One rare instance where I can contribute some inside information. I met the director of all of Amtrak's development efforts a few week ago over a cookout. And he laid down how the whole system works. Basically, Amtrak was created as part of a government bailout effort to rescue the train companies, who were making money from freight but losing money on commercial routes.
The train companies went to the US government and said "bail us out or we all go out of business". So the senate created Amtrak as a private company, and purchased all of the money losing assets (all passenger routes) from the train companies. The US government owns all voting shares, and the train companies owns all non-voting shares. So Amtrak is a quasi public-private utility companies, where the cost is bore by public in tax money.
Here is the kicker, the special interest group that lobbied for the US government to create Amtrak was ... the US airlines. They bet that passenger trains would be obsolete in a few decades, so the tax payers would not be holding the bag for too long.
50 years later, we are still holding the bag. According to my sources, they will be break even across all routes in a couple of years. By the way, all passenger trains in Europe , China and Japan lose money too.
Agree with the sentiment. I don’t know about Amtrak’s carbon foot print relative to airlines so I don’t know how the score tally out. The trains run on electricity but I’m not sure how it’s genrates. And fixed the link thank you
It’s purportedly [1] better at ~1500 BTU per passenger-mile versus ~2300 for air travel, but I couldn’t find any methodology there.
Most of the long-distance trains are diesel. If you read their full sustainability report [2] you can see that diesel fuel is their second largest cost, which means it’s also the primary source of energy.
Caltrain, for example, is still trying to raise funds for electrifying the system here in the Bay Area. It’ll soon be thirty years, since it was first proposed apparently [3].
Imagine the equation if the diesel fuel burned had a carbon tax on it, and the electricity (provided by overhead wire) was almost free due to the exceedingly low cost of renewables.
Of course it would be an enormous effort to electrify train lines across America; it should still be done.
I've wondered if battery electric multiple unit trains will/would make electrification cheaper since you then don't have to electrify the whole line. You can avoid having to put overhead lines where inconvenient, expensive, or potentially dangerous as well. Like tunnels and stations.
I love this idea, are you thinking a hybrid model? i.e. Overheads where possible, battery to bridge the gaps? that way in some places the trains could charge en-route.
That's exactly what I think is interesting. It's actually an old idea. I think with modern batters the design becomes a lot more viable. There are already some of these in service and more on order.
Well, we'll have to look at the cost benefit analysis. It should be done if it makes sense.
Would it be more economically feasible to invest the money into fourth generation nuclear technology and then research solid aluminum batteries that are hot swappable? That would cut down on costs for copper or steel for electrification.
"Enormous effort" and "it should be done" without an economic analysis is generally ill thought out.
It seems it is bolstered by profit by non-passenger _transport_ related revenue, but I think, is an example that train companies can have monetize their positive impact on society
The true story is also a little more complicated: in general the profitable ones are profitable now with operations, but the predecessor company JNR was privatized and split because it was struggling under the load of heavy Shinkansen construction debt, which got split into a sort of "bad corporation".
I wish I could but he didn’t give a very detailed answer. If I had to punt I’d say it’s because they didn’t think the government could offer a compelling experience once they took it over from private companies...
Postal employees would sort the mail while it traveled. Parallel processing!
Centralized automated sorting helped kill that model, along with the fact that an air and truck system really squeezes the viability of rail.
I read once that there's a sweet spot for passenger rail at about 250 miles, but it loses to air any farther and roads any shorter. Can't seem to find that source though, and I might be off by 50 miles. TSA might be one of the best things to happen to Amtrak, hard to say.
In both NYC and Boston, the track for boarding isn’t shown to the waiting area until arriving passengers had a chance to leave the train. Once your track number is shown you have only a couple minutes to board, and depart.
This means you can arrive at NYC’s Penn Station or Boston’s South Station 5 minutes before departure and walk right onto your train.
When I factor the terminal location and the boarding convenience at both ends of my trip, I can get door to door from midtown NYC to downtown Boston faster by Acela than by air.
Before TSA, the flights won. I could hit LGA curbside 15 minutes before gate close and be in my plane seat. Now, the variability is much too high, and TSA is too likely to be completely mismanaging surge situations, so typically 2 hours buffer is needed as a guarantee instead of 15 mins.
