Home Asides Death by a thousand cuts

Death by a thousand cuts

by Benjamin Kabak

During the MTA Board’s committee meetings this morning, New York City Transit unveiled plans to adjust its bus scheduling, and the result are a bunch of minor cuts to bus schedules throughout the city. No routes will be scrapped, and the J train will in fact enjoy two additional morning trains. But neighborhoods will see bus wait times inch upward as certain routes are cut.

The changes, which you can find right here as a PDF, come across as minor. Some buses will see headways increased from 10 to 12 minutes. Others will see wait times go from five minutes to five minutes and thirty seconds. A few routes in some of the outer boroughs and Staten Island will see off-peak headways increase from 15 minutes to 20. Because of the addition of a few routes and the increase in J service, these changes will actually cost the MTA $300,000 a year, but they are cuts nonetheless.

According to the committee documents, the MTA is putting forth this proposal to “ensure that bus and subway schedules accurately match current rider demand and operating conditions…These changes also address the need for running time adjustments to more accurately reflect observed operating conditions.” It all sounds good, but there’s a fundamental problem of supply and demand. When it comes to public transit options, supply often drives demand. If I know a bus runs frequently and regularly, I’m more likely to take it than I otherwise would be. If Transit cuts back bus service so that trips are less frequent, it will make the bus a less attractive transportation alternative and will further drive down demand until they can cut the supply to zero. In a world where public transportation is a public good, these scalebacks are just a part of death by a thousand cuts.

You may also like


Alon Levy January 24, 2011 - 4:24 pm

Are these actual cuts, or the usual adjustments based on preexisting loading guidelines?

Benjamin Kabak January 24, 2011 - 4:25 pm

Cuts based upon both demand and load guidelines. In a few instances, they’ve increased service, but by and large, these are cuts.

Alon Levy January 24, 2011 - 4:28 pm

Alright… so no actual cuts in capacity – just a response to lower bus ridership and higher J ridership.

John January 24, 2011 - 4:46 pm

True, but Ben does have a point. If headways become high enough, then nobody takes transit unless they literally have no other option. Where I live (not even close to NYC), buses run only from about 6am to 10pm, and with 30 or 60 minute headways. Needless to say, ridership isn’t high, with the potential of waiting 29 (or 59) minutes for the next bus. And since ridership is low, they can’t really justify adding service, so it’s a catch-22.

That said, I don’t think headways are in danger of that here. These are small changes.

BrooklynBus January 24, 2011 - 6:32 pm

The MTA has the right to make minor service adjustments four times a year. There is nothing wrong with that. In fact it is good that they do that. The result of these changes should be a net zero change in operating costs.

One problem is that the net change this time is $300,000. If this happens four times a year in the same amount, then you are talking about over $1 million in cuts and you are right, it is death by 1,000 cuts.

The other problem is the relationship between supply and demand that you point out that the MTA does not see. Whenever they speak about cuts, you will never see them make a relationship to demand or revenue. They always consider operating costs in a vacuum assuming all the riders are captive and the number of trips are fixed.

This is untrue as everyone else recognizes. There is a strong relationship to levels of service and demand which the MTA fails to acknowledge. This is why they are reluctant to lengthen any route because the assumption is that additional service will just cost more and not attract new ridership. Until this philosophy by those in charge (Operations Planning and Budget) changes, the MTA is doomed to fail.

Alon Levy January 24, 2011 - 8:21 pm

If the MTA has to increase service by $1 million every year, it’ll go totally bankrupt in 2,000-3,000 years.

The effect of service levels of ridership decreases as the service gets more frequent. Cutting a bus from a 30-minute frequency to 60 can be deadly to ridership; cutting a bus from 10 to 12, which involves the same reduction in operating hours, has a small effect.

BrooklynBus January 24, 2011 - 11:46 pm

I don’t understand your first point.

Cutting a route from 10 to 12 minutes has a small effect only if the service needs to be cut. However, if the bus is overcrowded at 10 minutes, and you still cut it to 12 minutes, the effect on the route can also be deadly, because the route will not be able to keep its schedule due to increased loadings and increased bunching so the buses may actually start arriving every 20 minutes.

Alon Levy January 25, 2011 - 3:14 am

The headways were revised due to loading guidelines; crowding levels are therefore unchanged.

BrooklynBus January 25, 2011 - 8:19 pm

That makes no sense at all. If there are fewer buses, they will become more crowded with higher guidelines.

