11 ms·

Somewhat tangentially, I watched a recent Not Just Bikes YouTube video that discusses the merits of commuter rail [1]. The video is obviously trying to discoura

Somewhat tangentially, I watched a recent Not Just Bikes YouTube video that discusses the merits of commuter rail [1]. The video is obviously trying to discourage infrastructure that supports car-based suburbs, but another point came out that struck me: commuter rail is very expensive infrastructure with very low utilization, say 4 incoming trains in the morning and 4 outgoing trains in the evening. Any public transportation option needs to be compared against the default choice of "run more buses," which it turns out is pretty hard to beat in terms of cost-effectiveness. [1] https://youtu.be/vxWjtpzCIfA https://youtu.be/vxWjtpzCIfA

closeparen 5y ago
Even in clear traffic, buses are slow. I find the breakeven point with walking is about 2 miles, and with cycling about 7 miles. Going more than 7 miles on a bus sucks badly enough that I'll order my life not to do it. As such, the bus is never a first choice, but can be a lifesaver in some edge cases: * Weather is bad enough to pay the (very large) time penalty vs. walking or cycling. * You have moderate cargo hauling needs, no car, and delivery is impractical. * The train is broken, your trip is too long for active transport, and you don't have a car. * You are too elderly, infirm, or handicapped to walk or bike. * You can't afford a bike. * You can't afford to be anywhere else that's heated and enclosed. There are enough of these edge cases, and you're sufficiently likely to meet one of them at some point, that I think we should all be in favor of bus coverage. Backups and resilience are important! But we should take it as a positive sign when ridership is minimal. If we want people out of cars and can't afford trains, the answer is active modes.
Spooky23 5y ago
Busses are similar in some ways to the Amtrak issue. They are cheap, but very variable — although things like bus rapid transit help. I live in a small city, and the return bus trip takes 45-60m. The morning trip takes about 20. I pay about $80/mo to park at work, and my commute is 6-10m each way.
nickpeterson 5y ago
More high end, electric buses is probably the best option. They’re comfortable, take the room of 3 cars, fit way more people, and use existing infrastructure.
- himlion 5y ago
  Yes I think so too. I don't see rail competing with that once the charging challenges have been solved.
jeffbee 5y ago
That sounds like the worst rail service imaginable. Very American. The Swiss town of Chur, which has the same population as Menlo Park, California, and which is a good distance from Zurich, as Menlo Park is from San Francisco, has five S-Bahn trains every hour, all day long, plus 9 other regional, interregional, and intercity trains per hour. There's nothing fundamental about trains that says you have to have insulting levels of service. It's a choice.
- ravitation 5y ago
  It's a choice funded by the automobile and airline lobbies.
- Fiahil 5y ago
  We went to switzerland for vacation last summer, and we were quite impressed by the rail system there. Everything runs on time, there is no shortage of space, you can buy your tickets online and the trains are quiet and clean. I'm from France, so I'm not easily impressed by a working train station. I don't think you can hold any American state to that level of expectation when they are 70 years late in their railroad infrastructure. Not even California would be able to match a 1/10 of your rail service before 2050.
  - jeffbee 5y ago
    I don’t need the whole USA or even California. I’d settle for fixing the Bay Area, which has close to the same population as Switzerland, about the same economic activity, and only half the territory to cover. It should be a piece of cake.
    - rsj_hn 5y ago
      > It should be a piece of cake. Eh, you can't compare a local area GDP to a national GDP like this. Yes, tons of software licensing and ad revenue are funneled to the SF region giving us a local area GDP of 591 Billion dollars[1] which is comparable to Switzerland's $750 Billion, but if you try to convert that regional income into infrastructure spending, you'll find that 1) the ability of the bay area to tax that 591 Billion is much less than the ability of Switzerland to tax it's economic activity, as Switzerland captures ~32% of it's national income, while the bay area captures something like 10% of its regional income, so a ballpark 1/3 deflator. 2) once revenue is obtained from taxing, the efficiency with which the distributed bay area governments can make use of the tax revenue is much, much lower than the efficiency with which the Swiss government can spend the money. I'd estimate a 1/10 deflator. This is because we will waste tons of money doing various studies, funnelling side payments to local machine politics, battling environment impact reports in courts, funding various diversity and equity initiatives, and set up a dozen committees to oversee the spending plans. 3) once a dollar is actually spent on the production of infrastructure in the bay area, you'll find that the we will get much, much less for that dollar in infrastructure spending then the same dollar spent in Switzerland on infrastructure. I'd say another factor of 1/3 as per https://www.vox.com/22534714/rail-roads-infrastructure-costs-america https://www.vox.com/22534714/rail-roads-infrastructure-costs... So when you multiply all those deflators, you'll find that our 591 Billion can be usefully channelled to 1/90 * 591/750 ~= 1% of the useful infrastructure that Switzerland has. Which seems about right. Remember it cost San Francisco $3 Billion for the DTX[2] which is 2 kilometers of transit. We are not Switzerland, we are a dysfunctional, corrupt, patchwork of local governments that doesn't really understand much of anything when it comes to infrastructure or efficient government. [1] https://apps.bea.gov/iTable/iTable.cfm?reqid=70&step=1&acrdn=5 https://apps.bea.gov/iTable/iTable.cfm?reqid=70&step=1&acrdn... [2] https://sf.curbed.com/2018/6/18/17464616/bay-area-subway-train-rail-costs-price-bart-muni https://sf.curbed.com/2018/6/18/17464616/bay-area-subway-tra...
