In 2006, Alon Levy moved to New York to attend graduate school. Levy took the subway to get to class and events around the city, as most New Yorkers do. But unlike most New Yorkers, Levy is the type of person who will instantly become inquisitive about the most seemingly mundane issues. For example, when told there are people waiting at the elevator to get to an event, Levy starts crunching numbers in their head about elevator capacity.
So naturally, Levy took an interest in the subway— especially the new Second Avenue subway, when construction began in 2007. Levy heard this first phase of the project with three new stations and 3.5 miles of tunnel would cost $3.8 billion. (It actually ended up costing $4.5 billion in the end.) “OK, really high cost,” Levy recalled thinking to themself during a recent panel discussion at New York University on transit construction costs. “But this is how much it costs to build a subway, right?”
This idle thought led Levy to poke around and find out how much it actually costs to build a subway. They recalled finding that in Tokyo, around the same time, the government had implemented a moratorium on subway building because costs had risen too much, to approximately $500 million per kilometer. At the time, New York was building the Second Avenue subway for an estimated $2.5 billion per mile, or well over a billion dollars per kilometer. Paris was building subways for $250 million per kilometer, some 10 times less. The more Levy looked, the more they found New York wasn’t just paying more for subways than every other city, but many times more.
Most troublingly, Levy found, nobody seemed to know why this was the case—why New York specifically and the U.S. in general are so expensive to build in. “I keep asking New York rail fans why,” Levy recalled, using the term for subway enthusiasts who often have encyclopedic knowledge of the system, “and I never get satisfactory answers.”
Some 15 years after these initial inquiries, Levy now has an answer. It is a dastardly combination of:
- Hiring contractors to do the work in a manner so bizarre it almost seems intentionally designed to drive up costs
- Hiring consultants to design and manage projects rather than having the staff to do so in-house (or not having the necessary staff expertise to manage consultants in a way that keeps costs manageable)
- Haphazardly coordinating complicated work with local utilities
- Viewing infrastructure projects as job programs and therefore hiring more workers than needed (union and non-union, blue collar and white collar)
- Over-designing projects (stations in particular)
- And, on top of it all, having local politicians who micro-manage, slowing down planning at best; changing projects in a way that makes them more expensive and less useful at worst
Levy has these answers because after writing about the problem for the better part of a decade on their blog Pedestrian Observations, Levy has, alongside a team of researchers at NYU’s Marron Institute of Urban Management—including Elif Ensari and Marco Chitti—spent the past three years crunching massive amounts of data, interviewing hundreds of experts all over the world, and conducting in-depth case studies on transit projects in Sweden, Istanbul, Italy, Boston, and New York. The group plans to release its final report this month on what the U.S. and New York in particular can do to start building transit at least on par with the rest of the world. It is part of a growing body of literature—the Eno Center for Transportation launched a similar project around the same time—examining the question of how the U.S. can be more productive with its transportation infrastructure dollars.
(Full disclosure: I moderated the panel discussion on the group’s findings on October 26. I did not receive any compensation other than half of a roast beef sandwich from a lunch provided to all attendees afterwards.)
It’s important to emphasize why this matters. All too often, politicians say they want an efficient, effective government but do the opposite. Two transportation-related examples come to mind. One is Maryland governor Larry Hogan’s failed experiment with the Purple Line, a light rail that will connect Bethesda and New Carrollton in the D.C. suburbs, built under a public-private partnership model ostensibly to save money; it has been such a disaster that the first contractor abandoned the project midway through construction and it is now set to cost $3.4 billion, or 75 percent more than budgeted. The other is in New York, where in 2019 then-governor Andrew Cuomo launched a “transformation plan” created by a consulting firm to “simplify a complex and inefficient organization,” the MTA (still) says on its website. The most noteworthy “accomplishment” was hiring six-figure executives who promptly left with golden parachutes; once Cuomo was gone the transformation team was shut down. There are countless more examples where that came from.
Although no U.S. city builds subways cheaply—or any urban transit, regardless of whether it goes on tracks or roads underground, on the ground, or elevated above ground—New York is the worst of the worst, the most expensive place to build mass transit on the planet. Each new project breaks records. First there was the Hudson Yards extension, then Second Avenue Subway Phase I, then East Side Access, and now Second Avenue Subway Phase II, each more expensive than the last on a per-mile basis, and all world-record breakers. Second Avenue Subway Phase II will cost a whopping $4.2 billion per mile, or about $2.2 billion per kilometer, almost doubling the then-record costs of Phase I.
With costs so high, U.S. cities cannot build as much new transit as they’d like, which restricts access to new jobs, housing, and locks in transportation emissions from driving. And for New York’s Metropolitan Transportation Authority specifically, those astronomical costs, increasing with each successive project, requires borrowing lots of money to build anything. Borrowing money means paying off debt, and those debt payments come from the same budget it uses to run buses, trains, and do routine maintenance. While the exact structure of each city’s transportation agency, funding, and governance is different, the basic underlying facts are the same. They all borrow money to build megaprojects that cost billions of dollars, and the more money they have to borrow, the more debt they’ll have to pay.