Andrew January 25, 2011 - 10:12 pm

Bus guidelines haven’t changed. Service is being reduced where ridership has dropped. The resulting crowd levels will be about the same as they were before ridership dropped.

Andrew January 24, 2011 - 8:36 pm

But the net change this time is an increase of $300,000!

(The result should not necessarily be zero. If systemwide ridership is increasing, it will generally be positive; if systemwide ridership is decreasing, it will generally be negative. These generalizations may not hold right now, since many of these changes are in response to ridership shifts due to last year’s service cuts. In particular, I’d guess that the increase on the J is to compensate for the loss of the Lower Manhattan M – I’ve ridden the J a few times recently and it’s been extremely crowded in the morning.)

BrooklynBus January 24, 2011 - 11:38 pm

I don’t know what you mean by net change unless you are comparing it to the June 27th cuts. If so, that would be comparing apples and oranges. The routine changes need to be considered by themselves, and cannot be compared which was supposedly a one time cut.

The point you raise in parentheses is interesting. Should the adjustments balance out or reflect the change in total ridership? The problem with them reflecting total ridership is that when you reduce service you are also causing ridership to reduce further.

Also, if the MTA is making a total cut to reflect total reduced ridership, how come when ridership was increased by 10%, service only increased by 3%? I don’t think it was because the capacity was already there but more because the crowding guidelines increased.

Alon Levy January 25, 2011 - 3:16 am

The MTA isn’t increasing train service one-to-one with ridership. Some service levels, for example peak trains based on capacity and night trains at 20-minute intervals, are fixed. The loading guidelines affect off-peak, non-late-night service.

The crowding guidelines changed during the service cuts, but that’s it. And they didn’t change by a huge amount – they went from 100% to 125%. Having ten standees on an R62 is not a big deal.

Andrew January 25, 2011 - 7:54 am

I’m not sure what you mean. There are loading guidelines for rush hours: 110 per 51 foot car, 145 per 60 foot car, 175 per 75 foot car. Where those are violated, it’s usually because there’s no track capacity for more trains.

In theory, the off-peak loading guidelines apply at night as well, but the overnight policy headway of 20 minutes is the driving force. (If overnight ridership were to increase on any line to the point that loads exceeded 125% of a seated load, service would be increased.)

The daytime policy headway is 10 minutes, which accounts for most of the off-peak 10-minute headways you see around the system. Loading only comes into play when headways are shorter than 10 minutes.

Alon Levy January 25, 2011 - 1:57 pm

My point is that there are peak hour loading guidelines, but the limiting factor on many lines is track capacity, as you’ve said.

BrooklynBus January 25, 2011 - 2:48 pm

Of course I understand that neither train or bus service is increased one-to-one with ridership, but I was thinking more of bus service than train service which is more inelastic than train service. You need a hell of a lot more passengers before you add an extra train, but not that many more bus passengers before you need an extra bus.

A change of 100% to 125% is quite significant for buses. You have to remember that those figures are only an average for the trip, not maximum numbers. If a load on a bus trip ranges between 25% and 175% of a seated load, the average may only be 100% of a seated load and the most time a passenger may be required to stand during the off peak may only be 5 or ten minutes. But raise that to 125%, and now many passengers may have to stand for 20 minutes or for their entire trip. Even worse, the maximum load may now reach 200% of a seated load, or full capacity requiring the bus to flag passengers at bus stops doubling or tripling their actual wait. Overcrowding also accelerates bus bunching. The effect on ridership can be quite significant. That is the danger when you are only concerned with averages.

Andrew January 25, 2011 - 6:23 pm

As i said, it depends on the specific conditions. For example, if a bus line is already operating at the policy headway (30 minutes during the day, I think, for NYCT), then reduced ridership won’t lead to reduced service, and increased ridership won’t lead to increased service unless loads are pushed above the guideline.

The change from 100% to 125% seated load was for off-peak subway service. There were no guideline changes for buses. I don’t know what the guideline is for buses, but it’s the same as it was two years ago. None of the adjustments presented here are due to last year’s service cuts, except to the extent that riders have shifted travel patterns in response (e.g., to the J to Lower Manhattan now that the M doesn’t go there).

And you are absolutely incorrect that they are an average. They most certainly are not. Guidelines apply to the peak load over a predetermined time period (probably 1 hour for buses; I’m not sure, but it may be as short as 15 minutes on the subway during rush hours).

BrooklynBus January 25, 2011 - 8:17 pm

Are you sure that the increase to 125% did not apply to buses? I thought that it did.

