Step One Toward Fixing the Subway: Be Honest

Governor Cuomo won't get far if the MTA isn't candid about the subway's problems and what it will take to fix them.

Gov. Cuomo. Photo: Governor’s Office/Flickr
Gov. Cuomo. Photo: Governor’s Office/Flickr

The New York City subway is not in good shape. High-profile breakdowns and a tangible increase in severe delays have brought widespread attention to years of decline in service levels. Train breakdowns are on the rise, and service is slowing down. After a long period of growth, subway ridership recently stalled and appears to be declining in 2017 despite the opening of the Second Avenue Subway.

The MTA has for the most part been running on autopilot. Senior management tries to avoid taking responsibility, and planners are afraid of doing or saying anything that would invite scrutiny from the media or their superiors.

This paralysis is accompanied and enabled by a lack of candor about what is going on. Some of the subway slowdowns involve slow orders on the tracks for maintenance, while others involve poor train maintenance. The MTA blames most delays on overcrowding, a claim that was uncritically parroted in the New York Times. But while crowding may be a problem, it has become a fallback excuse, cited whenever the MTA cannot identify another reason for a delay.

There are steps the agency could take to improve service relatively soon, but to make them happen, Governor Andrew Cuomo and his MTA CEO, Joe Lhota, will have to step in and put their authority at the MTA to good use. So far, however, they have done no such thing.

Cuomo wants to be perceived as taking action, so he recently declared a state of emergency, which according to one inside source at the agency has gotten MTA managers scrambling, but without any real purpose. That was preceded by his announcement of “genius” grants, judged by tech people who are mostly not in the transit field.

Lhota, for his part, was just appointed to the CEO position and has to catch up with the changes in New York’s transit situation since he left his first stint as MTA chief in 2013 to run for mayor. He’ll be getting up to speed while retaining his full-time job as an executive at NYU Langone.

What would it look like if Cuomo stopped making empty gestures and started exercising real leadership at the MTA? Here are three steps that, in the medium run, could make a difference for subway service.

First, Cuomo and Lhota should bring in outside experts for real, substantive transfers of knowledge, not the superficial stagecraft of last week’s event with the CEO of RATP, which runs the Paris Metro. Instead of one-time meetings and photo-ops, Cuomo and Lhota should hire people with experience running trains in large, complex subway systems, such as Tokyo or London, and give them space to study New York City Transit’s operations and make specific recommendations. This is likely to involve multiple people working together for a period of months before they can offer concrete suggestions for more efficient operations.

Second, they should declare speed a priority. The MTA has been too hasty to sacrifice service quality for dubious safety rationales. After two fatal crashes in the 1990s, the MTA installed signal timers to reduce train speed. But one crash had nothing to do with speed, and the other involved a train exceeding the speed limit. The MTA also has large and growing safety margins near work zones — stretches of track with slower speed limits. Some safety margins are necessary, but the MTA has gone too far, to the detriment of good service.

Third, instead of his flashy, tech-focused “genius” grants, Cuomo should have the MTA reevaluate routing and scheduling. The solution to a problem may involve a software fix, but it may just as well involve a change in where and when trains run. One option is pruning branches. New York is unique in how complex its subway branching is, and it should look into disentangling lines, even at the cost of removing some one-seat rides. Alternatively, the MTA could change the frequency guidelines to ensure trains that share tracks have the same frequency, even if one route ends up more crowded than the others, to maintain even headways. Having one line that comes every five minutes and another that comes every six minutes share tracks is a recipe for baking “ladies and gentlemen, we are being delayed because of train traffic ahead of us” into the timetable.

The thread tying these three solutions is honesty — honesty about the state of the system, about tradeoffs, and about the scale of effort required.

There is no silver bullet. But neither is the solution some grand undertaking requiring billions of dollars and a monument named after Cuomo. Everything on the above list should take months, not years.

Real tradeoffs will be necessary. It’s possible, for instance, that improving maintenance will require more extensive nighttime shutdowns than under today’s Fastrack initiative, which itself was controversial when it began in 2013. Lhota has indicated openness to this idea.

The changes may also lead to some pushback from the unions. Before he can make millions of New Yorkers less angry about the subway, Cuomo will probably have to make some people angrier. He’ll have to finally invest political capital in improving transit in New York.

Cuomo and Lhota have a choice: They can promise sexy quick fixes and cross their fingers that things will somehow work out (they won’t), or they can be honest with transit riders and get started on useful medium- and long-term solutions.