Therefore, there’s a tradeoff between wasteful megaprojects today and fewer trains, more delays, and higher fares tomorrow. And after decades of overpaying for so many transit projects in New York, that “debt service” is now 17 percent of the annual budget, meaning almost one out of every five dollars the Metropolitan Transportation Authority gets from fares, tolls, and tax revenue goes not towards running better transportation, but paying debt it accrued by being the most expensive place to build a subway on Earth.
Perhaps the most noteworthy conclusions of the years-long research is what Levy and their fellow researchers did not find. There is no magic bureaucratic structure for a transit agency to be run well and turn bad administrators into good ones. There is no magic funding mechanism that incentivizes transit agencies to be more fiscally responsible. There is no law or set of laws that can make transit agencies good. And the countless accountability agencies usually with “inspector general” in their names to detect waste, fraud, and abuse routinely miss the big picture, catching small-time crooks while ignoring how billions of dollars can simply go “poof” in the night as long as it’s attributed to a megaproject. Or that blaming unions misses the much more specific ways in which American labor unions add to labor costs while unions in other countries do not. What does have to happen, first and foremost, is political and transit agency leadership has to actually give a shit.
“Every theme comes back to institutions and decision-making processes and how they influence each of the aspects we study,” the group found. “Practitioners in high-cost environments must be ready to learn from their low-cost counterparts, reflect on internal practices and standards, and ask both ‘Why do we do things the way we do them?’ and ‘How can we adopt practices from more efficient places?’”
The good news is “lower costs can be achieved” without requiring “large-scale legal changes.” the research group concluded. The bad news is it requires “reassessing decision-making processes, institutions, the nuts and bolts of project agreements, and getting key decision makers to support projects rather than delay difficult decisions.” Or, to put it another way, it requires our politicians and their appointees to unite around a common good rather than bicker and intervene to promote their own particular brand of provincialism, itself a reflection of the fact that most voters do not have this issue on their political radar.
An optimistic example here is Italy, as Chitti explained during the panel discussion, which has gone from being a high-cost country in building transit to a low-cost one. In response to major corruption scandals in the late 1980s and early 1990s, a movement grew for good-governance reforms which culminated in a major change to the laws over how to spend money on big contracts. There were many aspects to this reform, but a key one, according to Chitti, was to require a publicly-available list of itemized costs.
And by itemized, Chitti means really itemized. As in, “One cubic meter of concrete of this particular characteristic in these circumstances, with a lot of other parameters, should cost, like 70 Euros per cubic meter. I’m just guessing a number, but this is something you can use as a reference.” This kicks off a positive feedback loop. Bidders know a baseline cost they are expected to meet. Competitors can try and beat that. If changes have to be made during the project—a huge reason costs keep skyrocketing in the U.S., as each change has to be negotiated on an individual basis and agencies will default to paying whatever contractors want to get the project done or engage in a years-long fight that will waste time and increase costs through delay—there is already an agreement in place about what things cost so there is no need to argue about it. And for future projects, the process doesn’t have to be repeated all over again. Plus, with an itemized cost baseline to work from, it frees transit agencies to evaluate bids more on technical merit rather than a “race to the bottom,” as Chitti put it.
None of this is done in New York. The MTA has its own internal cost estimates it regards as trade secrets; contractors almost always come in 20 or 30 percent higher, and then make more money on changes once the project is underway. How much everything costs beyond top-line figures is shrouded in mystery, not just from the public but from other companies that could potentially do it for less.
Although there are some relatively simple things agencies can do to lower costs like itemizing, building up in-house expertise, and exchanging ideas with other countries that do things better, the final report is not about The One Simple Trick to Save American Public Transit. A problem this big and this long in the making—New York has had high transit construction costs since the 1930s—does not, by definition, have quick fixes. And Chitti emphasizes that even the success story in Italy took time. Itemized cost reforms were first discussed in the 1970s, but it wasn’t until the 1990s they got passed. And it takes about 20 years for the virtuous cycle to really have an impact.
The final report highlights six areas for improvement with the kind of headings that one does not exactly put in a headline meant to drive clicks—”civil service and internal capacity” and “procurement and risk allocation” being two examples—but all come back to a central theme, what the report calls “institutions and decision-making.” Crucially, the report’s central finding speaks to issues not just with transit in the U.S., but all kinds of institutions. We need better political and institutional leadership which actually cares about costs, about treating this metastasizing cancer on American society that is hobbling our ability to build anything decent. And in order for that to happen, voters need to start caring, too.
“We need to shine a light on this and talk about it more,” said Eric Goldwyn, a longtime collaborator of Levy’s and program director at NYU Marron Institute. “I think that’s how you start.”