In that specific case, it's hard to justify taking Acela over a bus. Acela is at best an hour faster and something like 4x the cost (or more) It is certainly more comfortable, but unless we get true high speed rail that shortens the travel time between Boston and NYC to < 2 hours I can't ever justify personally paying for it. I've only ever taken Acela between Boston and NYC when an organization I was working for was willing to foot the bill.
In Europe, I took trains that hit speeds of 195 MPH for long stretches. That sort of speed would make rail a real viable third way between car/bus and air for long distance travel.
I’ve only taken the normal train between the two cities. For the regular train when you plan 2 weeks out the prices were comparable. The slow train is only 1/2 hour slower than the accela. I found it much more comfortable than the bus and easier than driving.
The main thing to me is the comfort level and that I can do other things than concentrate on driving.
I-95 congestion is so variable that you can’t depend on the bus if you have a meeting to make, though. I was once stuck on a Megabus that took 9 hours to get from College Park to the border of Maryland and Delaware.
I think in part because they also own all stations and really pushed very hard do the same as the airports: create a pleasant atmosphere and maximize the food and non-food offers.
> By the way, all passenger trains in Europe , China and Japan lose money too.
Do you have a source for that? I have a really hard time believing this is true when many of the areas you mention have multiple competing passenger train companies operating at the same time.
We need to fire the romantics and rail fans and have leadership that actually serves travelers. Approximately nobody actually relies on the long distance routes. Shut down the long distance routes, and spend that money improving service in routes where you actually have a hope of stealing significant market share from air and road travel.
We were having a similar problem with WMATA. People wanted to keep money losing late night service to help the “night life” (and invoking late shift workers as a fig leaf). New WMATA GM said it’d be cheaper to just provide Uber rides for the late shift workers and stood fast against extending hours. Maybe not the hero we want, but the hero we need.
So wait, we should have late shift workers depend on companies that subsidizes its rates relying on a "greater fool" theory of access to capital (VC, now Wall Street), exploits its labor force (drivers) and uses far less energy efficient vehicles (cars)?
Oh and btw most of these shift workers can't afford to ride Uber/Lyft much less if the fare is surging. A $7 ride is what, half an hour of minimum wage work? That's pretty steep. Maybe we have WMATA reimburse the worker, but who hasn't had a large gap between filing a receipt and getting reimbursed? Now add being potentially financially precarious, doesn't seem ideal.
If only we had a service that could transport large numbers of people using electricity to pre-defined locations so that they could have a far cheaper final leg of their trip. Oh wait...
I'd guess that late at night an Uber ride from downtown DC to Shady Grove will run you close to that. It definitely will cost more after say, a late night game at Nationals Park. Uber isn't public transit, it serves a different purpose and shouldn't be seen in the same way as public infrastructure.
When I lived in DC, I loved taking the metro but I couldn't rely on it when it got late. If I were king for a day, I'd force the creation of a new ring-type line and make the system run late with far more trains. This would cost money, it would be subsidized. It will make life better for all DC area residents and reduce the terrible traffic in the area. It would also may help mitigate downtown housing prices and slightly increase housing prices out in the suburbs. An enormous number of lives would be improved for a relatively small $35-$40 subsidy per ride.
Public infrastructure should still be subject to cost-benefit analysis, just like everything else. That doesn’t mean it can’t be subsidized, but it must still achieve the necessary results at the lowest cost.
A $40 Uber is what it costs to get me all the way to Annapolis late at night. That’s not a “small” subsidy, its an insane waste of money that could be put to better use. If the concern is really low-income late night workers, it’d be cheaper to just lease those people an EV.
A ring-type line in DC would also be a terrible idea and would do nothing to reduce traffic. Any ring would run through mostly low-density areas where neither the starting point nor end point would be walkable to the train station. (The possibility has been studied.)
There's zero reason to say that a ring line must through low density areas. You can place the ring (or partial ring) underground at a variety of locations that would be higher density, you could have it be further out from downtown as a mix of above and below ground, it's just about what we are willing to pay. Obviously, an all-above-ground follows-the-beltway route would be pretty low density.
As you know, the DC Metro has several U shaped lines. It takes a ridiculously long time to get from say, College Park to Dupont Circle. Or from the stadium to Silver Spring. A few linker lines would make the downtown system far quicker internally and also accelerate those kinds of trips, even if it doesn't create a full ring. The goal should be much like the NY system: I can get anywhere in DC relatively easily using only the Metro.
