As far as bus improvement efforts go, I’ve long maintained that the city’s and MTA’s Select Bus Service is something of a charade. It has so far taken these planners four or five years to identify a route, plan the service, hold the requisite community meetings, bid out the work, build the infrastructure and launch service along what is essentially a glorified express bus line. This isn’t Bus Rapid Transit with fully dedicated rights of way and constant service; this is New York’s “we have to please everyone all the time but especially drivers” middle-of-the-road stumbling toward transit upgrades.
From an assessment perspective, one of the frustrating elements has been the sheer lack of data made public about the success (or perhaps the failures) of the Select Bus Service routes. Are these improvements decreasing travel time while increasing ridership? If rides are faster, why? What is the effect of an SBS route on parallel local bus routes? Recently, in a report on the B44, some of these questions were answered, and the results highlight two simple reasons why buses are faster. We’ll get to those in a minute.
The B44 was one of those long-drawn out SBS routes. It debuted in late 2013 along Brooklyn’s third busiest bus corridor, and it’s a success story in a vacuum. According to the report [pdf], travel times have decreased by 15-31 percent depending upon the time of day and ridership is up 10 percent. Traffic crashes are down as well. Meanwhile, ridership has decreased on the local buses by only four percent, suggesting that the SBS route is a net gain. We don’t know how overall local travel times are affected by the shift in service though so it’s tough to analyze the overall impact.
Officials were pleased. “The B44 SBS along Nostrand Avenue is a tremendous success story, among the biggest successes in the eight years that DOT and MTA have coordinated Select Bus Service,“ DOT Commissioner Polly Trottenberg said.
But this is a lot of back-slapping for two simple infrastructure improvements. Overall, SBS B44 travel times have decreased by an average of 17 minutes, end-to-end, but times in motion haven’t changed. Buses move for around 37 minutes out of every hour, but instead of sitting at stations for nearly 26 minutes and stuck in traffic for 20, buses are stopped at stations for 15 minutes and stuck in traffic for only 12 minutes. Why? Pre-board fare payment, dedicated bus lanes and signal prioritization. It’s not exactly a secret combination, and improving bus service is as simple as that.
DOT and the MTA have made better bus service into a big deal and something that warrants special consideration during the planning process and special treatment after. It involves branded buses, painted lines and special infrastructure. But it shouldn’t. It should just involve the recognition that buses shouldn’t be subject to the whims of surface traffic through busy corridors and that our fare payment system is horrendously antiquated and inappropriate for city buses. If DOT and the MTA wanted to, they could improve bus service tomorrow by significant amounts simply by giving buses their own lanes, and the fare payment problems should be a part of whatever comes along to replace the Metrocard.
For all the handwringing about declining bus ridership and the need to expand transit access, the answers are right in front of our collective faces. That DOT and the MTA haven’t been aggressively pushing these measures is a stain on their records that deserves a closer look. Improvements and faster travel times don’t need to come through such a torturous process.
I’ll say this much, the Q44 SBS has been surprisingly successful at being a fast bus route when compared to the Limited service that came before it. It only ever really hits a snag around Downtown Flushing, and that’s just how Downtown Flushing is.
You might be looking at this wrong way – they reduced the trip time from 83 minutes to 65 minutes. A quick skim of the lushly-produced PDF didn’t list total route cycle time, but if the %age change on the opposite-bound part of the run was similar the productivity of the operator goes up and the utilization of the vehicles does up. Those aren’t things to sniff at.
I agree, for all the time and effort this took to implement, there wasn’t much payoff in *moving* time but again, those other reductions are real and useful. A better question might be “why isn’t MTA doing this for all the routes in a particular corridor?” No, you don’t have the protected lanes, but the signal priority and OBFC should be transferable to other routes. And with very little additional cost (if any).
There are now a bunch of SBSs running. Is there any evidence that the planning process has gotten shorter? 4-5 years for a service without dedicated lanes (though one that does include other construction) seems suspect.
