getting around

Why It Costs So Much to Build Our Subways

An NYU study finds that most of the overspending isn’t where you think it is.

Photo-Illustration: Curbed; Photos: Shutterstock, Alamy
Photo-Illustration: Curbed; Photos: Shutterstock, Alamy

Our subways cost too much. Not just to run — though we certainly know that to be true — but also to build, which is one reason New York has added so few of them in living memory. In this century, New York has added two tunnel extensions and four stations: one on the westward extension of the 7 line and three on upper Second Avenue. The latter had been on the planning boards since 1919. We spent $1.5 billion per mile for it. Berlin and Paris pay about a sixth as much for more ambitious projects. Paris, in particular, is in the middle of a 120-mile, 68-station subway-and-commuter-rail project that is known for its overruns and budget excesses, and its costs are still nothing remotely close to New York’s. (That is particularly humbling given that it’s France, of all places, a nation not known for weak unions or extreme efficiency.) Phase two of the Second Avenue Subway, which will reach 125th Street, does not look as if it’s going to be much cheaper.

None of this is newly discovered information, but three researchers from NYU’s Marron Institute of Urban Management have dug deep into the data and come up with a precise assessment of just where and how the MTA overspends. Most of their methodology involves fine-grained comparisons with practices elsewhere in the world, especially in Sweden, Italy, and Turkey, all of which have useful data and spend (give or take) about 10 percent of what we do per mile of subway. The results of NYU’s Transit Costs Project are a little distressing because there isn’t one graspable, fixable issue; everyone with whom the MTA interacts, from contractors to other agencies to its own internal processes, adds to the cash-draining effect. The obvious answers that may pop into your mind (“It’s the unions,” say, or “New York is dense and complex”) turn out to be small pieces of the whole that, taken together, roughly triple our costs on any given project. Yet the problems are categorizable, falling into a few broad areas. (The MTA’s Janno Lieber commented on the study’s findings at a press conference last week.) I spoke to the study’s lead author, Eric Goldwyn, about the more surprising aspects of each.

Our new stations are way too big.

The stops on the Second Avenue Subway, like those on the mid-century IND lines, have enormous mezzanines above the tracks. The new stations are also much, much bigger than you might realize. The 96th Street station is more than two and a half times the length of its platform. By comparison, the other cities in the study build stations just 3 to 20 percent longer than their platforms: “They don’t extend, extend, extend — the most expensive thing to do,” Goldwyn says. The giant mezzanines are typically justified as a fire-safety move, ostensibly allowing platforms to be evacuated in a hurry, but they are nearly unique to New York, even though every subway system has fires. Add to that the cost of the buildings needed for ventilation equipment and evacuation. “The ancillary buildings go 70 feet down and another 70 feet up, so you’re building a 14-story mid-rise tower, and of course that’s going to be expensive, especially when half of it is in hard rock.”

Those stations also include a lot of back rooms and service areas, way more than other cities. “Back-of-house space,” as it’s called, is surely convenient: You can store extra lighting or equipment and provide staff changing rooms and the like. But too much of it is wasteful and terribly expensive. Goldwyn notes that the 72nd Street station on the Second Avenue Line had higher costs per cubic meter than any other, and it’s no coincidence that it has more back-of-house space, too. “Some rightsizing of the stations, moving the technical spaces out of the underground space, moves the dial on costs.” (Paris and Istanbul stick most of it upstairs.)

One reason we build so much underground is pure territorialism. While interviewing MTA employees and consultants, the authors learned that each constituency — hydraulics, track maintenance, and so on — demanded its own space, partly “because each user group bears responsibility for cleaning and maintaining its own room; thus, how would those responsibilities be distributed if multiple groups shared a room?” New York ended up dynamiting spaces out of bedrock rather than impose the solution worked out by every multi-child family in history.

The oversize stations more than doubled the cost of the project just on their own, according to the study.

Their designs vary too much.

Some of this back-of-house bloat will be addressed in the next set of subway stations to be built, where spaces will likely be consolidated and shared. But that brings up another problem, which is that too much custom work goes into each one. The three stations built on Second Avenue “used two different escalator contractors and have a different number of exits, crossovers, and elevators,” thus forgoing economies of scale, the study explains. We do not necessarily want cookie-cutter stations — after all, the City Beautiful movement gave us IRT stops with distinctive finishes — but, as Goldwyn says, “in a utilitarian system, the industrial process can be good. In Copenhagen’s Cityringen, all the stations use the same form, and they’re beautiful. In Stockholm, often considered the most beautiful in the world, they’re relatively standardized. In a dense urban environment, it’s going to be harder to do that, but it’s not impossible.”

Other agencies want a piece of the action.

City and state agencies defend their turf and budget ruthlessly, a situation Mayor Eric Adams is trying to address. It goes way beyond poor communication: When the MTA needed public space to store construction equipment on Second Avenue, it used a section of a small playground at 96th Street — and the Parks Department charged the MTA $15 million in rent to borrow it for a few years. “When I talked to people in Istanbul,” Goldwyn says, “they had no comprehension of why you couldn’t use public land. All the benefits accrue to the City of New York; for the 7 extension, the city donated the land.” That kind of nonsense has not stopped: “If you look at phase two, MTA is buying parking lots from the city for millions of dollars.” (Some of this is surely owed to the MTA’s being a state agency, not a city one, and it’s worth remembering that Bill de Blasio and Andrew Cuomo were barely on speaking terms.)