I think the way you've phrased the cost-benefit is exceedingly narrow. Operating a train is a fixed cost, if it's highly utilized the cost goes down. As such, making the service more useful can shift the cost-benefit. Also, there are benefits (for example fewer cars on the road) which don't seem to be taken into account.
Finally a key assumption here is that Uber's price is real. I personally don't think Uber's current prices are sustainable. They are either 1.) subsidized by investor capital or 2.) dependent on paying drivers essentially minimum wage (maybe even less) or 3.) both. If you didn't have Uber around would your analysis shift? (genuine question) In the past I doubt we'd have said, "Taxis are a replacement for public transit." We should be careful of making a similar argument today.
Re. low wage workers, it's a concern. You make some good points, but I'm not sure I buy that the answer is Uber or leasing EVs. How many have access to a charger for that EV? How many can get a driver's license? The poor often live precariously, they may not be able to park that car a month or two from now, or who knows what other problems might come up.
In short, I think focusing on top line cost-per-ride is a very incomplete way to look at infrastructure.
p.s. $40 from downtown DC to Annapolis, wow Uber is cheaper out there than here in the bay area.
If we’re asking people to give up their cars, we need to handle all transportation needs, not just the ones that are convenient to serve. Otherwise we still need to pay all the fixed costs of cars and car infrastructure for the cases not served by transit.
Asking people to give up their cars is pointless social engineering. (Mandating EVs is a different proposition.) When you get where these trains go, you can’t do anything without a car.
The thing about long-distance rail is that it offers a lot of value for secondary rural stops.
There's no way that flights to, say, Winslow, Arizona will ever be economical. You might have someone offer them, but at the cost of enormous subsidy. You can't expect flights running LAX-ORD to stop there, even if there was the infrastructure to handle it.
But the railway runs right through the town. You're already running trains. All you need to do is add a 10 minute gap to the schedule for boarding and you've connected that town back to the world.
Yes, buses can do it too, but Greyhound doesn't serve a government mandate and will drop money-losing routes (I think there's been a huge fuss over this in Canada recently)
Given the political power of rural legislators, Amtrak is an easy sell.
The real issue is that Amtrak is really a collection of networks that should not be dragged down by each other.
If you want the government to run long-distance rail service, then there should be a dedicated company that we subsidize specifically for that purpose without dragging down the rest of the American rail network.
There were a techy startup or two making overnight sleeper-bus service between LA and SF over the last few years. Apparently the ride quality was rough, but they had a solid business model. How Amtrak hasn’t seen this opportunity of a far more comfortable, spacious travelling hotel (ie 9pm to 8am) between the two cities (and similar routes) is mind-boggling to me.
I know this to be true; it just surprises me Amtrak, as large as it is, doesn’t have more leverage / bargaining power here. I understand tying up a line can be an issue and the rail network is a relatively complex thing, but if it enables a service to make money and BNSF makes a cut, I’d think it would be viable. I don’t think freight lines are exactly at capacity (generalizing here), nor are they gushing with cash these days either. IIRC Union Pacific just laid off 7000 jobs.
I have used Amtrak a great over the past 13 years that I have lived in NYC. Back around 2006 a 10-hour trip from NYC to Pittsburgh cost me $60. Now it is much more expensive.
I still love Amtrak, though. Their iPhone app is pretty good, their customer service has always been excellent, and I really enjoy traveling by train -- it is so much more comfortable and efficient than any other mode of transport, and it is a great environment for getting work done.
I wrote some software that was used by Amtrak, the people implementing the software on the trains said that the food being sold on the trains was usually not what was being loaded on and off the trains, but was instead supplied by the employees, so that they could make extra cash. I'm not sure if that's still the case, but with a captive audience, I'm surprised Amtrak does so poorly.
Another issue is that cargo trains get precedence, so travel times are often excessively long.
By what metric should we rate the value of infrastructure like Amtrak's? Should it really be profitability? Maybe I'm not reading the right articles, but I never see highway or interstate infrastructure judged by their ability to turn a profit (with rare exception they never do). Rather, they are judged by factors like how many more people can travel, how they decongest traffic, or even the economic activity that they enable.
