Home MTA Absurdity The annals of bad ideas: Slowing down the subway

The annals of bad ideas: Slowing down the subway

by Benjamin Kabak

Handing policymakers numbers without contextualizing them often leads to Very, Very Bad Ideas from policymakers. For instance, if I told someone that 90 people were struck by subway trains in 2009, that person might be surprised. After all, 90 seems like a good number, and the 40 people who died from the impact represent tragic deaths indeed.

But what if I said 0.000006 percent of all subway riders were struck by trains in 2009? Would your response be to propose a measure that would add 20 minutes to a trip from Coney Island to Midtown? Or 12 minutes for those who ride the D from the Bronx to Herald Square? As Pete Donohue explores in his column today, that’s what one Assembly representative has done.

Dononhue writes of this Very, Very Bad Idea:

Commuting by rickshaw would be faster than the subway under a Bronx assemblyman’s proposal to increase straphanger safety.

Before entering each and every station on the line, a train would have to come to a complete halt just shy of the platform. Only after a required pause could the train move forward to pick up passengers. The maximum permitted speed when entering would be 5 mph.

The regulation by Democratic Assemblyman Marcos Crespo – introduced in the Legislature last week – would achieve its goal: preventing people from getting hit by arriving trains.

The bill — labeled A04796 in the Assembly — is available here, and Crespo’s justifications defy logic. He seems to think that 155 injuries over three years as well as liability of between $30-$60 million a year is rationale enough for this move. Crespo wants trains to come to a complete stop while drivers — he says conductors, but the drivers would do it — inspect the tracks for people before moving ahead very slowly. The cost in lost productivity due to slower commutes alone would wipe out the savings in human life and liability payments.

Donohue believes that Crespo’s “heart is in the right place” and claims that personal accidents are “harsh realities of the subway that are either routinely ignored or quickly dismissed as unavoidable.” From the perspective of the personal, perhaps Donohue is on target, but the numbers don’t support this sentimentality. Exceedingly few people are struck by subway trains every year, and even fewer die from those accidents. In a traditional model of economic efficiency, It’s not worth robbing the subway of speed to save a handful of lives.

Where Donohue hits the nail on the head though is in his attacks against those who slammed the MTA for requesting information about platforms doors. Crespo claims that it is “unlikely the MTA will incur that expense” of installing the new technology. He seemingly ignores the MTA’s proposal that someone else foot the bill for the platform doors and doesn’t know or pay attention to the difference between the authority’s operating and capital budgets.

As I noted a few weeks ago, critics who blasted the MTA’s Request for Information were barking up the wrong tree. Donohue, who notes that at least one company — Crown Infrastructure — is interested in the MTA’s proposal, agrees. “You would think” based on the reaction, he says, “subway executives had expressed an interest in building a space shuttle from the sharp criticism that emerged.”

Citing the high cost of fires caused by track debris and the safety benefits, Donohue says the authority “should test platform doors in a pilot program and not be rattled by critics.” I have no qualms with that claim, but we should not overstate platform dangers either. One train may hit one person every four days on average, but that represents one six-millionth of one percent of all subway riders. The city would suffer tremendous were slower subways the price the other 99.999994 percent would bear in the name of false security.

You may also like


BoerumHillScott February 14, 2011 - 11:47 am

Someone should propose a bill that cars do the same at every intersection, even if there is no stop sign or the light is green.

Someone January 15, 2013 - 4:07 pm

That’s dumb. Do you know how much traffic that would create?

Esteban February 14, 2011 - 11:56 am

And of those 90 people hit by trains, how many were attempting suicide by train? I knew somebody that did just that a few years ago, and I think it’s more common than people realize. Sorry for the morbidity.

mattmaison February 14, 2011 - 5:24 pm

I was thinking the same thing. I’m sure it’s a good portion of them.

Alon Levy February 14, 2011 - 11:58 am

Don’t blame him. The ICC and FRA have had the same approach to railroad safety.

JebO February 14, 2011 - 12:25 pm

Exactly. Look at Positive Train Control, which would create in-cab signaling to prevent fatalities from train-vs.-train collisions on commuter railroads. This is going to cost the LIRR $324 million, and Metro-North $350 million. Oh, wait, the railroads already have in-cab signaling? What’s that, you say? There’s never been a fatality from a train-vs.-train collision in the history of Metro-North or in the past 40+ years on the LIRR?

If the feds actually cared about overall transportation safety, they would focus on getting people out of their cars and onto trains. Instead, they fixate on preventing one exceedingly unlikely type of rail accident. And in solving that problem they’ll actually decrease overall transportation safety by forcing the railroads to raise fares, thereby encourage passengers to get back into their cars.