  • Larry Littlefield

    Being honest also requires being honest about the burden of the past. Yes Cuomo continues to harp on it to try to get people to believe things aren’t his fault, but how many riders are reading the articles? It has to be in their face.

    At the expense of being a broken record, fares (and tolls and taxes) should be cut to the level that is going exclusively to existing service and existing workers. All the additional money going to debts, the unfunded portion of retirement benefits, and catch up from deferred maintenance should be funded by a separate “Generation Greed surcharge” everyone can see. As in you put $50 in the Metrocard machine, and you get a $30 Metrocard and a receipt for a $20 Generation Greed surcharge.

    This isn’t just an MTA suggestion. Mayors and Governors all around the country are coming to be despised as they are forced to impose massive service cuts and tax increases to make up for what was done in the past. Christie in New Jersey. Malloy in Connecticut. Rauer in Illinois. Emmanuel in Chicago. Etc. Etc. Etc.

    In some cases this is well deserved, as they did little to improve upon the situation they inherited. But every politician is going to face the same thing for decades, including DeBlasio when the latest stock market bubble no longer allows they to hide and defer the damage. The result will be the widespread discrediting of public services in general.

    Politicians and unions won’t want to fess up to what their kind have done to people’s future. But for the unions, the alternative is for blame to attach itself to more recently hired workers who had nothing to do with it. Fessing up is the one and only way to perhaps, maybe, avoid an institutional collapse.

  • Maggie

    At the risk of going on a tangient here (to a cogent, excellent article), what do you mean by “every politician is going to face the same thing … when the latest stock market bubble no longer allows them to hide and defer the damage”?

    Are you saying that a stock market correction will imply lower revenues to service debts? Can you give more detail?

    Or do you mean if a stock market correction compels pension funds to walk down the expected returns on equity portfolios, the present value of unfunded future liabilities gets less affordable?

    Even if metrocards were priced with a 67% surcharge for paying down debt (to 0? I mean, that wouldn’t be politically feasible. The political will and corporate finance would permit some nonzero level of debt). Even if there was a 67% surcharge for retiree benefits, are you suggesting the public would insist on clawing those benefits back from retirees and workers?

  • Larry Littlefield

    “Or do you mean if a stock market correction compels pension funds to walk down the expected returns on equity portfolios, the present value of unfunded future liabilities gets less affordable?”

    Actually, if the value of stocks were to fall by half that it might actually be possible to achieve the projected rate of return — from those lower levels. But the lie of average or above average returns from a base of inflated asset prices would be exposed, and pension contributions would then soar even further. Explained here.

    Financial bubbles also temporarily inflate tax revenues for NYC and NY State. Including real estate-related tax revenues for the MTA. Setting off a massive grab among those on the inside, leaving those on the outside with the bill when things return to normal.

  • Alon Levy

    Okay, but the situation today is specifically not deferred maintenance on fixed plants. There’s deferred maintenance on rolling stock, but that’s a different discussion. The problem comes from ongoing mismanagement.

  • “One option is pruning branches. New York is unique in how complex its subway branching is, and it should look into disentangling lines, even at the cost of removing some one-seat rides. Alternatively, the MTA could change the frequency guidelines
    to ensure trains that share tracks have the same frequency, even if one
    route ends up more crowded than the others, to maintain even headways.
    Having one line that comes every five minutes and another that comes
    every six minutes share tracks is a recipe for baking “ladies and
    gentlemen, we are being delayed because of train traffic ahead of us”
    into the timetable.”

    I’ve recently posted a proposal applying both these principles here. Without capital works at Rogers junction and 138th/Grand Concourse, it doesn’t do much more the ‘A’ division, but a pretty good solution is acheivable for the ‘B’ division.

  • Maggie

    Sure, good governance always means smoothing out anticipated cyclicality in revenues.

    Are you saying the city is underfunding the MTA by overpaying teachers? I thought the MTA’s funding uncertainties were more state-driven.

    And apologies for not knowing, what’s a 20/50 pension?

  • qrt145

    I’m not Larry but I believe 20/50 means eligibility after 20 years of service and 50 years of age. See for example

  • Larry Littlefield

    Right, for those who did not follow in, following a transit strike in the 1960s TWU members forced the city to agree to a pension at age 50 after just 20 years of work, far richer than had been promised or funded. Knowledgeable workers all headed for Florida, costs soared, and the subway system collapsed.

