This piece is part of Vital City's exhaustive "Subway" issue. For all the content, click here.
New York’s subway was once an international model of modernity. It connected the United States’ largest city with speed, frequency and reliability. Even today, New York’s subway has some unique features. It has 24-hour service, a rarity across the globe, and its flat fare policy allows riders to traverse the full extent of the city at a reasonable price — in contrast to, say, London, which is divided into six zones, with longer trips costing more money.
Over the past century, however, New York’s system has lost its luster in international comparisons. It is now far from the world’s most modern urban rail transit system, and it is actually substantially shorter in length than it was in the 1930s because a number of elevated lines were removed.
Fortunately, there are many opportunities for New York to improve its system — especially if it’s able to leverage funding from congestion pricing. Having visited and studied many of the world’s biggest cities’ metros, I have come across several key technologies that New York can use to make the system more effective and more attractive for its riders. These technologies could improve the existing system — or be especially effective if implemented together on new express subway lines linking the region.
Innovation almost everywhere
Over the past few decades, metro systems in other parts of the world have seen major leaps in train reliability and passenger comfort. But here in the United States, technological improvements have largely bypassed underground intracity and intercity rail.
First, consider open-gangway trains, a fairly simple idea with potentially big benefits. These allow passengers to walk from one end of a train to another, which can help reduce crowding and improve a sense of safety. Open gangways expand train capacity by taking advantage of the space between the cars to carry more riders. They also allow passengers to find less crowded sections of the train while it’s moving, enabling riders to spread out more evenly rather than being stuck in one car. Paris’s Metro system was the first mainline metro network to debut this service, in 1993, and every train it’s ordered since includes the technology.
Open gangways have since become common on metro systems throughout the world, but not in the United States.
Second, many transit systems have implemented platform doors, which protect the tracks from the platform and only open when trains arrive. These doors have several advantages. They save peoples’ lives since they prevent people from jumping in front (or being pushed in front) of trains — a phenomenon that affects more than a dozen New Yorkers annually. They can reduce the frequency of trash fires, which happen dozens of times a month on the system and which can delay riders. They allow trains to roll into stations more quickly, which can speed overall service. And they make it more feasible to air condition stations — because they seal off the space where commuters are waiting. Saint Petersburg, Russia introduced these doors with its Line 2 in the 1960s, with modern glass versions first installed in Singapore in 1987.
Metro systems in other parts of the world have seen major leaps in train reliability and passenger comfort. But here in the United States, technological improvements have largely bypassed underground intracity and intercity rail.
Finally, metro automation replaces train drivers with computer signals. This technology can increase train capacity by freeing up the space onboard now used by train operators. It can make it feasible to run more trains per hour, which increases a metro system’s capacity and may be especially relevant now that transit agencies have had trouble attracting operators. And by lessening personnel costs, it can free up funding for other priorities.
Automation, combined with high-capacity trains and platform doors, makes more frequent trains a reality. In Paris, the automated Line 14 is designed to operate every 85 seconds — meaning trains pull up to stations up to 42 times an hour on just one track in each direction. Even London’s Victoria Line, which is not driverless but which has automated train operation (meaning the drivers simply monitor train operations), offers trains 36 times per hour during the peak period. Existing lines can be converted to driverless operation; cities such as Lyon and Nuremberg have already done so, and Madrid and Marseille have projects underway.
New York is stuck
On all three of these counts, New York’s subway has fallen far behind international standards. Almost all of the trains operating on metro systems in Beijing, Delhi, Shanghai and Singapore are open-gangway, as are about a third of those in London and half in Paris. Ironically, New York had open-gangway trains built in the 1920s — the first in the world! — but it didn’t order any after that point. Now it has two newly purchased open-gangway trains in testing, and it is considering ordering more.
How about platform doors? Virtually all stations in Beijing, Shanghai and Singapore have platform doors, as do about a third in Delhi and Paris and 4% in London. New York has none, though it is studying the implementation of a few.
As for automation, Singapore’s metro is fully driverless, with 15% to 28% of the line length of the metros in the other cities operating in a similar fashion, apart from London. None of New York’s lines are driverless — and, unlike any of the human-driven lines of those other systems, actually has two operators per train. And this has consequences. When running on schedule, the Lexington Avenue line does manage to run 44 trains at peak hour on the 4, 5 and 6 trains — but that’s on two tracks in each direction. That means it’s about half as efficient as Paris’s Line 14.
