Why are American rail projects so costly? The initial results of an ambitious project by three researchers at New York University’s Marron Institute suggest one culprit hiding in plain sight: pointlessly fancy train stations.

On December 9, the NYU researchers—Elif Ensari and Eric Goldwyn, trained as urban planners and architects, and Alon Levy, a mathematician and author of Pedestrian Observations, a widely read public transit blog—held a panel discussion on the project’s first case study: the Green Line Extension (GLX) built by Boston’s public transit authority, the MBTA. In principle, GLX is simple. It extends an existing light-rail line 4.3 miles into Boston’s inner suburbs, mostly in a trench that already holds commuter rail tracks. But GLX’s costs ballooned after planning began in 2006, reaching $3 billion in 2015—more expensive per mile than most subways.

Much of the cost bloat, the NYU researchers found, was for bells and whistles on the line’s seven stations. The first design for GLX envisioned basic stations much like existing Green Line stops in dense parts of Boston, comprising little more than a pair of concrete platforms with wheelchair ramps and costing about $500,000 each. But in later designs, the researchers write, “These simple stations morphed into bespoke neighborhood icons with headhouses, redundant elevators, escalators, personnel rooms, fare arrays, larger footprints, and additional landscaping and street grading extending beyond the stations.” These larger stations cost far more, especially in steel, concrete, and electrical and mechanical work. An internal MBTA review estimated $410 million in total station costs—more than 100 times the initial estimates.

Boston isn’t the only city that overengineers its rail stations. Levy once compared New York’s recent Second Avenue Subway with a subway extension in Paris. Simple tunnels in New York were 1.7 times as expensive as in Paris due to overstaffing of tunnel-boring machines, but New York’s stations were also larger than their Parisian counterparts and 6.5 times as expensive. One engineer in the San Francisco Bay Area has also pointed out that several suburban Bay Area Rapid Transit stations are overbuilt. For instance, Milpitas Station, part of a recent, expensive extension to San Jose, is a pointless hangar-like structure occupying more than an acre. A functional station would need just escalators from the street to the platforms, a simple bus shelter, and possibly a pedestrian bridge.

Why do agencies design bloated stations? Force of habit, partially. In the panel discussion, Levy hypothesized that many de facto American subway construction standards come from the IND, a subway network built by the New York City government in the 1920s and 1930s to compete with the city’s two private systems (which the city government bought and merged with the IND in 1940). The IND was built to very high—and in some ways excessive—engineering standards, especially in station design. For instance, IND stations have mezzanines between the platform and street levels that run the full length of the platform, requiring deeper stations with more excavation. (Older stations have short mezzanines or none at all; trains run just below street level and passengers need to ascend to the street to switch platforms.) This overengineering made the IND far more expensive than contemporaneous projects such as London’s Victoria Line.

Another driver for station bloat, which played a large role in GLX, is community involvement. In reaction to postwar urban-freeway construction, which threw hundreds of thousands of residents out of their homes with minimal due process, state and local governments began requiring infrastructure projects to pass through protracted community meetings and public comments. Community meetings are supposed to make planning more democratic, but they’re often scheduled at inconvenient times for workers and are thus dominated by unrepresentative segments of the population—especially wealthy, change-averse retirees who demand extras such as costly architecture and landscaping.

Planners are supposed to resist unreasonable requests, but it’s easier for them to “push the yes button” to placate a potential enemy, as one MBTA manager quoted in the NYU case study put it. GLX, for instance, was redesigned at a late phase to accommodate activists’ demands that it include an extension of a suburban bicycle path, requiring tens of millions of dollars of extra spending on wider retaining walls. New York built the Second Avenue Subway by removing soil through a narrow access shaft rather than cutting open a street from the surface—an approach that avoided a few temporary street closures but added several years to construction.

“Pushing the yes button” exemplifies a general problem with government: bad policies with obvious benefits for the few and diffuse costs for the many are easier to pass. One possible solution: transit agencies should make tradeoffs more explicit. Planners could propose multiple versions of projects: one that spends extra on fancy stations and other nice-to-haves, and a more economical version that rolls the savings into better service and faster construction. The public could even vote on these alternatives: in Switzerland, one of the world’s leaders in low-cost construction, most transit investments are made by referendum, encouraging planners to cater to the whole public, rather than just vocal minorities. Ultimately, planners need to be enabled to make economical decisions that benefit the entire population, overruling local busybodies if necessary.

Photo by Yana Paskova/Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next