    Some years later, newly hired workers were forced to accept a measly one year in retirement for each year worked, half pay retirement at age 55 after 25 years of work. And that was, once again, all that was paid for.

    So in the early 2000s, when the first workers were faced with having to work five more years, the TWU went on strike demanding a retroactive restoration of the 20/50 pension. The “New Directions” wing of the union had taken power promising this, and had to get it or prove it couldn’t be got. But it didn’t work. So was the transit system saved?

    It’s better off than it would have been, but instead of being destroyed by pensions it is being destroyed by debts. And neither the city of the state is putting up real $ because of the cost of other pensions. When NYC teachers, who had been promised retirement at age 62 after 30 years of work (with prior teachers getting 25/55), they actually got a retroactive, unfunded deal for 25/55 again. Causing city pension payments to soar. As a result.

    But don’t rush to be a teacher, because they subsequently screwed new hires again.

  • Larry Littlefield

    When did NYCT forget what it learned over the decades?

    After years of management wage freezes and downsizings, did the managers stop managing. And then despite wage increases did the workers stop working?

    I’m not sure of this.

    As for the grade timers, part of this may be imposed by the FTA.

  • Alon Levy

    I don’t think any of this was imposed by the FTA.

    As for NYCT forgetting what it learned: on the contrary. The lessons it learned from 1970s debacles like the R44 and R46 was “don’t try to innovate.” Hence, positive things like Fix It First and negative ones like indifference to speed.

  • Larry Littlefield

    I was at a meeting in the early 2000s (mid-2001?) when then NYCT President Larry Reuter said that the success of the Metrocard proved the agency could successfully implement a high-tech project.

    And on that basis ATS, CBTC, PA-CIS, the Rail Control Center, and a new fiber-optic communication system were launched. $billions were spent — with those German and French experts (Siemens and Alstom) getting a lot of the $.

    So what happened, and why? Perhaps they should have different “geniuses” look into it.

  • Alon Levy

    CBTC was specifically a reaction to the accidents in the 1990s. It’s also legitimately an investment in speed and reliability (the L has the least delays even though it’s in a near-tie for most overcrowded line). Train drivers go slower than they can because they don’t fully trust the signals, and CBTC lets them go as fast as they are allowed.

  • Larry Littlefield

    The problem is the price went to the moon, and the schedule was set way back. The assumption was that after the first few projects they could go to CBTC very quickly and much cheaper.

    Meanwhile, the auto industry seems to be rapidly moving toward driverless cars?

    So what happened?

  • sbauman

    As for the grade timers,

    The grade timers are necessary because the MTA’s emergency brakes don’t work very well. If they worked as per spec, 3.2 mph/sec, they would not be needed. If the dynamic brakes were applied in an emergency, they would not be needed. There would have been no derailment at Union Sq nor a collisions on the Williamsburg Bridge or at Roosevelt Ave.

    The MTA went from absolute system that would provide safety, despite what the operator did, to one that would provide safety, only if the operator followed the rules. This allowed them to cut corners in rolling stock design.

  • Joe R.

    The part about reducing speed is even worse than this article makes it out to be. Not only were timers installed in many places where they weren’t necessary, such as long stretches of straight track, but the trains themselves were neutered. The DC equipment had field shunting removed and the AC equipment had its traction computers reprogrammed to approximate the performance of the neutered DC equipment. This reduced balancing speed from about 55 to 57 mph down to 42 to 44 mph. It also severely retarded acceleration rates above speeds of about 18 mph. This means even in the absence of timers or other speed restrictions trains will run slower. Locals take about 10 seconds longer per stop. Expresses can have even worse delays in terms of percentage. With the reduced performance, there is also little chance delayed trains can recover schedules.

    The first thing the MTA should do is to return the trains to their unfettered state. If there are places where signal spacing might cause a problem if trains accelerate too fast, then you can post speed restrictions only in these areas. On every other system the train operators are responsible for controlling train speed. I’ve never heard of any other transit agency intentionally detuning equipment because of signal spacing issues.

  • Hux

    RE this and,

    The MTA is not honest because it is not a private company. They do not have an incentive to care about anything but looking good politically. That’s why they tried to stop reporting delays and OTP publicly. They think that it will make their lives easier to be dishonest or misstate facts or cherry pick good numbers, maybe they think it will give them time to do real work, but it usually backfires and then they end up spending time just scheming and trying to BS the public, the press, the board, etc. I bet even internally, people who aren’t in the press office don’t know that they are BSing the public and then they believe their own confusion.