Perhaps I didn’t make the point appropriately: Reduction in moving time doesn’t matter. Reduction in stopped time so that overall trip length is shorter is what matters, and to that end, the three improvements — dedicated lanes, signal prioritization, pre-board fare payment — are responsible.
the three improvements — dedicated lanes, signal prioritization, pre-board fare payment — are responsible.
Don’t be so quick to jump to conclusions.
Dedicated lanes and signal prioritization should have benefited the local buses too. Did they? Not according to the MTA’s schedule. The local’s average speed was 7.3 mph in the LTD days. It has hovered between 7.1 and 7.2 mile during the time SBS has been in operation. The passenger count has roughly been constant during this period. It looks like a wash. Dedicated lanes and signal prioritization have not made a dent in the local’s travel time.
It’s likely they did not contribute to the SBS reduction in travel time because dedicated lanes and signal prioritization help buses in motion. There’s no difference between a local and SBS bus in motion. If these features didn’t help the local, it’s not likely they helped the SBS bus.
That leaves prepayment. That’s an easy parameter to estimate. Prepayment helps only when boarding the bus. The exit time remains the same. Published data estimate the prepayment time saving at about 2 seconds per passenger. Multiply this by the average number of passengers on a bus trip and you have an estimate prepayment’s average saving per trip.
In round numbers there are 18,000 daily passengers on the B44-SBS and there are approximately 300 daily trips. That comes to 60 passengers per bus. That’s an average 2 minute estimated savings due to prepayment per B44-SBS trip. The average time savings per B44-SBS trip was 8 minutes (adjusted for shorter travel distance).
What about those remaining 6 minutes? If it wasn’t dedicated lanes, signal prioritization and prepayment, what caused the bulk of the dwell time reduction?
One obvious explanation besides falsifying data is that were fewer passengers on each bus. Fewer passengers on each bus means fewer passengers getting on and off which means less dwell time. However, total passenger counts were relatively stable.
There is an obvious method to decrease the number of passengers per trip while keeping the total passenger count stable. That’s to increase the number of bus trips. Did they?
The MTA’s GTFS schedule data provides the answer. They scheduled 292 LTD trips on 31 Oct 2013. The number of scheduled SBS trips was 328 on 12 Dec 2013. It stayed at 328 until 22 Apr 2014. It rose again 328 on 10 October 2014 and remained at this level until 25 Jul 2015. It bounced back to 329 on 14 Oct 2015 and remained at this level until the end of 2015. The study was of 2015 performance. There were 328 scheduled daily trips for most of the year. That’s a 12% increase scheduled trips without any increase in passengers. This would result in 11% fewer passengers per trip and an expected 11% drop in dwell time per trip.
Applying this 11% to the remaining 64 minutes from the LTD trips, reduces this number by 7 minutes to 57 minutes. The average SBS trip was 55 minutes or 2 minutes quicker. That’s due entirely to the prepayment time savings. The rest of the reduction was due to shortening the route and adding more trips. The dedicated lanes and signal prioritization did not make any contribution, confirming why the locals received no benefit from them.
Local buses that stop every block or two have a hard time benefiting from signal progression/priority compared with buses that actually move some distance between stops.
Correct me if I’m wrong, but I remember hearing that the B44 and B44 SBS don’t run the same route. This might explain why signal prioritization doesn’t benefit the local.
“You might be looking at this wrong way – they reduced the trip time from 83 minutes to 65 minutes.”
They also reduced the average distance traveled by the SBS buses vs. the LTD’s. That wasn’t factored into the analysis of reduced SBS running time.
If you analyze the GTFS data that the MTA distributes regarding their bus schedules, the picture is not quite the win. The MTA started issuing the computer schedules on 10/30/2013. That was the last month before the switch to SBS.
The 31 Oct 2013 schedule shows the average B44 LTD trip was 9.1 miles and took 71 minutes. The 12 Dec 2013 schedule shows the average B44 SBS trip was 8.2 miles and took 56 minutes. The percentage drop in scheduled trip time was 21%. That’s roughly the same percentage reduction that was shown in the bar graph.
However, the average SBS trip was going only 90% as far. To compare apples to apples, the LTD trip time should have been discounted by 10% or 7 minutes to account for the shorter distance. This means the trip savings attributable to SBS was only 8 minutes, from 64 to 56 minutes or 13%. That’s closer to a scratch single than the advertised grand slam.
