Wednesday, February 18, 2015

To Save Money on Marketing Buses, Try Running Enough Buses

I like bus improvements. I hate Governor Cuomo's terrible proposal to run an elevated "AirTrain" from Willets Point to LaGuardia Airport. So you'd expect that I would like Josh Barro's post last week arguing that we should improve the buses to LaGuardia instead of building the AirTrain. He writes, "Transit agencies are spending millions of dollars on new rail infrastructure that is no faster than existing bus service, simply because riders perceive a train as better than a bus." And then he goes on to make a nice argument that we should tell these poor deluded people that the train is actually not better.

This is wrong nationwide, but it's wrong on even more levels in the case of transit to LaGuardia. This is because people are already riding the bus to LaGuardia. On the first weekend of the M60 Select Bus Service I rode the bus, and it was packed. Since then I've ridden it twice more, and both times it was crush-loaded. The Q70, Q72 and Q48 aren't quite as heavily packed, but they have very healthy ridership. The Q70 probably gets even higher ridership than I give below, because it had only been in service for three months by the end of 2013.

2013 average weekday ridership17,013279057643716 (August 2014)*
2015 weekday buses139627196
Average riders per bus122458139
Loading capacity112707070
Seated capacity62404040

The M60 is packed, and the other buses are pretty full. If I were wealthy, or if my employer were paying, I would take taxis over the M60 almost all the time. I would probably take taxis over the Q72 or the Q48 as well; the Q70 experience is the only one that has been close to comfortable for me.

Why does the MTA not run enough M60 buses to bring the loads down to reasonably comfortable levels? I have no idea. but imagine that someone did what Barro suggests and spent a ton of money on "marketing" these buses. Imagine if that marketing succeeded in attracting the 70-90% of people who currently arrive by taxi or private car (PDF). The MTA would not be able to serve the people that they attract. They would have a horrible time and take a taxi from then on out.

Barro frames this with a quote from the National Bus Rapid Transit Institute, "Bus-based public transit in the United States suffers from an image problem." Yes, the BRT people keep repeating that buses are just as good as trains, and everyone just needs to be shown the light, but notice two things. First, the actual report (PDF) that Barro drew the quote from gives a much more nuanced picture and hardly makes a strong case that marketing is all you need. Second, this report and Barro's post, and this lame entry from EMBARQ a couple weeks ago, are just three more in a long line of bus scoldings where someone patronizingly tells you to love your bus without showing any interest in taking the bus themselves.

When Barro first tweeted the link to his post, I responded by telling him that the M60 is frequently packed. His response to me was simply, "even more reason not to spend $1 billion on a train." Well, I don't know about a billion dollars, but as Stephen Smith frequently reminds us, high bus ridership is actually one indicator that a potential train line is worth spending money on.

What bothers me most about Barro's piece is how since he posted it on February 10, several people have uncritically cited it as either an argument for more bus marketing, or an argument against subway expansion. It is neither, because it is based on inaccurate information. I hope that Barro will post something correcting those mistaken impressions as soon as he can.

* Thanks to @AHInQueens for the Q70 ridership figure.


itineranturbanist said...

I used to ride the M60 back in college. It had several challenges: 1) with 125th being so slow, it had huge bunching problems, seemingly regardless of frequency. 2) It was often crazy crowded, and as several people have noted only 10% of ridership goes to the airport. 3) At the time (this has since been remedied) most buses did not have luggage racks, which was a PITA for a college student.

Honestly, the M60 would be most useful as airport transportation if it were boarding-only eastbound in Manhattan. That would free up a ton of capacity for extra airport passengers immediately. There's still demand for Harlem--Queens service, though, so that's not necessarily a desirable result.

If MTA really wanted a dedicated airport bus, they could do a premium fare, boarding-only eastbound, with one stop at Astoria Blvd. in Queens. But I'm not sure that would be a great result, and yeah, basically, there should be a subway along 125th, over the Hell Gate, and to Astoria/LGA.

Capn Transit said...

Good points, Sandy, but all of the crush-loading I observed was east of Steinway Street in Queens. If that's only ten percent of the ridership, it must be hell in Manhattan!

I think building out the dedicated bus lanes the rest of the way on 125th Street would help, but in the RPA study (PDF) there was a map showing that most LGA passengers are coming from Midtown, so I still think extending the N train is the best solution.

Andrew said...

Your math is incorrect, since it's based on the average number of riders per bus over the length of the run, not the average number of riders per bus at the peak load point. On the Q70, it's close enough, since virtually every rider is on the bus between Jackson Heights and the airport (i.e., the number of riders between Jackson Heights and Woodside, or between airport terminals, is negligible), but on the other three it's way off. The M60, for instance, may carry 122 riders per bus, but there's quite a bit of turnover as the M60 proceeds along its route, so it doesn't carry all 122 riders at any single point on the line.

Your denominator is also off, since it looks like you're using crush capacities rather than anything remotely sustainable. You can find NYCT's guideline capacities at (page 23, and I think MTA Bus now uses the same guidelines), although they vary by headway, so a simple division across the entire day won't work even if you do get your hands on the peak loads.

Capn Transit said...

Andrew, I recognize that the M60 is not carrying 122 riders from end to end, 24 hours a day, but that bias and the crush capacity estimates work against each other. Even if the ridership is significantly lower off-peak and at the less popular points of the route, it's still standing room only.

I've ridden all four of these bus routes multiple times and it's only on the Q48 and Q72 that I've seen any significant drop in ridership at a certain time or place. The M60 and Q70 seem to be pretty full round the clock.

These averages still support my main point: people are already riding the bus to La Guardia, and the MTA isn't running enough buses to keep them comfortable. In this environment, marketing the buses in an effort to get more ridership is pointless and in fact counterproductive.

AlexB said...

I live in Astoria near the Astoria Blvd/31st St stop on the M60. I use the bus pretty regularly, but would use it a lot more if it were more reliable and less crowded. The bus is always packed and is usually bunched. You wait 20 minutes and 2 buses arrive at once. The bustime app doesn't even show all the buses and they appear and disappear from the next arrival list. It's definitely improved over the last year or two, but based on the crowds, I see no reason why that route shouldn't be running at 5 minute intervals. It's a work in progress at best.