How could the guidelines apply to a peak load on a bus route? If so, this situation, a regular occurrence, could never happen.


Andrew January 25, 2011 - 10:30 pm

Quite sure. See http://www.mta.info/mta/news/b.....0-nyct.pdf. “Revisions to Off-Peak Service Guidelines” is listed only under Subway Service.

Loading guidelines always refer to peak loads. Otherwise any subway line that’s much busier for the central section than at the ends would have to carry loads well in excess of crush capacity (which is physically impossible, by definition) before the average load would come close to guideline.

I’m not sure what that has to do with your B1 example. You yourself acknowledge that there are enough buses to handle the load (“I don’t think more buses are the answer in this particular case”), only that the passengers aren’t moving to the back.

Andrew January 25, 2011 - 6:50 am

I am considering these routine changes by themselves. You must have misread the article. These routine changes come at a cost of $300,000. Although fewer in number, the cost of the upward adjustments exceeds the savings due to the downward adjustments.

A small increase in headways will cause ridership to decrease slightly. And a small decrease in headways will cause ridership to increase slightly. Your point?

Service is increased or decreased as necessary to stay within the loading guidelines. I was only pointing out a general trend, that increased ridership usually leads to upward adjustments and decreased ridership usually leads to downward adjustments. That isn’t always the case – for instance, increased ridership during periods when there’s spare capacity on the train or bus, or decreased ridership during periods when the train or bus is operating on a policy headway, won’t lead to any adjustment at all. The numbers certainly won’t correspond!

BrooklynBus January 25, 2011 - 2:36 pm

Why do you say the changes come at a cost of $300,000 when the report says the MTA will save $350,000 as a result of these changes?

Benjamin Kabak January 25, 2011 - 2:38 pm

The news reports are incomplete and misleading. The changes I’m talking about here — bus schedule adjustments based on load guidelines and two additional runs for the J train — will cost the MTA $300,000. A different set of changes to a few Staten Island bus routes that were considered a different action item will save the MTA $350,000. It’s all in the PDF I linked to in the post.

BrooklynBus January 25, 2011 - 3:11 pm

I was referring to the PDF not news reports, but I read it too quickly. Now I understand.

So the entire plan seems cost neutral (a savings of like $17,000 a year.)

The problem I now have is the MTA taking from the buses to give to the trains. The changes should have been cost neutral just within the bus system. You are helping riders on one subway line by punishing riders all over the city on many bus lines. I don’t remember this having been done before.

Another question — Prendergast stated that the subway fares cover a higher portion of their costs than the buses cover. Do you know anything about how they determined that? It just doesn’t seem right. I realize there is more labor involved with the buses, but subways have an infrastructure to maintain while the buses do not. Do you know if they are considering the cost to maintain the subway stations and ROW in their calculation of subway costs or just equipment costs?

Andrew January 25, 2011 - 6:33 pm

There’s no “entire plan.” There are two independent things going on. One is a routine adjustment of bus and subway headways, which will cost $300,000. The other is a service change, which will save $350,000. The two are in no way linked.

Nobody’s taking from the buses to give to the trains. Buses and trains are not in competition. Service is being reduced where ridership has dropped and is being increased where ridership has risen. There’s no bias for one mode or against the other. Why should more buses run than are needed while J trains are overcrowded?

Labor is the driving force – a crew of 1 per busload vs. a crew of 2 per trainload. Once the line is in place, any well used train line is much cheaper to operate than the equivalent bus line. I believe the typical calculations include operating costs but exclude capital costs. Routine maintenance is usually covered by the operating budget; major overhauls are covered by the capital budget. But street maintenance isn’t included in the bus calculations, since it’s funded by the city.

BrooklynBus January 25, 2011 - 8:31 pm

I couldn’t reply to your other post, so I will respond here. Of course they are taking from the buses and giving to the trains. If J service is being increased and other subway lines are not being reduced, and the net change is near zero, the change has to be coming from buses.

As for your allegation that more buses should not run than needed, I doubt that would be he case. I was told by Bus Operations several months ago that they would put in a request to increase B1 service, yet B1 schedules are remaining unchanged, although drivers can’t meet the current schedule now because it is too tight.

These revisions need to be more transparent. Trust us is not good enough.

Andrew January 25, 2011 - 10:35 pm

They are taking from the lines that have reduced ridership and giving to the lines that have increased ridership. It so happens that, this time around, one of the lines with increased ridership is a subway line. There’s no bias toward one mode or against another.

