4 ms·

I'm skeptical for many reasons. First, where is he getting that car commuters outnumber rail commuters (just barely) in New York? Last I heard the split was 60

I'm skeptical for many reasons. First, where is he getting that car commuters outnumber rail commuters (just barely) in New York? Last I heard the split was 60/40 in favor of rail. Second, where did he get that it is unusual for management to not use a car? Rather famously Bloomberg himself used to commute to city Hall by subway. By the authors own admission, even the lawyers use the subway. New York doesn't have that weird anti-transit classism of the rest of the country. Third, what is meant by multi modal and why is it a relevant figure if merit. Fourth, he clearly used New York metro area, not the city proper in his calc. Can I get some footnotes on if the cross comparison did the same thing to other cities? Fifth, 10 min headways is exactly the threshold where you can just show up and not pre plan your trip. When you get to the station you will wait on average five min. Compare that to the amount of time spent looking for parking or waiting for an Uber. I am unsatisfied with the conclusion that of peak wait times are the reason for underutilization or that there even is an underutilization problem in the listed American cities.

Symbiote 7y ago
10 minute gaps sounds like a long gap for a city the size of New York. Most metro lines in London have 2½-5 minute gaps off-peak. 10 minutes is more like the waiting time for the next train from London to Birmingham, after the morning peak: https://traintimes.org.uk/london/birmingham/1000/monday https://traintimes.org.uk/london/birmingham/1000/monday
- jkaptur 7y ago
  Many of those gaps take place at night. Last time I visited London, I was nastily surprised to find that the tube has a several hour gap during that time.
resonantjacket5 7y ago
I am not sure why you're skeptical about any of these stats. A visit to any modern metro system outside of America will reflect what Alon says. 10 min headways while decent in America is still kinda annoying. Note that the 10 min headways is really more like 15 when taking into account American transit's unreliability in actually adhering to the schedule. That being said, I don't think the lack of off-peak usage is because of low frequency, but rather lack of density on the suburban stations. If there's only offices at the core stations then people can only really use it for going to work.
jcranmer 7y ago
> First, where is he getting that car commuters outnumber rail commuters (just barely) in New York? Last I heard the split was 60/40 in favor of rail. He's getting that number from American Community Survey data, and it's including the full MSA, which includes the entirety of Long Island and half of New Jersey, even if you don't work anywhere near NYC itself. The 60/40 split you're talking about is more representative of the mode share for people who live in NYC proper. > Second, where did he get that it is unusual for management to not use a car? Rather famously Bloomberg himself used to commute to city Hall by subway. He's talked with several staff at the NYC MTA, and it's fairly well-known that most of the upper management at the MTA don't use its own services. Bloomberg is an exception, not the rule. > Fourth, he clearly used New York metro area, not the city proper in his calc. Can I get some footnotes on if the cross comparison did the same thing to other cities? The author does spend a lot of time trying to compare like-to-like. Official city boundaries are notoriously bad indicators of urbanization--Staten Island is basically a suburb within NYC boundaries, to say nothing of places like Houston, while Boston and Washington, D.C. have core parts of their key business districts just outside official boundaries. The official metric that tends to be most comparable across countries is various approximations of the MSA.
jakobegger 7y ago
I agree that 10min is a threshold. We have a bus line nearby that has a 15min interval. It's close to useless for short trips, and I only use it rarely. The busses with 10min interval are okay. The average wait is still longer than I'd like, but it's not quite as frustrating. But a 5 minute interval should absolutely be the goal. It's probably not possible for many bus lines because of traffic, but for rail and tram lines its not a problem. With a 5 minute interval public transport becomes more appealing than driving. One reason why wait time is so important is that people often need to change lines. If you want to take two busses with a ten minute interval, you have to leave 20min early to make sure you get there on time. That makes driving a lot more attractive...
- ummwhat 7y ago
  The thing is in parts of New York City 5 min gaps aren't possible in the rail lines even at peak. The reason is traffic in a manner of speaking. The New York City subway uses extensive interlining which means that two or more services share track for part of their route (typically the part in Manhattan). If trains on the shared track are every 3 min, by conservation of mass, trains on the split part need to be every 6 min. NYC also is still upgrading it's signaling system so for now train are a 5 min interval even in the best parts and 10 min intervals everywhere else. There is physically no easy fix.
bobbiechen 7y ago
Tiny nitpick, you'll probably wait longer than five minutes: >In the case of a nominally 10-minute bus line, sometimes the span between arrivals will be longer than 10 minutes, and sometimes shorter, and if you arrive at a random time, you have more opportunities to encounter a longer interval than to encounter a shorter interval. And so it makes sense that the average span of time experienced by riders will be longer than the average span of time between buses, because the longer spans are over-sampled. https://jakevdp.github.io/blog/2018/09/13/waiting-time-paradox/ https://jakevdp.github.io/blog/2018/09/13/waiting-time-parad...