Another distortion of the DOT/MTA statistics that I missed.
What is the point in comparing average travel times of all trips? Your analysis penalizes MTA’s change to the route that reduced the number of trips that traveled south of Ave U. If the MTA data sampled only average travel times of comparable trips (Full length trips from Williamsburg to Sheepshead Bay), it appears their data checks out.
Sure, it’s not enough, but it is a real gain. I’m a frequent SBS rider (m34/m34a and m15) and I can tell you that it’s a real improvement over local service.
But here’s the key difference between half assing it with SBS and half assing it on the subways. When MTA half-asses things on the subways, it is stuck with that decision for all time, because building new tunnels is too expensive. With SBS, they can always build on what they’ve already done and finish making dedicated lanes and building more bus bulbs.
But you need to compare SBS to the Limited that proceeded it, not the local. How many former Limited passengers are now using the local and gave slower trips now? That is not discussed in the report.
The most important factor is passenger trip times, not bus travel times and the report says nothing about passenger trip times.
And as far as adding more dedicated lanes, there is no room for more. Other than adding one or two more SBS stops, the only other way to increase SBS patronage is to extend every other SBS one more stop to Kingsborough College as I suggested to the MTA over a year ago.
Ben, I guess you didn’t read the Progress Report too closely or was mislead by DOT. They did not say that ridership rose by 10 percent. That would be great if it did. They said that SBS ridership rose 10 percent during the second year and the local lost 4 percent. So the net gain was only 6 percent.
They also said there was a decline in B44 ridership the first year but intentionally do not tell how large that decline was. The annual ridership statistics show it to be 8 percent. So after all is said and done “ridership” did not rise by 10 percent. In fact B44 ridership is now 2 percent lower than it was before SBS started.
Throughout the entire report, DOT attempts to mislead. The report is plagued with numerous omissions. They try to imply that first year ridership was not favorable due to ongoing construction. How do we not know that SBS ridership declines during the first year was not due to massive fare evasion and little enforcement which was obviously stepped up the second year but which will be lowered again as additional SBS routes roll out?
So we have a route that cost $2 to $3 million more to operate annually, cost like $20 million to implement and after two years has lower ridership than before SBS. How in any world, can that spell success?
You also did not look at my 9 page critique of the B44 Progress Report and one page summary that I e-mailed you last week. It is available here:
I have written numerous B44 SBS articles over the years, and all my predictions regarding how this route would fail have come true. The success of many other SBS routes are also in doubt. M15 ridership is much lower today than before SBS debuted there in 2010. And that is the route the MTA was always pointing to in order to justify other SBS routes.
“You also did not look at my 9 page critique of the B44 Progress Report and one page summary that I e-mailed you last week.”
I mean, would anyone blame him for not reading it? I’m sure it was filled with incoherent arguments, poor analysis and typos as well as some pro-car BS you hide under the guise of transit advocacy.
Typical response from you. You have already arrived at your conclusions before even reading my report. Why is it that SBS supporters refuse to examine all sides of the issues for fear of having their minds changed.
But you are right about one thing. I did find two typos which I will correct. I mistakenly reported the borough decline for two years was 1.9 percent when that was the last years decline, so I concluded the B44 performed similar to other non-express routes. However even if you use the correct borough decline for two years of 3.5 percent, you really can’t conclude that the B44 performed much better than your typical Brooklyn route because it only lost 2 percent.
That is because of the change in northbound routing of the B44 for SBS buses to use the same street as the B49 (Rogers Avenue) for the entire length of the street. So anyone boarding at an SBS stop may also be able to use the B44 SBS in addition to the B49. So some of the increased B44 patronage is due to the change of route causing riders to take which ever bus comes first, stealing B49 riders. In fact B49 patronage declined in excess of 10 percent since 2013. So you really have to consider both routes as a group unless you can determine which passengers boarded on Rogers Avenue. When you consider this, the B44 did not perform that much better than other routes.
