Category: United States

» New York’s Subway is at a breaking point with an exploding number of riders. Is it time to expand the system deeper into Brooklyn?

It’s hard to fathom, but between 2009 and 2014—just five years—the New York Subway system’s ridership increased by 384 million annual rides, far more than any other U.S. rail system carries in total. This change was accomplished with no system expansions during the period, pushing more and more people onto the same already-crowded routes.

New York City’s increasing population is riding on the bench seats of the city’s subway cars. Now the City is contemplating ways to expand the system down Utica Avenue in Brooklyn; is the time right for expansion when the existing system is so crowded?

While growing ridership is a manifestation of the city’s relatively strong economy and a seemingly insatiable appetite to live there, a more crowded Subway system means lower quality of life for many of the people who rely on it daily. It means fewer available seats—if you’re lucky—a higher probability of having to wait for the next train during rush hour, lower service reliability, and, often, longer commutes.

The Metropolitan Transportation Authority (MTA) has proposed a five-year, $32 billion capital plan designed to address some of these concerns, including through the completion of the first phase of the Second Avenue Subway on Manhattan’s Upper East Side and initial work on that project’s second phase to 125th Street. The plan would also provide billions for the addition of communications-based train control (CBTC) to existing lines, which would ramp up capacity by reducing feasible train headways.

New York Governor Andrew Cuomo, who supervises the MTA, has so far failed to fill the $15-billion hole in the plan. Despite his support for fully funding the capital plan at the state level, New York City Mayor Bill de Blasio has been unwilling to commit additional City funds to bridge the gap, and has not yet announced his support for the reasonable Move NY tolling plan, which would add tolls to free bridges into Manhattan while reducing costs for many Outer Borough drivers.

The de Blasio Administration has, however, laid out a broader vision for improving the city’s transportation system in the OneNYC citywide plan, released last week (the plan’s ambitions spread far beyond transportation). The plan recommends deploying CBTC more quickly, the continued construction of the Second Avenue Subway, the conversion of Brooklyn’s Long Island Rail Road Atlantic Branch into Subway-like operations, and increased availability of bus rapid transit.

Most dramatically, OneNYC recommends that the MTA study the extension of the Subway south along Utica Avenue through East Flatbush and Flatlands, a roughly four-mile route that would, if built, include the first new Subway stations outside Manhattan since 1989 and the first in Brooklyn since 1956. In the interim, the MTA plans to implement a bus rapid transit route along Utica this year.

In the city’s collective imagination, a Subway extension along Utica is practically as mythical as the Second Avenue Subway; it’s an idea that’s been floated around for a century. De Blasio’s most recent plan doesn’t help much to de-mystify the proposal, since it includes no clear financing source for the project. But the plan does suggest at least studying it.

As shown in the following map, a Utica Subway would fill a significant gap in Brooklyn’s transit network, offering faster commutes on the city’s third-busiest bus line, the B46, which currently serves about 46,000 daily riders. It’s also a route that serves a relatively low-income area, meaning it would bring significant transit benefits to people who are already very reliant on public transportation and who deserve a hand up. It will provide an important boost in equitable access to transportation to a currently underserved neighborhood.

Drag vertical line from left to right to see images (if this does not work for you, view the article in a web browser).

But does the fact that no Subway line currently serves that section of Brooklyn mean that the MTA should prioritize investing in a new corridor there? Is this the right place to be investing?

As the map below illustrates, the area around stations along a potential Utica Avenue Subway line has a density of 38,000 residents per square mile within a half mile of stops, and a total density, including jobs, of around 44,000 people per square mile. At the national scale, that’s a very built-up environment; a recent comparison of Los Angeles and San Francisco, for example, extolled the fact that Central L.A. averaged about 17,500 residents per square mile. 

But compared with other existing Subway corridors and the proposed second and third phases of the Second Avenue Subway, the Utica corridor comes up short. The existing line under Nostrand Avenue serves a corridor that is, in total, 57 percent denser. And the future Manhattan stops serve neighborhoods that are almost six times as dense when including residents and jobs.