TulliusCicero 5y ago
It depends on the commuter rail. The S-Bahn in Germany is sort of 'commuter rail' but also gets regular use for other types of trips. And actual commuter-specific train lines tend to go on just general tracks that are used for other lines as well, so it's not a huge deal. It's just that US-style development -- more or less mandated by the government -- makes public transit in general infeasible. For example, in Munich where I lived, there are suburbs nearby surrounded completely by farmland, where the whole town is still a handful of minutes by bike to the train station. That means that even in these sleepy little towns, it's dense enough to where a car is unnecessary. In the US, that kind of development is a rarity, because the local zoning makes it literally illegal, and the roads have been designed to be hostile to anything that's not a car.
- kumarsw 5y ago
  This sort of density pattern is present in Colombia. The government taxes unbuilt land on a lot at a higher rate than built land, so this incentivizes having houses share a wall, which is the norm.
- curryst 5y ago
  That's not true at all, those towns are common across the Midwest and the South of the US. The only difference is that the rails are freight, so you have to use the highway to get anywhere. I grew up in several of those towns. I don't think I lived more than a 15 minute bike ride from the railroad tracks until I was in my 20s. If Wikipedia can be trusted, upon construction over 80% of the farms in the Midwest Corn Belt were within 5 miles of a railroad (which is an area roughly 5 times the size of Germany). I never lived close to a passenger train station until I moved to a big city. There are lots of small towns where there's a semi-dense urban core with a grocery store or two, a couple bars, a few restaurants, apartments and some shops. A lot of them even exist within a fairly small radius of major cities. Nobody builds rail anyways. My personal guess is that the major city doesn't want to pay for the rail, and the sleepy towns can't afford to pay for the rail. That's why most of what we see is major cities connecting, because they get enough economic activity out of those rails to justify the connection.
- jeffbee 5y ago
urthor 5y ago
I forget where I read it, but the cost per km of rail vs buses is pretty appalling. I suppose that rail has theoretically the advantage in being more reliable. But I do wonder if you could carry the same quantity (outside the subway cities) of passengers on a double lane bus road as a train.
- MereInterest 5y ago
  Is the cost of roads handled in that comparison, or are the roads assumed to already exist? (And are the roads being continually expanded to keep up with the induced demand from previous expansion?) If not, then that gives buses an unfair advantage in the comparison. Also, per kilometer probably isn't the best metric, unless you're building into a low-density area. It would be better to compare the cost per kilometer, divided by the throughput (people/hour).
- bombcar 5y ago
  The big advantage of rails is people don't try to drive on them. Build a dedicated bus lane across the country and everyone will want to use it for their cars, and either do so or bitch to the politicians that they can't.
- occz 5y ago
  You need to consider the price of road infrastructure in the total cost of buses. If that is not considered, then you could just as easily just calculate the cost of trains by excluding the rail and signalling-cost (and you would be just as incorrect).
  - sokoloff 5y ago
    You should certainly consider the buses’ share of the road cost, just like you should consider only the portion of the rail cost that is built and maintained for the commuter usage. If either of those shares is 100%, logically it should be burdened with 100% of the cost. Bus depots, bus maintenance yard, dedicated bus lanes, and bus stop turnouts should all be 100%.
    - 5y ago
TigeriusKirk 5y ago
Los Angeles did sort of a compromise in the San Fernando Valley. Instead of light rail, they put extra long busses on a dedicated busway. Part of the busway was an old rail line. It's worked out pretty well. It seems to attract people who wouldn't ride metro buses, for one thing. https://en.wikipedia.org/wiki/G_Line_(Los_Angeles_Metro) https://en.wikipedia.org/wiki/G_Line_(Los_Angeles_Metro)
- fulafel 5y ago
  The WP article explains that it was just to work around a bill forbidding surface rail on that route.
  - mst 5y ago
    That might have been why they decided to do it but the question of return-on-investment in terms of infrastructure costs is still interesting and doesn't really depend on the underlying motivation for the choice.
mushbino 5y ago
Reliable rail has been supported very heavily wherever it exists and it's the preferred option nearly everywhere. Busses are slow, uncomfortable, and less reliable, even if cheaper to operate. They can be a decent connector or last mile option.
- xabotage 5y ago
  Buses would be way better if roads were designed to prioritize them over single passenger cars. Instead (in the US), you're just stuck in traffic with extra stops, so you might as well drive.
  - Symbiote 5y ago
    Even on their own road, they're still bumpy and rough compared to a tram or especially a train.
Doctor_Fegg 5y ago
> commuter rail is very expensive infrastructure with very low utilization, say 4 incoming trains in the morning and 4 outgoing trains in the evening Post-COVID travelling patterns in the UK seem to suggest that the peaks are being ironed out. Whereas previously rolling stock procurement was dictated by the number of bums on seats for the morning and evening peak, now there's much more consistent demand across the day, and at weekends too - on some lines, weekend usage is now reportedly exceeding weekdays. No doubt it'll vary from place to place, but it looks like the days of buying enough rolling stock for the peaks, then having it sit empty for the rest of the day, might be over.