    Like… the whole thing with wait assessment being better than OTP. Wait assessment was sample-based, it was a straight average for all lines, and it is still percentage-based so that it can pass a train that’s 5 minutes late on a 30 minute headway but fail a train that’s seconds late on a 2 minute headway. That doesn’t make sense for the customer experience (though it is measured along the line, rather than at the terminal, which is better… most customers don’t ride all the way to the terminal, which is why the new journey time metrics are good for them.)

    But WA certainly doesn’t make sense for operations… which is why rail control never really managed to WA anyways (how would they even calculate it? and it doesn’t have cause data…), but leadership wasn’t paying attention to OTP/delays… and look at the mess now! Yes, in an incident, trains are held, but that harms WA and OTP because holding trains lengthens headways and makes them late. It will end up making service less bad, yes, but it does – and should – harm performance metrics. So much confusion about what WA was… it’s not evenness… it’s BS.

    Look now, at all our rising delays! They passed all these signal mods, flagging rules, they weren’t looking at the results. They didn’t care and/or they were ignorant. Now we need so many more trains just to meet the adjusted running times/schedules, or else it’ll appear as a service cut. We don’t even have the space in the yards.

    Why didn’t anyone do something? Well, again, no incentive to care. If your leadership isn’t good, if it lacks vision, it’ll just try and look good… the easiest way… Alon, thanks for bringing all of this to light, other reporters don’t seem to get it, except Dan…


  • Hux

    Yes, and all the looking good happens at every level of government. So bad news won’t get passed up the chain. Maybe they are just ignorant and just want their cherry picked talking points and no inconvenient truths.

  • Hux


    I don’t think the writers of that article fully understand CBTC on the L line… the L can speed into 8 Av now, it’s an automated train, but it doesn’t because of safety concerns, which may not be justified. Also, even though it’s automated, they need operators until they have platform screen doors, in case someone jumps onto the tracks, etc. The operators don’t actually drive the L, but they’re there in case they need to take control. Systems with fully automated trains usually have platform screen doors at all stations on the line. (Think about all the advertising space on those platform screen doors!)…

    They think the main benefits of CBTC are the reduction in maintenance and operating costs… I don’t know if CBTC will be cheaper, but it’s main benefit, in my opinion, is increased speed. With CBTC, there aren’t any slow crews and the signal system allows trains to go faster. Congestion pricing, value capture, new payment technologies, new subways… these should not be built to reduce crowding, but to make things faster! If we ran our trains on-time, we wouldn’t have crowding. We are a rapid transit system, we need to stop saying “reliable” service and start saying “fast” service, because fast service is reliable – it means we’re not waiting long on the platform or on the train, it means more capacity, etc. If service is fast, it’s because everything else works out – more reliable rolling stock, track, signals…

  • Hux

    There are many disconnects in the MTA. The people who run the trains don’t write the schedules, and the people who write the schedules often don’t know much about the conditions underground. The people who make the policies may have never learned how to operate a train. Capital construction builds things that maintenance won’t be able to maintain easily, or they build things without understanding operations (some of our newest terminals have low capacity.) Hard for everyone to work together and think about the whole subway system… let alone all of the MTA modes and connections to the buses and railroads… let alone to City and other agencies in the region!

    For instance, the Montague Tunnel (R train) was rebuilt without adequate clearance for older rolling stock. Instead of fixing this, the MTA is spending more money in the long-term rerouting older trains away from this tunnel, which takes longer and disrupts more service, in order to get them to yards for repair. They reopened the tunnel after Sandy repairs early, and didn’t want to close it again after finding out about the clearance issues, in order to save face.

    Or, let’s say there’s a track problem causing trains to break down. Track will easily blame Car Equipment and say they aren’t maintaining things properly, rather than just fix the problem, to save face. Service delivery (they run the trains) will blame slow crews on the department that trains the crews. Management will blame unions, unions will blame management. Governor blames Mayor, Mayor blames Governor.