Density is an essential characteristic when determining the appropriateness of a corridor for new transit services; indeed, it is often used as a proxy for potential ridership levels. Urbanist Vishaan Chakrabarti’s book A Country of Cities, for example, suggests in order to support rail from a cost-benefit perspective, neighborhoods must be “hyperdense” and feature 30 or more dwellings per acre.* At the Brooklyn average of 2.8 people per household, that equates to 54,000 people per square mile, or generally the areas colored red or blue in the maps above. In other words, that’s far more than the Utica corridor on average, but pretty typical for areas along Nostrand or Second Avenues.

This metric suggests that a Subway line on Utica would provide too high a level of transit service for a neighborhood that is significantly less dense than many other parts of the city served by the Subway.

The question of whether this corridor is adequately populated to support a Subway extension is relevant given New York’s exceedingly high construction costs, which have now depressingly risen above $2 billion a mile for Subway lines. If the City has the opportunity to devote funding to the construction of a line**, it better make sure that it is investing in the project that can provide the biggest bang for the buck.

To put it simply, is it worthwhile to spend $8 billion on a Utica Subway extension when the second and third phases of the Second Avenue Subway, which would serve many more people, are not yet funded? One might argue that in fact New York needs both projects, but it’s hard to square that idea with the hard, cold fact that the MTA’s capital plan, which would mostly fund maintenance, is missing $15 billion.

The problem with adding a new route along Utica extends beyond the question of whether there is an adequate population to support the line. Indeed, given the mounting congestion on the Subway system, additional ridership from Utica—assuming riders who currently drive or take the bus switch to the train—would make the already-difficult crowding worse. Is that a policy the City should be pursuing?

Above all else, I contend that the City’s priority must be to find ways to relieve congestion on existing lines before adding to the problem with new ridership from new lines. One way to do that is to encourage transit ridership growth on the city’s bus rapid transit network, which, unfortunately, has not absorbed much of the city’s increasing transit ridership. Another would be to, as OneNYC suggests, significantly speed up the installation of CBTC. A third would be to convert the region’s commuter rail lines into higher-capacity rapid transit.

One way to add service to Utica without necessarily worsening existing congestion would be to add capacity elsewhere in the system. In 1996, the Regional Plan Association’s Third Regional Plan proposed linking an extended Second Avenue Subway under the East River to the exact same Long Island Rail Road Atlantic Branch that de Blasio’s plan would convert into Subway-like service. If that branch were then to split off down Utica Avenue, new passengers would do little to worsen congestion.

But even if the major goal of transit investments were to serve new parts of the city with Subway service, would you start with Utica Avenue, assuming the Second Avenue Subway were completed? The density map of the city, shown above, suggests otherwise; indeed, Jackson Heights in Queens is denser than Utica Avenue and a new line along Northern Boulevard, combined with some other congestion relief into Manhattan, would probably address more peoples’ needs than a line along Utica. The same could be said of a line on Third Avenue in the Bronx.

Given these facts, the concept of spending billions of dollars on a Utica Subway line becomes less and less appealing. Certainly if the City committed to upzoning neighborhoods along the route to ensure that the line would attract adequate ridership to justify its cost, the logic behind its prioritization would become less murky. The density of neighborhoods near existing Subway lines, of course, is in itself a direct consequence of the existence of the Subway network. And if the MTA were to find a way to somehow significantly reduce its construction costs, many more train lines would be possible within the same budget.

A reduction in construction costs, however, is the holy grail that American transit systems seem unable to track down. Funding for the second phase of the Second Avenue Subway, while supported in OneNYC, is hardly definite, and it’s not like the federal government is offering generous expansion grants at the moment.

Utica Avenue does deserve improvement in its transit service. The new bus rapid transit line planned for the route will speed up commutes. But missing from the discussion is any intermediary between buses and Subways—it’s as if the vocabulary of high-capacity surface rail has been excised from the minds of transportation planners in New York City. As I’ve written before, Brooklyn is filled with opportunities to provide fast, surface-running light rail at a cost significantly lower than Subway service and a capacity higher than possible with New York-style bus rapid transit.*** If more of New York deserves access to high-quality, faster transportation, we should be looking at options other than just Subway extensions.

* Chakrabarti adapts this estimate from Boris Pushkarev’s Urban Rail in America (1982).