closeparen 5y ago

Even in clear traffic, buses are slow. I find the breakeven point with walking is about 2 miles, and with cycling about 7 miles. Going more than 7 miles on a bus sucks badly enough that I'll order my life not to do it. As such, the bus is never a first choice, but can be a lifesaver in some edge cases: * Weather is bad enough to pay the (very large) time penalty vs. walking or cycling. * You have moderate cargo hauling needs, no car, and delivery is impractical. * The train is broken, your trip is too long for active transport, and you don't have a car. * You are too elderly, infirm, or handicapped to walk or bike. * You can't afford a bike. * You can't afford to be anywhere else that's heated and enclosed. There are enough of these edge cases, and you're sufficiently likely to meet one of them at some point, that I think we should all be in favor of bus coverage. Backups and resilience are important! But we should take it as a positive sign when ridership is minimal. If we want people out of cars and can't afford trains, the answer is active modes.

Spooky23 5y ago

Busses are similar in some ways to the Amtrak issue. They are cheap, but very variable — although things like bus rapid transit help. I live in a small city, and the return bus trip takes 45-60m. The morning trip takes about 20. I pay about $80/mo to park at work, and my commute is 6-10m each way.

nickpeterson 5y ago

More high end, electric buses is probably the best option. They’re comfortable, take the room of 3 cars, fit way more people, and use existing infrastructure.

himlion 5y ago
Yes I think so too. I don't see rail competing with that once the charging challenges have been solved.