As for your accusation about pro-car BS, there is on,y about one paragraph out of 8 1/2 pages devoted to car travel times which according to you need not even be mentioned because only bus riders need to be considered.
Anyway, if you want to get into actual numbers, overall B44 ridership is down 2% from 2013 while overall Brooklyn bus ridership is down approximately 3.5% over the same time period. It’s a bleed and it’s a questionable investment considering you can get time savings from low-cost improvements (which was Ben’s point), but your claim that it’s a resounding failure defies reality.
No it is not a questionable investment. It was a downright bad investment. When your goal for SBS is to reverse the decline of ridership when subway ridership is soaring, and ridership still is declining on an SBS route after SBS investment, it is a resounding failure. If costs were the same, it wouldn’t be so bad. But SBS costed $20 million to implement and $2 to $3 million extra each year to maintain on the MTA side, not even considering the DOT side, (maintenance of lane markings and signage). And what happens to all the millions spent on far equipment when we switch to contactless payment in a few years? It is much more than a questionable investment.
“And what happens to all the millions spent on far equipment when we switch to contactless payment in a few years?”
I’ve wondered about this. It goes without saying that we should already have contactless payment by now, and the delays are costing us millions in additional spending on the MetroCard. Hopefully these machines can easily be converted to smart cards, but they would not be necessary at all with simple tap readers inside each bus door. At the same time, we shouldn’t have to wait years for the new farecard to get better bus service.
They said that SBS ridership rose 10 percent during the second year and the local lost 4 percent. So the net gain was only 6 percent.
Wow … no. You have to look at the actual numbers, not the percentages.
Total ridership, SBS and local combined, rose 2.8%, not 6%
I did look at actual numbers. Since in 2014 ridership was pretty much equally split between SBS and local, it works out to six percent which is the same number DOT uses.
Looks like you are also quick to criticize without even reading the DOT report or my response.
When you think of it that the SBS buses are 50 percent longer and service is more frequent than the local, yet just as many prefer the local to SBS and it is way more crowded, speaks volumes what riders think of SBS. The first year there were complaints of routine 40 minute waits on the local, yet more still used the local than SBS if SBS numbers were not strongly influenced by fare evasion which could have been the case.
Clarification: My quote is using weekday averages which are the only numbers the MTA provides in the report. So the weekday average during the past year is up by about 6 percent.
But when you look at annual ridership, B44 ridership is only up by 2.5 percent the past year. But that still ignores the fact that some B49 riders on Rogers will take the B44 SBS only because it arrives first.
Which buses can use the signal prioritization feature ? Is it just the SBS buses or is it all buses ? The SFMuni version of light-grabbing requires an on-vehicle transmitter as well as a sensor, usually mounted on a stoplight, that detects the bus’s signal. Not all of the buses in SF have the requisite hardware.
Parking/muni meters can take credit cards – why not sidewalk sbs fare machines?
Come to think of it, could there not be a savings if fare machines/muni meters were combined – stores and vending machines sell multiple products from multiple manufactures why not these machines?
Having the machines take creditcards (even if a 3% or 10 cent surcharge cover the creditcard fees is added) would be a major convenience to many riders and could actually encourage ridership from those who do not have a metrocard or exact change.
The most ridiculous part is how the coin fare machines are basically muni meters with the credit card reader filled in with plastic. They’re the same machine, capable of doing all the same things, but the MTA chose not to allow it.
I think it means the MTA’s machines aren’t actually connected to anything (they just accept currency and spit out paper).
I just hope they’ll finally make Avenue R an SBS stop.
The Progress Reports states they have rejected that idea because of the low transfer volume at Avenue R. That decision is just as biased as the rest of the report because transfer volumes should not be the only factor considered. They needed to consider all passengers boarding at Avenue R but also those boarding at Quentin Road including B100 transfers and those boarding at Avenue S as well because if Avenue R were made an SBS stop, those riders would walk a block and use the stop as well.
The fact that the closest SBS bus stop is a half mile in either direction should also have been a factor in their decision. They also used the excuse about residential driveways being in the way of a longer bus stop when many bus stops have commercial driveways in them. Commercial driveways are worse because they are more active than driveways in front of a single family house.