Open gangways, tunnel rehabs, and making better decisions

People are dying walking between subway cars, which an existing technology the MTA chose not to test 20 years ago would have prevented. How can the MTA start making better decisions?

Thanks to everyone who submitted mailbag questions and/or put their names in the ring for a tote bag. I only had three totes up for grabs, so apologies if you missed out. In any event, it’s not too late to send in some questions for the mailbag if you’d like by responding to this email.

I had written this special edition last weekend and was planning to send it out then until the whole L thing happened and we had more newsworthy matters to discuss. But I think it’s still useful to ponder/think about, so here it is.


Four people died in December walking between subway cars, most recently someone on a Q train. Yes, it’s a bad idea to walk between cars and people shouldn’t do it, but lots of people do lots of things they shouldn’t. A few days ago, a beggar came through the car I was in. He didn’t seem to be all there, mentally speaking, and when he got to the other side of the car, he opened the door and stood between the two cars for a solid 20 minutes.

As @MaketUrbanism pointed out, there is an internationally recognized solution to this problem: open-gangway subway cars, which allow people to walk from one end of the train to the other without going through any doors. The Q uses R160 model cars, which were purchased in 2002, one year after Paris bought 805 open-gangway cars.

Interior of an open gangway train in Toronto. Photo credit user Raysonho under Creative Commons 1.0

At this point, New York City is one of the few major subway systems in the world without open gangway cars, an obvious point to anyone who has done any semblance of world traveling over the last decade. Open gangways not only eliminate this needless risk but, importantly, also add around 5-10 percent capacity to each train because people can stand in the space currently between cars. I suspect the capacity improvement would be on the high end for the subway, because passengers are poorly distributed throughout the length of the train, especially on the Lexington Avenue express (4/5) line where stairways often put passengers in the same places on consecutive busy platforms. With open gangways, passengers can maximize use of the available space.

The MTA did order some open gangway cars in their latest car order about a year ago, but only 20 for a “pilot” program. When the R211s were being showcased at the Hudson Yards station I asked the MTA why the open gangways are only a small pilot program and not the whole 535-car order. A spokesman replied that they considered open gangways “new and unprecedented.”

This is, in the most technical sense, accurate for NYCT, but only in the way that the first time I go to a yoga class it will have been “new and unprecedented” because I have never personally done yoga before. Even worse, NYCT can only claim this technicality because they opted against testing open gangways 25 years ago.

In the 1990s, NYCT ordered a two-set prototype from Kawasaki explicitly to test out new technologies in advance of a huge car order. Kawasaki proposed one of those models be open gangway, but NYCT rejected the idea because, as an NYCT spokesman told transportation researcher Yonah Freemark in 2009, “due to the impact it would have had on the project’s budget and schedule.” In other words, they made an overly conservative, risk-averse choice that had both foreseen and unforeseen consequences decades down the road. NYCT has purchased 3,250 cars since then, or half the existing fleet. If the open gangway pilot had been successful—and there’s little evidence to suggest otherwise, given its international popularity—it’s likely most if not all subsequent orders would have been open gangway. Imagine if, for example, the 4/5 lines had about eight percent more capacity today!

Freemark revisited the topic in 2015, by which point open gangways had become more or less standard around the world but NYCT still had not purchased any. He found full-train open gangways were available on 75 percent of non-U.S. Metros, but precisely zero in this country (Honolulu is getting some soon). And no, there’s no legal or trade law reason for this. Nor have I ever heard a technical specification or operational reason to suspect open gangways might run into issues here, beyond so if one car smells bad does that mean all the cars will smell bad? , a truly profound case of New York exceptionalism, as if we are the only big city in the world with smelly and/or inconsiderate transit riders.

(Of course, it’s worth dwelling on that no American metro system has adopted open gangways while the rest of the world has. This isn’t purely a New York problem, although the NYC subway is the only American mass transit system with serious capacity issues and relied on to the extent of its international peers.)

Meanwhile, as the MTA tries to figure out how it can possibly discourage people from walking between cars using stickers and announcements, they will run their “pilot program” for open gangways a bunch of years to find out if this thing the rest of the world does could work in our exceptional city because it opted not to test that very thing a generation ago. If it concludes open gangways can work, it will still be another 40 or 50 years in all likelihood—if not longer, as we’re currently finding out with the R32s that can’t be retired because the R179s are delayed and broken—until the closed cars recently ordered are finally taken out of service.

This is what conservative—or just straight-up bad—institutional decision-making looks like and how it rears its ugly head decades later. This is the system, the bureaucracy, that needs to be overhauled. It it also where one might be thinking: aha! This is exactly the type of mindset the Governor is trying to correct with the L shutdown reversal! Indeed, Cuomo may have been a driving force pushing the MTA to do even the small open gangway pilot.

Does a governor jumping in and overriding this process solve the problem? In the short run, perhaps (although you know how I feel about the L shutdown process specifically). But definitely not in the long run. The Governor can’t—and shouldn’t—be running the MTA. He’s got a whole state to worry about. Furthermore, micromanagement from multiple degrees above the top of the org chart is rarely a recipe for success.

If nothing else, the L shutdown debacle has brought a very important question outside of transit circles and into everyday NY conversation: how does the MTA make decisions? The key going forward will not be figuring out how the governor can make more of them, but rationalizing the MTA’s governance structure to match reality. The MTA uses billions of dollars in tax revenues to provide a public good, which is a pretty solid definition of a government service. The authority needs to ditch the structural remnants geared towards self-sufficiency and have its governing structure proportionally represented by the citizens who use it—yes, this is an argument for greater city control—while creating a clear chain of accountability regardless of who controls it. Without these things, we’re unlikely to ever get consistently better decision-making, whether it’s about a tunnel rebuild affecting hundreds of thousands of people or a small pilot program nobody thinks is a big deal until 20 years later when it’s entirely too late.