jeffbee 5y ago

That sounds like the worst rail service imaginable. Very American. The Swiss town of Chur, which has the same population as Menlo Park, California, and which is a good distance from Zurich, as Menlo Park is from San Francisco, has five S-Bahn trains every hour, all day long, plus 9 other regional, interregional, and intercity trains per hour. There's nothing fundamental about trains that says you have to have insulting levels of service. It's a choice.

ravitation 5y ago
It's a choice funded by the automobile and airline lobbies.
Fiahil 5y ago
We went to switzerland for vacation last summer, and we were quite impressed by the rail system there. Everything runs on time, there is no shortage of space, you can buy your tickets online and the trains are quiet and clean. I'm from France, so I'm not easily impressed by a working train station. I don't think you can hold any American state to that level of expectation when they are 70 years late in their railroad infrastructure. Not even California would be able to match a 1/10 of your rail service before 2050.
- jeffbee 5y ago
  I don’t need the whole USA or even California. I’d settle for fixing the Bay Area, which has close to the same population as Switzerland, about the same economic activity, and only half the territory to cover. It should be a piece of cake.
  - rsj_hn 5y ago
    > It should be a piece of cake. Eh, you can't compare a local area GDP to a national GDP like this. Yes, tons of software licensing and ad revenue are funneled to the SF region giving us a local area GDP of 591 Billion dollars[1] which is comparable to Switzerland's $750 Billion, but if you try to convert that regional income into infrastructure spending, you'll find that 1) the ability of the bay area to tax that 591 Billion is much less than the ability of Switzerland to tax it's economic activity, as Switzerland captures ~32% of it's national income, while the bay area captures something like 10% of its regional income, so a ballpark 1/3 deflator. 2) once revenue is obtained from taxing, the efficiency with which the distributed bay area governments can make use of the tax revenue is much, much lower than the efficiency with which the Swiss government can spend the money. I'd estimate a 1/10 deflator. This is because we will waste tons of money doing various studies, funnelling side payments to local machine politics, battling environment impact reports in courts, funding various diversity and equity initiatives, and set up a dozen committees to oversee the spending plans. 3) once a dollar is actually spent on the production of infrastructure in the bay area, you'll find that the we will get much, much less for that dollar in infrastructure spending then the same dollar spent in Switzerland on infrastructure. I'd say another factor of 1/3 as per https://www.vox.com/22534714/rail-roads-infrastructure-costs-america https://www.vox.com/22534714/rail-roads-infrastructure-costs... So when you multiply all those deflators, you'll find that our 591 Billion can be usefully channelled to 1/90 * 591/750 ~= 1% of the useful infrastructure that Switzerland has. Which seems about right. Remember it cost San Francisco $3 Billion for the DTX[2] which is 2 kilometers of transit. We are not Switzerland, we are a dysfunctional, corrupt, patchwork of local governments that doesn't really understand much of anything when it comes to infrastructure or efficient government. [1] https://apps.bea.gov/iTable/iTable.cfm?reqid=70&step=1&acrdn=5 https://apps.bea.gov/iTable/iTable.cfm?reqid=70&step=1&acrdn... [2] https://sf.curbed.com/2018/6/18/17464616/bay-area-subway-train-rail-costs-price-bart-muni https://sf.curbed.com/2018/6/18/17464616/bay-area-subway-tra...

TulliusCicero 5y ago

It depends on the commuter rail. The S-Bahn in Germany is sort of 'commuter rail' but also gets regular use for other types of trips. And actual commuter-specific train lines tend to go on just general tracks that are used for other lines as well, so it's not a huge deal. It's just that US-style development -- more or less mandated by the government -- makes public transit in general infeasible. For example, in Munich where I lived, there are suburbs nearby surrounded completely by farmland, where the whole town is still a handful of minutes by bike to the train station. That means that even in these sleepy little towns, it's dense enough to where a car is unnecessary. In the US, that kind of development is a rarity, because the local zoning makes it literally illegal, and the roads have been designed to be hostile to anything that's not a car.