Bolwerk February 14, 2011 - 1:12 pm

Yep. It’s probably deliberate. Fear of failure for rail is unwarranted, given its advantages despite the overwhelming regulatory hurdles even a non-passenger freight railroad non-FRA passenger service must endure. The real fear is the proven success of rail. In this case, it seems to just be a less-than-intelligent state assemblyman who internalized a zany narrative about trains and safety.

Anybody else notice the sudden flurry of anti-rail propaganda that showed up on Google News over the weekend? Robert “Voting Affects My Objectivity” Samuelson had an op-ed about Obama’s latest rail push that was so riddled with factual errors that it could Glenn Beck’s fact checkers blush. It even cites such paragons of analytical integrity as Randal O’Toole, and completely misrepresents (likely deliberate lying) the facts of highway subsidies.

Alon Levy February 14, 2011 - 2:00 pm

PTC is about more than in-cab signaling. It’s, at the weakest, a system that automatically stops a train if it passes a signal at danger. The subway has something somewhat stronger, called ATS, automatic train stop, which also stops a train if it speeds.

Mandating nationwide ATS is a good thing. The problem is that the FRA has a zillion other dysfunctional rules – in-cab signaling and PTC required for speeds higher than 79 mph (the global standard is 125 mph), 900 tons buff strength for lead cars (the EU mandates 200, Japan 100), special testing required for off-the-shelf European and Japanese imports, loud horns at every grade crossing without four quad gates.

Bolwerk February 14, 2011 - 2:03 pm

What exactly is becoming of buff strength standards when PTC goes into effect? I thought they were being dispensed with by 2016 or so.

Alon Levy February 14, 2011 - 3:38 pm

My understanding is that the buff strength rule will go, but the FRA will make up new rules. It’s already harassing Shinkansen vendors, whose product has a zero fatal accident record, with questions like “Can your train survive a 6 kg steel ball impact?”. (The answer: “yes, but we’ll have to modify the trains and add weight to them.” For the record, Japan is the world leader in both safety and train weight reduction.)

Nathanael March 3, 2011 - 4:09 pm

Does the FRA have some knowledge of planned cannon attacks on passenger trains?!? That is a pretty insane question otherwise.

al February 14, 2011 - 4:13 pm

Combining lower buff strength requirements with composite construction would lead to very light cars. 35-40 ton bi-level commuter rail units could become the norm.

Alon Levy February 14, 2011 - 7:19 pm

In Tokyo, single-level commuter EMUs are in the 23-30 ton range, depending on whether the car is a motor or a trailer. The bilevels are heavier, but they’re also less advanced, since the focus for an overcrowded system like Tokyo is fast boarding and alighting (i.e. more doors) rather than in-vehicle capacity.

Andrew February 14, 2011 - 10:52 pm

Watch your terminology. That’s not what ATS means here.

The ATS on the subway is Automatic Train Supervision – a train tracking and automated routing system on the IRT. It doesn’t provide safety functionality.

With the exception of the L, the subway uses (and has always used, since 1904) a traditional stop-arm wayside signal system, as described in detail here: http://www.nycsubway.org/articles/signals.html

Most of the system has no speed restrictions. Where there are speed restrictions, they can be crudely enforced by the signal system: http://www.nycsubway.org/artic.....mesig.html

Alon Levy February 15, 2011 - 2:45 pm

Yes, the signaling system enforces (some) speed restrictions and protects from SPAD… that’s what’s called ATS both on the national system and abroad. Though, there are a lot of different flavors of ATS, some not providing robust protection from speeding. And, of course, this says nothing about how the system works, how it keeps track of the trains, how it communicates the signal aspect to the driver, etc.

Someone January 15, 2013 - 1:19 pm

ATS is also used on some upgraded parts of the BMT, it’s not exclusive to the IRT.

ant6n February 15, 2011 - 10:19 am

have you heard of the recent train crash in Germany? Ironically, it was due to lack of ptc not preventing a freight train going through a stop signal at a siding, so that it went back on the track and hit a lightweight DMU, with 10 dead. I guess they’ll change the regulations now to require ptc even on single track, which previously was regulated to 100km/h max.

Alon Levy February 15, 2011 - 2:43 pm

Yes, I have.

Bronx Thru Express February 14, 2011 - 10:54 pm

Actually, wasn’t there such a collision about 20 years ago in Mount Vernon? I remember seeing the news reports of rail cars propelled up to street level.

Al D February 14, 2011 - 12:13 pm

Actually Mr. Donohue is incorrect in his conclusion. It would not prevent people being hit by a train. Rather, it would lessen the chances. The only way that you can prevent people from being hit by a moving object is to remove either the person or the moving object. Otherwise, the potential will always exist.