You are asking for the schedules to be rewritten to allow for more running time. That has nothing to do with loads or guidelines. I don’t know how often NYCT reevaluates running times (it’s probably on a rolling basis systemwide, given the number of bus routes) or how long the process takes, but it’s a completely different process from increasing or decreasing service in response to ridership fluctuations.

BrooklynBus January 26, 2011 - 11:31 am

Have they ever taken service from subways and given it to buses? I don’t think so.

Rewriting running times is a different process, but the effect is the same. If running time is added, it now costs more to run the same service, at least on paper, which is all the MTA is concerned about. Therefore they are very reluctant to add running time. They also only consider traffic, not heavy passenger loads, which also slow down buses, when revising running times, which is why they are often inadequate.

Andrew January 27, 2011 - 10:32 pm

For the third time, there’s no competition between modes. Has bus ridership ever increased while subway ridership was decreasing? (In the 70’s, perhaps?) That’s when you’d find bus service increases at the same time as subway service decreases. In the late 90’s, ridership was increasing on both the bus and subway systems, but on the bus system at a much greater rate (due to MetroCard), and service was increased accordingly.

Calibrating running times is completely unrelated to adjusting the supply to match the demand. Yes, increasing scheduled running times and increasing service frequencies both cost money, but for different reasons, and the resulting schedules look totally different. Say your favorite bus line has a 60-minute running time and a 10-minute headway. Revising the running time might give it a 70-minute running time and a 10-minute headway. Revising frequencies might give it a 60-minute running time and an 8-minute headway. They both require more buses, but they’re not the same thing.

Revising frequencies is what’s done to ensure that passenger loads aren’t overly heavy (or light).

BrooklynBus January 26, 2011 - 11:57 am

I checked the link you cited. It says nothing about the 12 to 16 standees being at the peak point on the line. The way I recall it was that the load had to be sustained for a period of 15 minutes. In other words reaching a load of in excess of 16 passengers for one or two stations would not trigger an increase in service. It would have to be maintained from say Brooklyn Bridge to 86th Street for example.

I also sent them an e-mail requesting where on the website is the information regarding planning guidelines. Response time is 15 days, but I doubt they will answer because the info does not seem to be available on the website.

Also, when the request was made to increase off peak loads from 100% to 125%, I do not remember them stating that this would only apply to subways and not to buses, like you claim. Just because it isn’t cited in the service reduction report does not mean they have not been revised for buses as well.

Andrew January 27, 2011 - 10:43 pm

Loading guidelines are always compared to loads at the peak load point. That’s basic Transit Scheduling 101 material. It’s not mentioned in the service reductions book because there’s no change in where the loads are taken. The book is an outline of proposed changes, not a textbook.

There’s certainly no requirement for the load to be sustained for 15 minutes – very few peak loads are maintained for more than one stop! – but I think I know where your confusion is coming from. Peak loads are always averaged over a period of time. That is, rather than focusing on the peak load on a single bus (which may have an atypical load due to bunching or connections or whatnot), the peak loads are averaged over each bus during a predetermined time – 15 minutes, 30 minutes, an hour, etc. But the numbers that enter into the average are peak loads!

If you hear back about the loading guidelines, please post them – I’d like to see them.

The book of proposed service changes is a legal document. The public was not asked to comment on a proposed revision to bus loading guidelines; the MTA can’t just go ahead and revise them. If you scroll down to page 4, you’ll see everything spelled out in detail, including a list of affected subway routes. Buses are not affected, because the proposal was only for the subway.

Kai B January 24, 2011 - 5:12 pm

Of the five bus lines that serve my neighborhood (Greenpoint), it looks like four are being left alone and one, the B43, is being increased.

I too prefer to call this an annual adjustment rather than a cut. Especially in light of the fact that it involves increased spending.

Sean January 24, 2011 - 10:21 pm

While it is true that the less convenient it is, the more it discourages people to use the system however just as you assume that b/c there is less frequent service people will just drive or not use the service, you also assume that people have alternatives. I believe they have the upper hand where most of the riders of the MTA have no alternative. As far as the MTA going bankrupt, that would probably be the best thing. The pensions and horrible union contracts are the number 1 killer of the budget, besides the overlapping jobs and high executive pay.

Al D January 25, 2011 - 11:16 am

If the ‘gypsy cabs’ are legalized and regulated, with a certain, fixed fare structure, bus ridership could decrease further because there would be a more legitimate, realiable alternative.


Leave a Comment