Symbiote 7y ago

10 minute gaps sounds like a long gap for a city the size of New York. Most metro lines in London have 2½-5 minute gaps off-peak. 10 minutes is more like the waiting time for the next train from London to Birmingham, after the morning peak: https://traintimes.org.uk/london/birmingham/1000/monday https://traintimes.org.uk/london/birmingham/1000/monday

jkaptur 7y ago
Many of those gaps take place at night. Last time I visited London, I was nastily surprised to find that the tube has a several hour gap during that time.

resonantjacket5 7y ago

I am not sure why you're skeptical about any of these stats. A visit to any modern metro system outside of America will reflect what Alon says. 10 min headways while decent in America is still kinda annoying. Note that the 10 min headways is really more like 15 when taking into account American transit's unreliability in actually adhering to the schedule. That being said, I don't think the lack of off-peak usage is because of low frequency, but rather lack of density on the suburban stations. If there's only offices at the core stations then people can only really use it for going to work.

jcranmer 7y ago

> First, where is he getting that car commuters outnumber rail commuters (just barely) in New York? Last I heard the split was 60/40 in favor of rail. He's getting that number from American Community Survey data, and it's including the full MSA, which includes the entirety of Long Island and half of New Jersey, even if you don't work anywhere near NYC itself. The 60/40 split you're talking about is more representative of the mode share for people who live in NYC proper. > Second, where did he get that it is unusual for management to not use a car? Rather famously Bloomberg himself used to commute to city Hall by subway. He's talked with several staff at the NYC MTA, and it's fairly well-known that most of the upper management at the MTA don't use its own services. Bloomberg is an exception, not the rule. > Fourth, he clearly used New York metro area, not the city proper in his calc. Can I get some footnotes on if the cross comparison did the same thing to other cities? The author does spend a lot of time trying to compare like-to-like. Official city boundaries are notoriously bad indicators of urbanization--Staten Island is basically a suburb within NYC boundaries, to say nothing of places like Houston, while Boston and Washington, D.C. have core parts of their key business districts just outside official boundaries. The official metric that tends to be most comparable across countries is various approximations of the MSA.

jakobegger 7y ago

I agree that 10min is a threshold. We have a bus line nearby that has a 15min interval. It's close to useless for short trips, and I only use it rarely. The busses with 10min interval are okay. The average wait is still longer than I'd like, but it's not quite as frustrating. But a 5 minute interval should absolutely be the goal. It's probably not possible for many bus lines because of traffic, but for rail and tram lines its not a problem. With a 5 minute interval public transport becomes more appealing than driving. One reason why wait time is so important is that people often need to change lines. If you want to take two busses with a ten minute interval, you have to leave 20min early to make sure you get there on time. That makes driving a lot more attractive...

ummwhat 7y ago
The thing is in parts of New York City 5 min gaps aren't possible in the rail lines even at peak. The reason is traffic in a manner of speaking. The New York City subway uses extensive interlining which means that two or more services share track for part of their route (typically the part in Manhattan). If trains on the shared track are every 3 min, by conservation of mass, trains on the split part need to be every 6 min. NYC also is still upgrading it's signaling system so for now train are a 5 min interval even in the best parts and 10 min intervals everywhere else. There is physically no easy fix.

bobbiechen 7y ago

Tiny nitpick, you'll probably wait longer than five minutes: >In the case of a nominally 10-minute bus line, sometimes the span between arrivals will be longer than 10 minutes, and sometimes shorter, and if you arrive at a random time, you have more opportunities to encounter a longer interval than to encounter a shorter interval. And so it makes sense that the average span of time experienced by riders will be longer than the average span of time between buses, because the longer spans are over-sampled. https://jakevdp.github.io/blog/2018/09/13/waiting-time-paradox/ https://jakevdp.github.io/blog/2018/09/13/waiting-time-parad...