** As the City did, through back-end means, for the 7 Line extension currently under construction in Manhattan.

*** Many bus rapid transit services in Asia and South America, for example, operate in highway or highway-like rights-of-way that allow corridor capacities at or above those offered by light rail. But the New York environment makes such corridors impossible (and undesirable) to implement.

Image at top: Utica Avenue Subway Station, by Flickr user Ed Yourdon (cc).

» Virtually every new metro or subway train purchased by transit agencies over the past ten years has been built with open gangways—allowing passengers to walk from one end of the train to the other. Except in the United States.

New York City’s Second Avenue Subway project, which in its first phase will bring transit service north from 63rd to 96th Streets in Manhattan, will provide many benefits for commuters, offering three new stations and much easier access from the Upper East Side to western Midtown. It will reduce congestion on the Lexington Avenue Subway (4/5/6) by as much as 13 percent—a boon for commuters on the single-most-used transit corridor in the country. And it will respond to the simple fact that New York City is growing quickly; it has added half a million people since 2000 and continues to expand.

But the Second Avenue Subway project has its issues—notably the fact that at $4.5 billion, it’s outrageously expensive given its 1.7-mile length. Given these construction costs, few projects of this magnitude are possible. So what alternatives do congested, growing cities like New York have to increase the capacity of their transit systems?

All around the world, cities investing in their metros—a term I’ll use here to describe systems like New York’s Subway, the Bay Area BART, and others—are choosing to include open gangways on their trains.* It’s a simple concept to understand: Basically, people who board a train are able to walk from one end of the train to the other without opening doors or stepping outside of the train.

Open gangways provide a number of advantages: One, they expand capacity by allowing riders to use the space that typically sits empty between cars. This added capacity means that a metro line can carry more people with trains of the same length. Two, it allows passengers to redistribute themselves throughout the train while the vehicle is moving, reducing problems associated with many people boarding in the same doorway, such as slow exiting times and poorly distributed standees. Three, it increases safety at times of low ridership by increasing the number of “eyes” in the train. There are no obvious downsides.

Open gangways offer passengers the benefit of an improved, less congested, and safer environment as compared to trains with individual cars, the standard you’re used to if you live in the U.S. And it’s no surprise that transit agencies all around the world are choosing open-gangway trains for virtually every new vehicle purchase. This is documented in the following map, where green cities represent places where the metro systems run at least some trains that are all open-gangway. Those that are red do not. Click on the map for a higher-resolution, larger version.

I used the World Metro Database to help me create the map below and the table at the end of this article, but the Database is out of date and, in some places, incorrect and as a result, I collected the information shown here one agency at a time. The vast majority of metro systems are investing in trains with open gangways.

Yet American transit agencies have ignored the concept. New metro trains have been or are being purchased in Chicago, New York, San Francisco, and Washington, among others, but they all continue to be built with individualized cars, with no open gangways. It’s as if the agencies simply have not gotten the message. Only Honolulu, which has a new purpose-built metro currently under construction, will adopt this technology. Perhaps the other agencies will get the message once that system opens in two years.

I wrote about this issue six years ago, interviewing representatives from New York and Washington transit agencies to ask why their new trains did not feature open gangways. The responses were anemic: In Washington, a spokesman told me that the agency had “no plans to change it just to change it,” as if the concept of open gangways was frivolous. In New York, I was told that open gangways would only be possible if “we have a budget for Research and Design for an entirely new subway car.”

Others have suggested that the handicap in the U.S. is that transit agencies have specifications that make them incapable of handling such vehicles. Some say that U.S. agencies need trains with short cars, but the Paris region features a commuter train with open gangways with cars that are shorter than even the notoriously short Chicago L vehicles (43’5″ versus 48′). Some say that the maintenance expense would be too high to transition to these trains (since maintenance facilities might have to be altered to handle cars that are permanently affixed to one another), but many of the European agencies, with metro systems just as old as those in the U.S., have been able to accommodate the trains in their facilities, probably with the assistance of the train manufacturers. Some suggest that these trains would be more expensive, but evidence suggests otherwise.**

London, which has resisted adding open gangways to its “deep tube” fleet (it has such trains already on its “sub-surface” lines) because of issues with tight curves, has recently come around to the concept. In its future metro vehicle feasibility study, London found that open gangways were not only possible, allowing walk-through trains, but that they would increase train capacity by up to 10 percent, while reducing train weight and energy consumption.