JP February 14, 2011 - 12:22 pm

So let’s put those plexiglass doors at the three stations most likely for jumpers or accidents: 42nd, 14th, 125th. done.

Nathanael March 3, 2011 - 4:11 pm

Agreed. Place platform doors at the most trouble-prone stations and leave it at that.

I’d actually start with the remaining station with curved platforms and mechanical gap fillers — the trains have to spot perfectly at that station anyway, and it is one of the busier stations. Which one is it again?

Peter Smith February 14, 2011 - 12:50 pm

i’m assuming suicides were factored out, but even if they weren’t — why can’t we make trains with cow-catchers out front to push people off the tracks?

Kid Twist February 14, 2011 - 1:12 pm

When the 2 runs local at night, it makes more than 60 station stops. This effectively turns that into 120. Put another way, even if the stop-and-crawl routine adds just 30 seconds to each station stop, that’s an extra half-hour to get an overnight 2 from the Bronx to Flatbush.

Bolwerk February 14, 2011 - 1:16 pm

This proposal could probably be likened to the safety procedures in place when workers are on the tracks. The train has to slow down to a crawl, and can easily take more than a minute to come to a stop in a station.

Joe Steindam February 14, 2011 - 1:59 pm

Especially on the IRT in Manhattan, some local stations are so close that I even wonder if there is enough space of a train to leave a station completely without having to stop before entering the next station. This is just a stupid proposal to the problem of high cost injury suits against the MTA. The Legislature can do better by setting a limit on damages. I know that’s not a popular decision, but it’s better than this stupid rule.

What are the accident rates on the MNRR or LIRR (or for that matter, the Buffalo Metro)? Surely this threat exists on our other rail networks, so why is this legislation limited to just New York?

Bolwerk February 14, 2011 - 2:04 pm

It’s a great idea that won’t happen as long as Sheldon Silver is in charge.

Bolwerk February 14, 2011 - 1:21 pm

Crespo likely has an ideological desire to make sure transit doesn’t work. Such plans are often couched in terms of safety. He also is one of the clowns who blamed the MTA for service cuts after voting to slash its funding last year.

Donald February 14, 2011 - 1:44 pm

Making traisn come to a complete stop before entering the stations is completely insane. However, having trains slow down while entering is a good idea. Currently, the MTA requires train operators to enter statiosns at full speed. The TWU has been trying to get the MTA to allow trains to enter at slow speeds to prevent accidents, so this proposed bill must be the result of their lobbying.

Benjamin Kabak February 14, 2011 - 2:01 pm

It’s completely unnecessary to have trains slow down as they enter stations. Operating at slower speeds simply robs the system of its efficiencies. Trains already are capped at unnecessarily slow speeds because of the 1991 Union Square accident. There’s no need to make things worse.

Joe Steindam February 14, 2011 - 2:03 pm

Why would the TWU want to operate trains so that they’ve started slowing down before they enter stations? Is there any reason the MTA has the rule that trains start slowing down once they enter a station? I guess I don’t really notice a difference, it always seems like the trains have started slowing down before they enter a station, but trains probably still enter stations above 20MPH.

Donald February 14, 2011 - 2:11 pm

Because TA workers are sick and tired of going through the nonsense that they get put through every time there is a 12-9 (person hit by train). This is from the TWU website:

As of March 11, there have been 31 12-9’s in 2010. 31 people have been killed or injured by subway
trains. 31 crews have had their lives disrupted.

None of the 12-9’s was the fault of the train
crew. But some of them might have been
avoided if trains came into the stations more
We want to reduce the number of deaths
and injuries in the subway. We want to
reduce the number of crews traumatized by
something they had no control over.
Do NOT enter a station at full speed.
Operate with caution —
especially if the
platform is crowded. When the TD or ATD
asks where you lost your time, say you were
preventing 12-9’s and name the stations. If
supervision pushes you to go faster entering stations, let us know who and where.

VLM February 14, 2011 - 2:15 pm

So what you’re basically saying is “We want to create our own work rules and slow down the system so more people blame the MTA for their delayed commutes, and we can bilk them for higher wages.”

You’re still talking about, as Ben noted, a de minimus number of accidents. 31 people hit by subway trains over 10 weeks is still less than 1/100,000th of a percent of all riders. Come on.