kumarsw 5y ago
This sort of density pattern is present in Colombia. The government taxes unbuilt land on a lot at a higher rate than built land, so this incentivizes having houses share a wall, which is the norm.
curryst 5y ago
That's not true at all, those towns are common across the Midwest and the South of the US. The only difference is that the rails are freight, so you have to use the highway to get anywhere. I grew up in several of those towns. I don't think I lived more than a 15 minute bike ride from the railroad tracks until I was in my 20s. If Wikipedia can be trusted, upon construction over 80% of the farms in the Midwest Corn Belt were within 5 miles of a railroad (which is an area roughly 5 times the size of Germany). I never lived close to a passenger train station until I moved to a big city. There are lots of small towns where there's a semi-dense urban core with a grocery store or two, a couple bars, a few restaurants, apartments and some shops. A lot of them even exist within a fairly small radius of major cities. Nobody builds rail anyways. My personal guess is that the major city doesn't want to pay for the rail, and the sleepy towns can't afford to pay for the rail. That's why most of what we see is major cities connecting, because they get enough economic activity out of those rails to justify the connection.
jeffbee 5y ago

urthor 5y ago

I forget where I read it, but the cost per km of rail vs buses is pretty appalling. I suppose that rail has theoretically the advantage in being more reliable. But I do wonder if you could carry the same quantity (outside the subway cities) of passengers on a double lane bus road as a train.

MereInterest 5y ago
Is the cost of roads handled in that comparison, or are the roads assumed to already exist? (And are the roads being continually expanded to keep up with the induced demand from previous expansion?) If not, then that gives buses an unfair advantage in the comparison. Also, per kilometer probably isn't the best metric, unless you're building into a low-density area. It would be better to compare the cost per kilometer, divided by the throughput (people/hour).
bombcar 5y ago
The big advantage of rails is people don't try to drive on them. Build a dedicated bus lane across the country and everyone will want to use it for their cars, and either do so or bitch to the politicians that they can't.
occz 5y ago
You need to consider the price of road infrastructure in the total cost of buses. If that is not considered, then you could just as easily just calculate the cost of trains by excluding the rail and signalling-cost (and you would be just as incorrect).
- sokoloff 5y ago
  You should certainly consider the buses’ share of the road cost, just like you should consider only the portion of the rail cost that is built and maintained for the commuter usage. If either of those shares is 100%, logically it should be burdened with 100% of the cost. Bus depots, bus maintenance yard, dedicated bus lanes, and bus stop turnouts should all be 100%.
  - 5y ago

TigeriusKirk 5y ago

Los Angeles did sort of a compromise in the San Fernando Valley. Instead of light rail, they put extra long busses on a dedicated busway. Part of the busway was an old rail line. It's worked out pretty well. It seems to attract people who wouldn't ride metro buses, for one thing. https://en.wikipedia.org/wiki/G_Line_(Los_Angeles_Metro) https://en.wikipedia.org/wiki/G_Line_(Los_Angeles_Metro)

fulafel 5y ago
The WP article explains that it was just to work around a bill forbidding surface rail on that route.
- mst 5y ago
  That might have been why they decided to do it but the question of return-on-investment in terms of infrastructure costs is still interesting and doesn't really depend on the underlying motivation for the choice.

mushbino 5y ago

Reliable rail has been supported very heavily wherever it exists and it's the preferred option nearly everywhere. Busses are slow, uncomfortable, and less reliable, even if cheaper to operate. They can be a decent connector or last mile option.

xabotage 5y ago
Buses would be way better if roads were designed to prioritize them over single passenger cars. Instead (in the US), you're just stuck in traffic with extra stops, so you might as well drive.
- Symbiote 5y ago
  Even on their own road, they're still bumpy and rough compared to a tram or especially a train.

Doctor_Fegg 5y ago

> commuter rail is very expensive infrastructure with very low utilization, say 4 incoming trains in the morning and 4 outgoing trains in the evening Post-COVID travelling patterns in the UK seem to suggest that the peaks are being ironed out. Whereas previously rolling stock procurement was dictated by the number of bums on seats for the morning and evening peak, now there's much more consistent demand across the day, and at weekends too - on some lines, weekend usage is now reportedly exceeding weekdays. No doubt it'll vary from place to place, but it looks like the days of buying enough rolling stock for the peaks, then having it sit empty for the rest of the day, might be over.