When I analyzed this subject in 2009, I didn’t realize the degree to which the world standard had shifted. 75 percent of non-U.S. metros now offer open-gangway trains in their fleets, representing systems as varied as the brand-new networks in China to the ancient facilities in Berlin or Budapest. The last time Mexico City, Madrid, Oslo, or Amsterdam bought a train with individual, separated cars was back in the 1990s. Even our compatriots just across the border in Montreal and Toronto have come around. Every major train manufacturer offers trains with open gangways off the shelf. What is holding U.S. systems back?

Back in 2013, New York’s Metropolitan Transportation Authority announced in its long-term capital needs assessment that “consideration should be given to” trains with open gangways. We’ve heard no more on this subject in the intervening time, despite some positive coverage of the news.

Yet the agency, like others around the country, has the opportunity to address some of its problems through the purchase of these trains. On the congested Lexington Avenue Line, which I discussed at the beginning, about 45.6 feet of each train’s 513.3-foot length is used up by the empty four feet between each car and the 10 feet reserved for the cabs at the center of the trains.

That means that, if the Lexington Avenue Line were transitioned to trains with open gangways, the line could gain almost an entire car-length of capacity on every train. That’s practically as much relief as the Second Avenue Subway will provide—at the cost of trains that would be purchased anyway.

Open gangways are hardly the end-all be-all of transit operations. They won’t guarantee better service or necessarily attract more riders. And they may not be able to resolve some issues, such as the fact that Washington’s Metro runs trains of different car lengths on each line.

But the fact that every U.S. transit agency—with the exception of Honolulu’s—has failed to adopt to this trend and has no plans to change, raises important questions. Just how much are the management of these transit agencies isolating themselves from world best practice? This is hardly an isolated case. The fact that transit agencies around the world are transitioning infrequent suburban rail operations into frequent regional rail services seems to be lost on most U.S. commuter rail agencies.

If the problem is simply a lack of knowledge, that’s no excuse given the existence of this website or Wikipedia or countless other sources. If the problem is petrified management, stuck in an older technological age and unable to try something new, staffers at those agencies should be working to convince them of at least the possibility of change. If the problem is some sort of U.S.-specific regulatory problem enforced by the federal government, let’s work to adjust it.

I’m skeptical that this technology is just “not possible” on historic U.S. systems; it’s been adapted to too many places around the world in all sorts of conditions for that to be the case. But if the problem is that transit agency management simply doesn’t care enough to adjust their operational standards to respond to improvements that can be offered to passengers, well… it’s time to kick the bums out.

* You could call trains with open gangways “articulated,” but this typically refers to a specific type of gangway, often where the truck (the bogies, where the wheels are) is right below the gangway. A traditional train would have two trucks supporting each car (a 10-car train would have 20 trucks), but an articulated train might have every two cars sharing one truck, such that a 10-car train could have as few as just 11 trucks, vastly reducing weight and energy consumption.

** For example, I compared two contracts conducted in the early 2000s with one metro manufacturer, Alstom. In 2001, Paris bought 805 metro cars (each 49.6 feet long, in open-gangway train configurations) for €695 million. In 2002, New York bought 600 subway cars (each 60.2 feet long, without open gangways) for $962 million. When converted to U.S. dollars (at the July 2001 rate of 1.16 dollars to the euro) and inflation-adjusted to 2002 dollars, the Paris contract was $820 million. This means that, per foot of subway car, Paris paid $20,535 and New York paid $24,200, despite the fact that New York’s contract included, as this article notes, lots of empty space!

Image at top: Potential future London Tube, from Transport for London. World map of metros based on world map base SVG by @F1LT3R of Hyper-Metrix on Wikipedia.

Edit, April 11: I updated values for Moscow, Kazan, Kiev, Kharkiv, Sofia, and Novosibirsk to reflect the fact that they do not currently have metros with open gangways.