Home View from Underground No one is cutting 24-hour subway service, but let’s talk about it anyway

No one is cutting 24-hour subway service, but let’s talk about it anyway

by Benjamin Kabak

The mayor called 24-7 subway service a ‘birthright’ as the RPA’s Fourth Regional Plan kicked off a debate over the MTA’s approach to modernization.

Every generation or so, the Regional Plan Associate releases a major vision for the New York City metropolitan area and the ways its residents travel. The third Regional Plan, released in 1996, included calls for the Second Ave. Subway, East Side Access, and a streamlined transit hub at Fulton St., projects we know and love (or love to hate). Not everything becomes a reality — the 1996 plan also included calls for a one-seat ride from Lower Manhattan to JFK Airport and the Triboro RX circumferential line — but the plans serve as a blueprint for years’ and decades’ worth of discussions. So when the RPA unveiled its Fourth Regional Plan last week, the moment was something of a watershed for the next few decades’ worth of transportation plans.

Or at least it should have been, but one part of the RPA’s Fourth Plan stole the headlines. In it, the RPA may have proposed, to some degree or another, curtailing some or all 24/7 subway service. It was a vague, off-handed mention that that should have been clarified before the plan saw the light of day, but it cut at something New Yorkers believe to be sacred. Even if only approximately one percent of subway rides occur overnight, you can pry our 24-hour, seven-day-a-week subway service from our cold, dead hands. But as complaints about subway service the whole rest of the time mount, something has to give.

The spark to this fire was a brief mention of a plan to close, well, something at some time that arose on page 120 of a 374-page report [pdf] in the section on “accelerate the adoption of modern signaling systems.” The MTA’s need to quickly replace its collapsing signal system is hardly controversy; this paragraph was:

Guarantee track access and extended work windows. Track work is complicated and expensive on a 24/7 system. Closing the subways on weeknights and/or for more extended time periods would create more opportunities for track installation and testing of the equipment—and reduce costs. Only 1.5 percent of weekday riders use the system between 12:30 am and 5 am. The overwhelming majority of people who ride the subway during the daytime would benefit from the better, more reliable, cleaner and better-maintained system that weeknight closures allow. Of course, whenever lines are shut down, the MTA will need make sure that riders are not left stranded. New bus service should be provided to mimic subway service on traffic-free streets, and with shorter waiting times than today’s overnight subway service.

Now, it’s not immediately clear what the RPA is proposing here, and their comments surrounding the plan, including a subsequent blog post, did little to offer clarity. They say they want to “close the subways” for “more extended time periods” to allow for concentrated periods during which workers can access tracks. It’s indisputable that the current system, which allows for track access during limited overnight shutdowns, is a barrier to a quick signal replacement effort as the MTA itself believes it cannot complete a wholescale signal replacement effort in less than 40 years. But the scope of work completed during FASTRACK — night shutdowns, in which workers are afforded at most 4-5 hours of access — is limited to cosmetic repairs and track clean-up efforts. Replacing a light bulb isn’t the same as replacing an entire signal system.

Since the RPA mentioned the paucity of overnight riders, everyone latched onto this idea as a bad one put forth by a think tank insulated from the reality of low-income workers who rely on the late-night subway service, as bad as unreliable as it is, to get from work to home. Officer cleaners and late-shift healthcare aides can ill afford to lose their rides home. The outcry was immediate.

The mayor sounded an alarm and New York exceptionalism at its finest. “I’m a New Yorker, ” he said. “Twenty-four hour subway service is part of our birthright. You cannot shut down the subway at night. This is a 24 hour city.” (Of course, other 24-hour cities have more reliable late-night options via buses which are better suited for late-night ridership volumes, and some are only now introducing limited 24-hour service a few nights a week, but I digress.)

One City Council member is considering introducing a bill mandating overnight subway service be maintained if the MTA wants city funding. The law of unintended consequences is screaming in protest as this mandate could inhibit the MTA’s ability to do any work, let alone large-scale capital work it seemingly cannot do now.

Even MTA Chairman Joe Lhota pushed back hard as he seemingly objected to something the RPA wasn’t even proposing. “I believe a permanent closure of the entire subway system every night is a bit draconian,” he said. “The MTA has successfully been closing certain subway lines in evenings and on weekends as needed for maintenance and repairs. A permanent closure, I fear, would be inappropriate for the ‘city that never sleeps.'”

Had the RPA been more careful in its initial release, no one would have been talking about a permanent closure, and it’s not clear to me that the RPA intended to infer that a permanent closure was even on the table. We can unequivocally say that a permanent overnight closure of the subway system shouldn’t be on the table. The MTA doesn’t have enough maintenance workers to make this worthwhile, and there is no real underlying need to stop 24/7 service.

But to improve subway service at all other times, it may be time to consider line-by-line shutdowns for extended periods of time so that the MTA can make modernization a reality in the next 15 years rather than the next 40. My proposal comes with many conditionals and is modeled on the L train shutdown and what I believe to be the proper mitigation plan. Close entire lines, or certain discrete segments of them, for the number of months needed to make all repairs and replacements to the signal system. Provide adequate advanced notice and adequate replacement service (via bus bridges and increased service on nearby lines) and institute a night bus service as robust as the one in London. If this is sold as a short-term pain that’s necessary to deliver long-term gains, New Yorkers will accept it and plan around it. At this point, considering the state of the subway system, do we have another choice?