New York’s limited progress on these fronts is due in part to its age; Shanghai’s Metro only opened 31 years ago! And, certainly, high costs of construction in the United States make any improvements difficult. But London’s Underground is older than the subway. And Singapore’s per-capita domestic product is larger than that of the United States.
The biggest problem facing New York, then, is neither the subway’s age or its construction; it is the perverse form of American exceptionalism to which its management clings. Even as open gangways were introduced broadly on metro systems internationally in the 1990s, transit agencies in the United States resisted altering standard operating procedure. In 2009, a New York City Transit spokesman told me that making such a change would require research and design for an entirely new subway car and was thus impossible. A representative of Washington’s Metro was even more dismissive at the time, telling me the agency had “no plans to change it just to change it.” In general, American transit agencies seem to have dismissed progress elsewhere as not worth copying.
Platform doors, similarly, are largely missing from U.S. metro systems (though they are common on airport trains, like AirTrain JFK). In New York, it would admittedly be difficult to install these features on some older platforms, which have structural columns close to the track and/or are served by trains with alternate door placement on different models. But, even on the few new stations the system has added over the last decade, such as those along Second Avenue or at Hudson Yards, platform doors have not been installed despite no such constraints.
Finally, line automation has not progressed. The system has been slowly upgrading its existing lines to a more advanced technology known as communications-based train control on the L and 7 lines, making them close to automated, though the pause in congestion pricing has eliminated funding for additional signals upgrades. Taking the next step to driverless operations would require a commitment to renovation, changes in operating practices and agreement from unions.
Looking ahead
New York once had the longest metro system in the world. Since the 1980s, however, it has lost that status as other cities — particularly in China — have constructed enormous networks. New York’s subway simply hasn’t added a fully new line for 80 years, unlike all of its peers; it is now 8% shorter than it was in 1950, even as London’s is 16% longer.
New York’s subway is now not even among the top 10 globally in terms of system length. Its urban area has a similar population as Beijing’s, but Beijing’s system is now about twice as long as New York’s. Forty-three percent of residents in the New York region live near rapid transit — compared to 51% in the Paris region.
An east-west line could link downtown Manhattan with either JFK or Newark Airport in just 15 minutes — compared to more than an hour by transit today.
Much of the discussion of the future of the subway has focused on relatively small-scale projects like the expansion of the Second Avenue line to 125th Street. While such projects would benefit the neighborhoods served, they would not do a massive amount for improving regional accessibility. Nor would they necessarily incorporate the new technologies described above.
There’s room for improvement. Recent expansion projects have produced train lines that are relatively slow. Travel on the recently built — and incredibly expensive — portion of the Second Avenue line, from 63rd to 96th Street, averages just 17 miles per hour; that’s the average speed systemwide. The fastest sections of the subway, like the A and D express lines from Columbus Circle to 125th Street, average about 29 miles per hour.
But metros can travel much, much more quickly. Consider the metro in Guangzhou, China. Its 36-mile-long Line 18, opened in 2021, has long interstation distances, few curves and modern train technology — including platform doors and open-gangway trains. It’s also ready for driverless operation. As a result, trains on that line average 50 miles per hour when serving each of the line’s eight stations — or up to 74 miles per hour when operating express services. Certain trains can travel end to end in just 29 minutes!
What would it look like if New York implemented express subways on the scale and with the speed of Guangzhou’s Line 18? They would transform travel in the region.
An east-west line could link downtown Manhattan with either JFK or Newark Airport in just 15 minutes — compared to more than an hour by transit today. A north-south line could connect Midtown with Yonkers in the same period of time, versus 50 minutes today, or give travelers an option to get between the north shore of Staten Island and Union Square in 15 minutes — compared to more than an hour today.
Brand-new subway lines may not be on New York’s horizon anytime soon, given challenges identifying even enough funding to run current lines capably. But subway system leaders would do well to consider how to progressively incorporate now globally standard technologies onto existing lines even as they plot out a more ambitious future. Doing so would improve the rider experience and make the system fit for the city’s residents.