Bolwerk February 14, 2011 - 2:23 pm

Jesus, that’s obnoxious. They shouldn’t be deciding for themselves when they can slow down the entire city’s transport system. I’m sure it’s traumatic when a motorman hits somebody, but it’s an occupational hazard. If you can’t handle it, you should really consider another line of work.

nycpat February 14, 2011 - 3:42 pm

That’s very easy to say. Certain stations at certain times I come in slower because of crowds, curves etc. I have no trouble maintaining the schedule.
A guy can come into a station at 45mph and still take 40 seconds to make a stop. Or you can come in at 20mph and stop in 19 seconds. The speed does’nt matter, it’s the braking.
The rules say come in at full speed. The rules also say you will have the train under control at all times.
I would’nt take that union missive too seriously. No one would ever tell a dispatcher that. There’s always ample nonsense going on in the subways to explain away any lateness.

Bolwerk February 14, 2011 - 3:50 pm

I don’t think they should have no discretion to slow down if danger is possible, of course. And if you have no trouble maintaining the schedule* I don’t see why it’s a problem. OTOH, increasing time has its own safety implications. Rush hour platforms fill up quickly, and making them more crowded is also a serious danger.

* Out of curiosity, what does that mean? Does it mean being on time to the minute, or within the range the MTA defines as “on time”?

nycpat February 14, 2011 - 4:23 pm

It means arriving at “gap stations” at the scheduled time. These stations have “holding lights”e.g. BowlingGreen, Nevins, TSQ etc. I was speaking non-rush hour. Rush hour you always lose a few minutes in the CBD. If you arrive at a gap station 3 minutes hot you should be held by control center but sometimes they have other priorities.

Andrew February 14, 2011 - 10:31 pm

If you’re sick and tired of doing your job, then quit and find a different job.

Slowing the trains would require more trains to provide the same level of service. The TWU has the same motive here as in its opposition to OPTO and its support of multiple station agents per station: maximizing membership. If you’re looking for safety, I suggest looking elsewhere.

nycpat February 15, 2011 - 12:15 am

So, all people with PTSD should just buck up and get over it or just transit workers. If someone gets PTSD working for NYCT, they are disposable? I guess they are in this new society we are having imposed on us.
This was a onetime flyer handed out by the union. There is no organized campaign to slow down the subways. Whoever wrote it accomplished what they wanted, i.e. they got their organizing cred established.

Bolwerk February 15, 2011 - 11:43 am

I don’t see anything wrong with providing therapy and leave if someone gets PTSD (which isn’t too likely in the case we’re talking about anyway) after hitting someone with a train. Still, hitting people is a routine part of NYCT operations. This probably especially affects bus drivers, but also affects train operators. Just as the TWU shouldn’t be deciding for itself its safety procedures, especially when they negatively affect the services’ practicality, the MTA should certainly budget and prepare operationally for when these (relatively predictable) things occur.

Fairness is a two-way street.

Alon Levy February 15, 2011 - 2:47 pm

If Walder handed a one-time flier proposing laying off the entire workforce and hiring strikebreakers, you wouldn’t be so cavalier about it.

nycpat February 15, 2011 - 5:57 pm

Nor would anyone. You’d need the National Guard, no joke.
I’m cavalier about it because, unlike Buses, trying to effect union discipline and solidarity in Rapid Transit Operations is like herding cats. The only time I’ve seen a rule book slowdown it was a spontaneous response to a truly capricious disciplinary decision by a superintendant. The control center had to announce over the radio that the superintendant changed his mind, no one would be disciplined. TWU had nothing to do with it.
In the end the Train Operator is charged with the safe operation of the train. Both the union and supervision can’t tell a T/O to slow down or speed up if he follows the rules and operates safely.

Alon Levy February 15, 2011 - 6:56 pm

Strictly speaking, Walder can’t really fire the entire workforce. There are legally enforceable union agreements preventing that. (And that’s a good thing, by the way – companies don’t become more professional by hating their employees.) And yet, such a suggestion would be taken as a declaration of ill intentions, and justifiably so. It’s the same thing.

Nathanael March 3, 2011 - 4:18 pm

Trains should actually travel even faster than they do now. This will probably have to wait for resignalling, and given general paranoia, possibly for platform doors.

Spendmore Wastemore February 14, 2011 - 6:39 pm

Let’s just slow the trains to average a bit less than a kid on a bicycle cranks out – a steady 15-18 mph.

The current system speed is around 15 mph.

The same level of transport could be accomplished by removing the trains and converting the tunnels to bike paths, which would allow cutting the MTA budget by 99%.
They’d still need a crew to change the light bulbs and a dozen or two to lobby Albany for more money.

Donald February 14, 2011 - 8:01 pm

The current system speed is 15 MPH? Since when? It’s nowhere near that. Stand on the 81st St. platform on the B/C line one day and tell me how fast the express traisn are zooming by. It’s certainly not 15 MPH.