Last Thursday, RPA Chair Scott Rechler made just that point as the brouhahah over his organization’s plan developed.

Some late-night workers voiced their support too.

Of course, this is where the MTA’s credibility gap is a big negative. We don’t know what the L train mitigation plan is yet, and DOT and the MTA will have to collaborate on surface-level replacement service to a degree not seen in recent years. Worse though, New Yorkers don’t believe the MTA can complete work on time or provide adequate replacement service, and riders fear the MTA won’t restore service that is lost to short-term cuts. Based on the agency’s recent track record, I don’t blame anyone for this skepticism.

Yet, I’m left with the feeling that options are limited. Normal service these days is an unreliable toss-up of failing signals and endless delays, and experts expect unreliable transit to have a negative impact on NYC’s economy. If Option A involves short-term 24/7 shutdowns with the promise after of much better service and Option B involves muddling through until the entire system collapses, I’ll take A. How we get there from here involves careful messaging and a reasoned debate over the past way to fix the NYC subway system. The RPA’s plan, and the outcry over one vague paragraph in an otherwise thorough document, was not the way to start the discussion.

You may also like


Aaron M. Renn December 4, 2017 - 9:41 am

I agree that line shutdowns can be a good model. Chicago has done it multiple times successfully, albeit in a much lower ridership environment.

But it shouldn’t be just about signals. If you’re going to do a line shutdown, all important systems should be overhauled or replaced – power, tracks, etc. Stations should be done where possible as well. If you’re going to close things down to overhaul the line, then overhaul the line. Otherwise you have to do it again the next time some system needs upgrading.

AMH December 11, 2017 - 4:22 pm

Speaking of 24/7 shutdowns, it looks like the MTA is taking advantage of the (relatively low-ridership) holiday week to do a major shutdown:


Larry Littlefield December 4, 2017 - 9:57 am

“If Option A involves short-term 24/7 shutdowns with the promise after of much better service and Option B involves muddling through until the entire system collapses, I’ll take A.”

However, those in charge are over the age of 60, expect to be retiring and no longer on the subway eventually, and don’t give a damn about the generations to follow. They’ve been taking B for 30 25 years.

And not just at the MTA.

Did you hear that the federal government is going to pass yet another huge tax cut for the rich. One that will automatically expire, leading to a huge tax increases, a decade from now once all the members of Generation Greed are finally out of the workforce.

Not put that way in the media you didn’t. Yet that is what the proposal actually is. No one dares to criticize it, or anything else, no generational equity grounds. Because Generation Greed not only feels entitled to pillage. It feels entitled to not be criticized for doing so, so they don’t have to feel bad about it.

Daniel Lafave December 4, 2017 - 10:39 am

“But to improve subway service at all other times, it may be time to consider line-by-line shutdowns for extended periods of time so that the MTA can make modernization a reality in the next 15 years rather than the next 40.”

I’m just not convinced that works like the Flushing Line Signal Modernization would have been any quicker if they had shut the line completely. Is there any information to show that it would have been? The work is performed on a section of track at a time. They have done the work by skipping stops and running on the express track while doing local track work. Even if there were speedup, would it merit stranding 800,000+ riders a day?

tacony December 4, 2017 - 10:48 am

I don’t understand this debate. The MTA already does overnight shutdowns to work on the tracks. For a while recently it was being branded and publicized as FASTRACK, but both before and after, they have been doing this regularly, along with shutting down lines for entire weekends. This debate seems to be based on some alternate reality where all lines run normally all night. Even when there’s not work going on, the MTA already reduces service to the point where it’s barely useful. Anyone with money takes cabs and Ubers during the overnight hours. Almost all trains run local-only, and most are cut to 3 trains an hour. 3 trains an hour is a lot closer to zero trains an hour than it is to good service. You want to make a trip that involves a transfer? Your travel time can easily double what it would be during rush hours. It’s mostly only poor people who have to put up with it.

The idea that 24/7 service is why the signal system is falling apart is total nonsense. Do people honestly think that when the DC metro, Boston T, London tube, Hong Kong MTR, etc are shut down for the night, every inch of track undergoes a battery of repair work each night that commences just in time for the beginning of service the next day? Nonsense. I’d bet the percentage of time the tracks are being repaired or cleaned during the nightly overnight closure is close to zero. They close them due to lack of demand, not because they have the insight that constant repair necessitates it.

Overnight transit service was common everywhere when the NYC subway was being built. I don’t think it’s quite as historically unique as people think it is. The els and trolleys were 24/7 before we put the trains underground. The State Public Service Commission found in 1912 that the 20 minute streetcar headways the New York City Railway Company was running from midnight to 6am weren’t adequate, and they needed to make them 10 minutes tops, because “persons should not be obligated to wait twenty (20) minutes upon a street corner, especially during the winter, for a car to take them home”: https://books.google.com/books?id=p_pLAQAAMAAJ&pg=PA82&dq=nyc+railway+headway+night&hl=en&sa=X&ved=0ahUKEwiFpqb31fDXAhWHPpAKHSkuAYEQ6AEIOTAD#v=onepage&q=nyc%20railway%20headway%20night&f=false

If only we felt the same way today.