Many other entities seem to regard a subway project as a chance to extract a few infrastructure updates without paying for them. When the MTA dug up Second Avenue, the Department of Environmental Protection got the transit agency to pay for a replacement water main, rather than doing it out of DEP’s own budget. (It tried to introduce an expensive upgrade in the process, and was negotiated back down.) Why did the MTA agree to pay? “They need to get underground,” Goldwyn says, “and there’s pressure to not slow things down. So if you just did—I’m going to make up numbers here—a $5 billion project, and escalation costs are 5 percent a year, that’s $250 million a year. You might make the calculation that, as long as it’s less than $250 million, it’s worth doing. They can’t just do what they want to do— they need to get a permit from the City of New York to open the street.” There are similar conflicts within the MTA itself: The capital-construction division of the MTA found itself pushed around by the larger New York City Transit arm (which, for example, was the source of many of those demands for back-of-house space). Or if—and this reportedly happened, more than once—Con Edison agrees to turn off the power under the street but the utility’s crew doesn’t show up on the appointed day to actually do it, dozens of expensive construction workers and their equipment sit around idly.

Politicians can help, but often they get in the way.

The obvious answer to something like that playground fee is to go over everyone’s head. The governor or mayor can simply say, “Give them the space.” But calling upon political power carries with it a separate risk. As Goldwyn put it, “You want politicians meddling — macromanaging — to say, ‘This is a project that the MTA is doing’” and demand that everyone else get out of the way. Instead, what sometimes happens is that elected officials make particular, inefficient demands. They’ll promise some group of constituents an extra station, say, or a hard deadline, and then the designers are stuck even though they know it’s a terrible idea that will add costs. “What you don’t want,” Goldwyn says, “is meddling where interest groups want it. Sometimes you get politicians saying, ‘I’m committed to this alignment,’ and then they can’t back down. They’re boxed in. What we’ve seen abroad is that there’s a real master plan that the mayor approves, but the mayor is not out there grabbing anyone by the lapels, Lyndon Johnson style. We need politicians to say, ‘This is important,’ but also, ‘We’re going to get out of your face and let you go.’” In short, they need to defend the ability of career subway builders to build the subway.

Labor is expensive, but so is outsourcing of expertise.

Sure, union labor costs a lot. Our crews are big, our work rules are generous, our wages are good. Our projects also tend to be well stocked with white-collar management. “But we would argue,” says Goldwyn, “that, more recently, there’s more going on here. There are other issues, which we refer to as procurement — like contractors charging a premium because the MTA is hard to work with. Talking to contractors and vendors, who make their money that way, they say, ‘Ugh, they’re the worst to work for,’ even though it’s the MTA that pays for their yachts.”

“Or you aren’t using your labor productively,” he continues. “If you have high labor costs, then really treat labor as this precious thing. You want to maximize its utility, not have construction timelines that take forever.” And we do not consistently do that, even if Con Ed does show up on time. “Compare the digging to get the tunnel-boring machine underground,” he says. “In New York, it took three years. In Italy and Istanbul, it took a year. When you interact slow-moving processes with high-cost labor, you’re going to get exponential increases — and I talked to some people who did that work and they said absolutely they can do this quicker.”

Another problem is the gradual outsourcing of expertise to expensive consultants. “The MTA used to have a capital-management arm that designed 90 percent of their projects, and I don’t want to say it was a golden age, but they had 1,600 people on staff. In 2011, when MTA was building Second Avenue and the 7 extension and East Side Access and Fulton Street, MTA Capital Construction had 124 employees. They’re spread extremely thin, running with their heads cut off, trying to put out every fire. There’s a logic — a lot to do with pensions and that sort of thing — and if you’re building one thing in the next 50 years, then, sure, hire a consultant. But if you’re building generations of projects and your core business is maintaining a capital plant, why not have people on staff who know all that stuff?”


The funny thing about reading this report is that you (or at least I) come away not furious at the MTA but cognizant of its plight. Many of these problems do not stem from horrible corruption, at least not in the “guy slipping an executive an envelope full of cash” sense; they’re more the result of clashing from often old practices, many of which aim to make one thing work better and have the effect of making others nightmarish. Goldwyn, perhaps surprisingly, agrees. “You know,” he says, “I interviewed a former head of DEP, and when he was running the agency — this was like 2010 — he couldn’t get a digital copy of where the utilities are. They’d bring over paper blueprints. It’s parochialism, not wanting to share information, or really holding on to what you have, squeezing every ounce of power from it. That whole story is super-interesting and fascinating, and it made me very sympathetic to the MTA. And that’s not something I expected. This is written very sympathetically to the MTA — although they would disagree with that.”

Why It Costs So Much to Build Our Subways