North America with 42,900 riders on an average weekday. That’s more than all but a handful of routes in New York City.
Despite its high service levels in low-density areas, Toronto’s transit system is also far less subsidized than any major American system. It gets about 70 percent of its revenue from fares, compared with 47 percent for the New York City Transit Authority, 30 percent for Boston’s MBTA, and 21 percent for Miami-Dade Transit—strikingly, a pretty clear inverse relationship with service level.
For people who can’t drive or choose not to, the dependence on cars in the U.S. is a huge mobility barrier. It also means that many drivers suffer long commutes on congested roads. Cars and trucks are the largest source of carbon emissions in the U.S., worsening air quality and driving climate change. America’s dependence on them also contributes to the extremely high cost of housing in urban neighborhoods that are served by transit. Because so few places fit that description, people who prefer not to drive everywhere pay an enormous premium.
Take Nashville, for example. There, a well-organized anti-transit lobby defeated a referendum to significantly improve transit service. In most of the city, if you’re not in a car it’s all but impossible to get around after 10 p.m., even if you live on some of the major routes. Other parts of the city lack Sunday service, and many bus trips that should take only 30 minutes end up tripling that travel time because of the long waits for connections. This isn’t the kind of transit that would encourage anybody to leave their car at home.
Sustaining higher gas prices won’t single-handedly save transit systems, as advocates sometimes wishfully hope. European countries shows that, while pricier fuel may drive some people to transit, it doesn’t make a meaningful difference if service isn't improved first.
In some ways, the story of American transit is not so unique. Europeans and Canadians also like to drive. Their countries have also built big expressway networks. The difference is more basic, yet profound: When transit service isn’t good, few will choose to use it.
Fortunately, improving American transit doesn’t necessarily demand multi-decade, hundred-billion-dollar infrastructure projects: It can be done by better advantage of existing space and existing vehicles, and then deploying them in ways that encourage people to actually use them.