There were periods when overnight ridership was very high due to millions of people working “third shift” industrial jobs. As I like to cite, the IRT 3rd, 6th, and 9th Ave Els ran 10 minute headways from midnight to 5:30am during their heydays. Only the 2nd Ave line closed overnight, forcing people to walk a single extra avenue west during the slow hours: https://www.nycsubway.org/perl/caption.pl?/img/maps/irt_1906_railways_guide_56.gif

Whereever there were a lot of people creating demand for travel, there was generally overnight service, barring blue laws that kept things closed in especially Puritanical states like Massachusetts. Even in Boston I’ve seen old timetables that showed overnight service on streetcars.

The overnight service we have today is probably the least overnight service we’ve ever had, and completely killing overnight service will not fix the subway.

Matt F. December 4, 2017 - 11:50 am

The MTA has taken a third option in Astoria: local stops get only one-way service middays and weekends on the NW. They do this about half the time, with random choice of direction for the local stops. Perhaps they do some maintenance at those times, but the line is no longer reliable.

AMH December 6, 2017 - 2:46 pm

There are also midday express runs on most other lines for track work; it’s hardly unique to the Astoria Line.

smotri December 4, 2017 - 12:22 pm

When there is 24-hour service, it is too infrequent to justify waiting, frequently all alone on a rotting platform, for filthy subway train to show up. Then, when the MTA does its overnight ‘work’, the concept becomes nothing more than a cruel joke. It is so confusing sometimes to figure out whether a given line is actually operating overnight or not that I wind up simply taking a bus or a taxi. The same holds true for weekend service shutdowns, where you need to read the advisories several times in order to make any sense of them. This 24-hour service is a flim flam. Either do it right, or don’t do it. In Paris there are overnight bus lines that replace the metro, which is not a 24 hours service in Paris. The world does not end when there is not 24-hour service!

J Adlai December 5, 2017 - 8:05 am

It seems pretty clear that they were talking about some significant weeknight suspension of service based off the following:

Track work is complicated and expensive on a 24/7 system. Closing the subways on weeknights and/or for more extended time periods would create more opportunities for track installation and testing of the equipment—and reduce costs

They identify 24/7 sevice as being the root of a problem and then suggest closing the subway. This is why everyone is reading this for what it is: a suggestion that NYCT do what the vast majority of other systems do and close at night time. Of course, many on this site frequently talk about New York Exceptionalism, but in this instance it does apply: There’s no reason to shutter the subway ovenight because significant portions of the system have 3 or 4 tracks. This is something fairly NY specific. As a result, the opportunity to re-route service in NY allows you to keep trains running while performing necessary work.

If anything, NYCT needs to embrace the ideas of shutting segments down earlier than usual, as is done on FASTRACK, and also embrace the idea of longer term shutdowns as has been done with the Myrtle rebuild, the Montague tube and soon the Canarsie Tube. CTA seems to be able to use this approach to great effect.

SEAN December 5, 2017 - 9:39 am

It took the CTA trial & error to figure out this issue as they lost ridership when the entire green line was being rebuilt. The riders did return, but it took quite some time. I don’t know if the MTA has the stomach to deal with that.

AMH December 6, 2017 - 2:50 pm

I was amazed at such a sloppy offhanded statement. I would expect a lot more careful analysis from the RPA, such as detailed suggestions of the best kind of shutdowns to maximize productivity while minimizing cost. Because guess what–track work is complicated and expensive on a 24/7 system, but so is shutting down a large system. How much would it cost and how much time would it take to lay up all the trains, then roll them out again? Would that leave any productive time for work? I think the answer has always been no.

Brooklynite December 5, 2017 - 4:54 pm

Before increasing shutdown time, we need to make sure that the existing shutdown time is being used efficiently. IG reports and empirical observations make me far from sure that that’s the case.

smotri December 6, 2017 - 11:01 am

Yesterday, December 5th, going home from work on the Q, at around 6:30 PM, I heard an announcement that the 7 train was delayed because of signal problems. This in itself seems to be standard practice for the 7 line, and I am fortunate not to have to use the 7, but the location of the signal problem was noteworthy – the signals at Hudson Yards were malfunctioning. If signals at a relatively new subway station are not working right, then the problems of the subways are even more extensive than might be suspectd. Could one of the newest stations in the whole system not be built with functioning signals??? Why is this happening???

Kyron December 7, 2017 - 10:06 pm

Not really surprised, given that the Second Avenue Subway experienced delays because of “signal problems” just days after the opening date.

Maclondon December 8, 2017 - 6:35 am

You may be interested in knowing that the new NYCT head, Andy Byford has mentioned 24 hour service being adjusted to better meet engineering works to upgrade the subway in an article in London’s Evening Standard – link below:



Leave a Comment