If we take into account the number of cars that are removed from the road, and the economic activity enabled by fast eastern seaboard commutes to financial and political centers, then where does the balance lie for this service?
To not sound wishywashy, I bet it looks more favorable.
44 comments
[ 0.20 ms ] story [ 119 ms ] threadThe train companies went to the US government and said "bail us out or we all go out of business". So the senate created Amtrak as a private company, and purchased all of the money losing assets (all passenger routes) from the train companies. The US government owns all voting shares, and the train companies owns all non-voting shares. So Amtrak is a quasi public-private utility companies, where the cost is bore by public in tax money.
Here is the kicker, the special interest group that lobbied for the US government to create Amtrak was ... the US airlines. They bet that passenger trains would be obsolete in a few decades, so the tax payers would not be holding the bag for too long.
50 years later, we are still holding the bag. According to my sources, they will be break even across all routes in a couple of years. By the way, all passenger trains in Europe , China and Japan lose money too.
PS your website is down
Most of the long-distance trains are diesel. If you read their full sustainability report [2] you can see that diesel fuel is their second largest cost, which means it’s also the primary source of energy.
Caltrain, for example, is still trying to raise funds for electrifying the system here in the Bay Area. It’ll soon be thirty years, since it was first proposed apparently [3].
[1] https://m.amtrak.com/h5/r/www.amtrak.com/travel-green
[2] https://www.amtrak.com/content/dam/projects/dotcom/english/p...
[3] https://en.m.wikipedia.org/wiki/Electrification_of_Caltrain
Of course it would be an enormous effort to electrify train lines across America; it should still be done.
Found this:
https://www.youtube.com/watch?v=1NpqPP5SMfY
https://www.youtube.com/watch?v=xxZQeS__6sE
Would it be more economically feasible to invest the money into fourth generation nuclear technology and then research solid aluminum batteries that are hot swappable? That would cut down on costs for copper or steel for electrification.
"Enormous effort" and "it should be done" without an economic analysis is generally ill thought out.
It seems it is bolstered by profit by non-passenger _transport_ related revenue, but I think, is an example that train companies can have monetize their positive impact on society
The true story is also a little more complicated: in general the profitable ones are profitable now with operations, but the predecessor company JNR was privatized and split because it was struggling under the load of heavy Shinkansen construction debt, which got split into a sort of "bad corporation".
Generally available social services are not for profit. Who knew?
https://en.wikipedia.org/wiki/Railway_post_office
Postal employees would sort the mail while it traveled. Parallel processing!
Centralized automated sorting helped kill that model, along with the fact that an air and truck system really squeezes the viability of rail.
I read once that there's a sweet spot for passenger rail at about 250 miles, but it loses to air any farther and roads any shorter. Can't seem to find that source though, and I might be off by 50 miles. TSA might be one of the best things to happen to Amtrak, hard to say.
The first passenger airplane that made money without carrying mail was the DC-3 with 21 passenger seats.
In both NYC and Boston, the track for boarding isn’t shown to the waiting area until arriving passengers had a chance to leave the train. Once your track number is shown you have only a couple minutes to board, and depart.
This means you can arrive at NYC’s Penn Station or Boston’s South Station 5 minutes before departure and walk right onto your train.
When I factor the terminal location and the boarding convenience at both ends of my trip, I can get door to door from midtown NYC to downtown Boston faster by Acela than by air.
Before TSA, the flights won. I could hit LGA curbside 15 minutes before gate close and be in my plane seat. Now, the variability is much too high, and TSA is too likely to be completely mismanaging surge situations, so typically 2 hours buffer is needed as a guarantee instead of 15 mins.
In Europe, I took trains that hit speeds of 195 MPH for long stretches. That sort of speed would make rail a real viable third way between car/bus and air for long distance travel.
The main thing to me is the comfort level and that I can do other things than concentrate on driving.
Not all. The Dutch Railway company is profitable: https://www.statista.com/statistics/748940/net-profit-of-ns-...
I think in part because they also own all stations and really pushed very hard do the same as the airports: create a pleasant atmosphere and maximize the food and non-food offers.
Not true. Russian passenger rail is profitable and growing.
Do you have a source for that? I have a really hard time believing this is true when many of the areas you mention have multiple competing passenger train companies operating at the same time.