    It is hard to hold departments/divisions accountable (car equipment, service delivery, track, signals, MOW, third rail, etc etc), especially with these long-term changes. Signal modifications are supposed to not slow down service, but in reality, whenever there is an incident, it takes a lot longer to recover, since the control lines and all of the grade time signals make the system less resilient. Also, crews don’t trust the signals since they are sometimes miscalculated, so they go slower. There is a lot of distrust. Signals will blame crews, crews will blame signals. Maybe they will both blame ridership or the weather. So much bureaucracy and paperwork to change anything.

    I’m sure it’s the same for buses. Let’s say a bus runs into a pot hole and breaks down. Who is to blame? The operator, for running into the hole? The City, for not maintaining the road? The weather, for causing it? The bus maintainers, for not making the bus more resilient?

    All of this extra time arguing and not coordinating things costs extra time and money. Easy to blame others, esp. ridership or “decades of deferred maintenance”… even though we spend more and more. I don’t know how to change this culture. Maybe the fish rots from the head. But it seems many people are happy to just work slowly and collect their pension, after years of working here, they are disillusioned and demoralized.

  • Hux

    It’s truly saddening that in 2017, we’re pitting safety against rider experience by slowing trains, rather than using modern technology to improve safety without diminishing the rider experience.

    The MTA has been slowing down with more and more safety rules and signal modifications, many of which don’t actually make us safer but serve a political purpose.

  • Railfan89

    Basically, management has been adding more and more rules, some of which make things safer and some of which are supposed to make things safer, but really aren’t needed and are done mainly to save face and appease unions. All of these things are slowing down service and making it less reliable, meaning that when there is an incident, it is more impactful and it delays more trains. There’s also better data. Over all, incidents themselves are pretty stable. If there is a signal incident, it will end up delaying more trains – not because there are more signal failures, but because the system is slower.

    Many of the changes also don’t actually increase safety. Crews operate slowerr, trains move slower, the signals make them go slower, and they go slower for track workers. Some of this is needed but a lot of it is not and it’s been allowed to get worse because of ignorance in the agency and mismanagement. The incentive is not to speed up service, but to look good and make the politicians who run the agency look good… often this means quick, short-term fixes, ignoring big problems, hiding facts from the public, misstating the truth, etc. It also means confusion and ignorance, because if you just want to look good, and your boss just wants to look good, the goal of not having bad press can pervade the agency and make everyone hide bad news from each other. Ignore it until it’s a crisis…

    If the media pressures them, then they will want to speed up service, but it’s hard to know if they really want to and if they care or if they are just saving face. And they think about things in the wrong way because they are always trying to find things that make themselves look good, so it is hard to get them to admit problems and fix things that could mean they’re blamed.

  • Prince Train

    I don’t think getting rid of 24/7 service to perform maintenance is practical. Since service is slowing down, they need more trains to meet demand, and they don’t have enough space to store these trains in the yards. Laying them up along the rest of the system would impede work trains. Also, there is no way to close many stations, so they would need to invest in gates and such. Over all, I think it is not necessary and they can continue with closing down sections of the system at a time, but not the whole entire system. I don’t think they could really do that… even during snow storms, they keep the trains running to keep the snow off the tracks, and MTA employees can ride the trains to work!

    Also, they now have adjacent track flagging, so work on one track will slow down trains on the adjacent tracks. Despite having no evidence to show it would increase safety, this rule was implemented a few years ago and is a big problem now. More and more flagging rules have been added since then (revenue trains can now “taxi” workers from stations to job sites, trains need to go even slower for longer sections of track, farther apart from work trains, etc.)

    More things slowing down service? Their R179s now are undergoing 30-day testing and they are trying to make sure the door chimes are played at the same time that the doors start closing, or else more running time will be added, these trains will run slower than other ones, there will be gaps in service, etc.

    (The R211 is years away but they’re about to make the same mistakes in car design that they made with the R143/R160. Doors are still not offset from each other. There’s still poles in front of the doors. Windows are still smaller than they could be. The open gangway is soft-shell. And they’re still using non-convertible cabs. But perhaps these can be changed if they provide feedback.)

  • Hogan

    Even after adjusting for route length and the number of merges?


He's the boss. Photo: Flickr/NYS Governor's Office

When the Going Gets Tough at the MTA, Andrew Cuomo Disappears

Remember when Andrew Cuomo announced that he'd sealed the deal on a new contract with the TWU? Or when he empaneled an "MTA Reinvention Commission" to shape the agency's five-year capital program? Or when he ordered the MTA to quit dragging its heels on cashless tolling, and the agency promptly delivered? The governor would like you to forget all that.