Alon Levy February 14, 2011 - 9:27 pm

Thin air is a perfectly good source of information. For the rest of us, there’s the NTD, which says it’s 18.2 mph, averaged on a per-car basis. (i.e., the snail-speed 7 gets weighted more than the D.)

Spendmore Wastemore February 15, 2011 - 5:29 pm

Whoops, that’s 31 miles. And 2 hours on the train.

Spendmore Wastemore February 15, 2011 - 5:22 pm

You expect an exact figure from a snarky comment? I did say _around_ 15mph, and to my mind 18 mph is about as broken as 15.

Hop on an Interstate and drive 18 mph, then 15 for 34 miles (the length of the A train’s run) and tell me how swift that is. Yes you’ll have to use an access road parallel to the highway or the breakdown lane.

As for the 81st street example, time how long it takes to traverse the station and figure its speed at that one point. For something running on a 3 mile straightaway even that brief top speed is underachieving. Then time the A/D trains from 59th to 145th and divide into the distance traveled.

Alon Levy February 15, 2011 - 6:59 pm

18 is pretty normal for legacy rapid transit. (Newer lines do a bit more, about 20-30.) You can’t get these speeds driving in a city, unless you make massive sacrifices to capacity. The capacity of just one two-track subway line is about the same as that of all north-south roads in Manhattan, including the FDR and Henry Hudson. Try to move all that subway traffic in cars and your speed will be deep into single-digit territory.

Nathanael March 3, 2011 - 4:21 pm

It’s the large number of stops which keep the speed down. Driving in a city, you get just as many stops for red lights or for stop signs (or for other traffic stopping at red lights or stop signs or taking entrances or exits).

Anon February 14, 2011 - 8:58 pm

What is this the 19th Century?

There are object detection systems that can be added on the cars…


Joe Steindam February 15, 2011 - 11:02 am

This is an interesting idea, although it would need to be adapted greatly for use by subways. After all a single subway car weighs about 20 times as much as a car, and has 9 subway cars behind it, so the system would need to be stronger, and it would likely have to trigger breaks in all cars in the train at the same time.

I see this better being adapted to ensuring that trains are evenly spaced along the system. Although this system is probably reliant on lines of sight, and its effectiveness is probably lessened when there are turns involved, so it’s not a perfect solution for detecting trains ahead either.

Someone January 15, 2013 - 1:20 pm

No, a subway car could also have 7 or 10 subway cars behind it, which can be confusing where the platforms are one length.

Andrew February 14, 2011 - 10:24 pm

It’s far worse than that. The signal system is designed under the assumption that trains will generally be moving at speed, except when stopped at a station. Violating that assumption comes at an extreme cost to capacity – my guess is on the order of 10-15 tph. So not only will the trains be slower, but they will be much, much more crowded, since they won’t be able to run as frequently. The platforms will also be more crowded (increasing the likelihood of someone falling to the tracks!).

And on the flip side, increasing running times means that many more trains will need to run. That’s more crews to be paid and more cars to be purchased and stored and maintained – to provide less frequent service.

I have trouble coming up with a more boneheaded idea.

Peter February 15, 2011 - 12:51 am

Just as there is unfortunately no intelligence test required before becoming a parent, similarly, there is not one required for legislators. Come to think of it, there isn’t one for voters either.


Pondering Crespo’s crazy slowdown idea :: Second Ave. Sagas February 15, 2011 - 1:15 am

[…] way. As I reported yesterday morning, Crespo has submitted a bill to the State Assembly that would over-legislate a non-existent safety issue. Since 0.000006 percent of subway riders are struck by train cars, Crespo would like the other 1.6 […]

Tubeprune March 16, 2011 - 4:50 am

This is a system issue: A subway train is an expensive piece of kit – $15million+. So is the tunnel it runs in and the train control system designed to keep it safe. It is designed to carry the maximum number of people in the quickest time over a given distance. To do this you need maximum acceleration (1m/s^2) and as close to that as possible for braking, so you can get trains into and out of stations as quickly as possible. You should only keep a train in the platform for less than 40s. This way, you can get 30 trains per hour along a single track. Many cities around the world get this or close to it every day.

Slowing to 5mph to enter stations is a nonsense. It would reduce the possible throughput to about 18 trains per hour.

As for platfrom screen doors – they could be used to air condition stations as well as for safety – but you’re talking $2million per platform edge. Does the MTA want to spend that sort of money?

Just paint a line along the platform edge and put up safety notices reminding passengers to say behind the line for their own safety.


Leave a Comment