We were having a similar problem with WMATA. People wanted to keep money losing late night service to help the “night life” (and invoking late shift workers as a fig leaf). New WMATA GM said it’d be cheaper to just provide Uber rides for the late shift workers and stood fast against extending hours. Maybe not the hero we want, but the hero we need.
Oh and btw most of these shift workers can't afford to ride Uber/Lyft much less if the fare is surging. A $7 ride is what, half an hour of minimum wage work? That's pretty steep. Maybe we have WMATA reimburse the worker, but who hasn't had a large gap between filing a receipt and getting reimbursed? Now add being potentially financially precarious, doesn't seem ideal.
If only we had a service that could transport large numbers of people using electricity to pre-defined locations so that they could have a far cheaper final leg of their trip. Oh wait...
When I lived in DC, I loved taking the metro but I couldn't rely on it when it got late. If I were king for a day, I'd force the creation of a new ring-type line and make the system run late with far more trains. This would cost money, it would be subsidized. It will make life better for all DC area residents and reduce the terrible traffic in the area. It would also may help mitigate downtown housing prices and slightly increase housing prices out in the suburbs. An enormous number of lives would be improved for a relatively small $35-$40 subsidy per ride.
A $40 Uber is what it costs to get me all the way to Annapolis late at night. That’s not a “small” subsidy, its an insane waste of money that could be put to better use. If the concern is really low-income late night workers, it’d be cheaper to just lease those people an EV.
A ring-type line in DC would also be a terrible idea and would do nothing to reduce traffic. Any ring would run through mostly low-density areas where neither the starting point nor end point would be walkable to the train station. (The possibility has been studied.)
As you know, the DC Metro has several U shaped lines. It takes a ridiculously long time to get from say, College Park to Dupont Circle. Or from the stadium to Silver Spring. A few linker lines would make the downtown system far quicker internally and also accelerate those kinds of trips, even if it doesn't create a full ring. The goal should be much like the NY system: I can get anywhere in DC relatively easily using only the Metro.
I think the way you've phrased the cost-benefit is exceedingly narrow. Operating a train is a fixed cost, if it's highly utilized the cost goes down. As such, making the service more useful can shift the cost-benefit. Also, there are benefits (for example fewer cars on the road) which don't seem to be taken into account.
Finally a key assumption here is that Uber's price is real. I personally don't think Uber's current prices are sustainable. They are either 1.) subsidized by investor capital or 2.) dependent on paying drivers essentially minimum wage (maybe even less) or 3.) both. If you didn't have Uber around would your analysis shift? (genuine question) In the past I doubt we'd have said, "Taxis are a replacement for public transit." We should be careful of making a similar argument today.
Re. low wage workers, it's a concern. You make some good points, but I'm not sure I buy that the answer is Uber or leasing EVs. How many have access to a charger for that EV? How many can get a driver's license? The poor often live precariously, they may not be able to park that car a month or two from now, or who knows what other problems might come up.
In short, I think focusing on top line cost-per-ride is a very incomplete way to look at infrastructure.
p.s. $40 from downtown DC to Annapolis, wow Uber is cheaper out there than here in the bay area.
There's no way that flights to, say, Winslow, Arizona will ever be economical. You might have someone offer them, but at the cost of enormous subsidy. You can't expect flights running LAX-ORD to stop there, even if there was the infrastructure to handle it.
But the railway runs right through the town. You're already running trains. All you need to do is add a 10 minute gap to the schedule for boarding and you've connected that town back to the world.
Yes, buses can do it too, but Greyhound doesn't serve a government mandate and will drop money-losing routes (I think there's been a huge fuss over this in Canada recently)
Given the political power of rural legislators, Amtrak is an easy sell.
If you want the government to run long-distance rail service, then there should be a dedicated company that we subsidize specifically for that purpose without dragging down the rest of the American rail network.
I still love Amtrak, though. Their iPhone app is pretty good, their customer service has always been excellent, and I really enjoy traveling by train -- it is so much more comfortable and efficient than any other mode of transport, and it is a great environment for getting work done.
https://synchrony.nyc/
Another issue is that cargo trains get precedence, so travel times are often excessively long.
If we take into account the number of cars that are removed from the road, and the economic activity enabled by fast eastern seaboard commutes to financial and political centers, then where does the balance lie for this service?
To not sound wishywashy, I bet it looks more favorable.