[sokoloff]

Suppose you ran 20 trains per day each way between those city pairs. What do you guess the average passenger load would be on each of those trains? Could you average even 100 pax per train? Would whatever the average total fare collected be enough to cover just the direct operating costs of that service?

I don’t see any non-stop airline flights between those cities for the smattering of dates I checked, suggesting that the airlines don’t find a lot of direct travel demand between that city pair.

[bombcar]

Los Angeles to San Diego is 120 miles, and the Pacific Surfliner serves it - and it only has 9 trains a day (round trip). (I notice that they've added a early morning service getting to LA at working time).

The key with these kinds of service is you have to run them consistently for 10-20 years before they start seeing the kind of ridership that can support the train density. People don't start building their lives around a transport option that they can't rely on.

[antisthenes]

> The key with these kinds of service is you have to run them consistently for 10-20 years before they start seeing the kind of ridership that can support the train density.

That's a good point. Whenever I've had the misfortune of using public transport in the US, it ended up being extremely unreliable.

Commuter bus at 6:30PM on a weekday? Just doesn't show up. Have to wait 40 minutes instead of 10.

NY to DC bus? Breaks down midway, have to wait an extra 2 hours for a relief bus to arrive.

DC to NY amtrak? Union station shut down for 3 hours due to weather-related power outage.

Now maybe I'm just super unlucky, but I've never heard of weather straight up shutting down an entire train station in other countries, especially in a nation's capital.

[yakireev]

> I've never heard of weather straight up shutting down an entire train station in other countries, especially in a nation's capital.

How about shutting down the entire network? Happens after any snowstorm in the NL.

That being said, it's amazing to have a reliable and fast passenger railroad network, which functions like an intercity metro and reliably and predictably brings you where you want to be in those 363 days out of 365 when there's no snowstorm, or a system failure, or a general strike.

[iso1210]

It's interesting how the American view of transport (other than private cars) has to make a profit. Other countries fund transport to various degrees as it increases economic output and provides social needs.

It's a similar distance as Penzance-Exeter in the UK which has 22 trains in each direction on a weekday

[bombcar]

Replace "profit" with "worthwhile" and it may become more clear. Literally getting to fare-box neutral is one way of determining if something worthwhile but it's not the only one. But people are bad at comparing the value of projects that are in the billons, and the value may take decades to appear. It can take 10+ years for people to decide to start using commuter rail that was just added, even if it would have worked fine the whole time.

[iso1210]

You're assuming the direct profit is the benefit, and ignore all externalities.

Do city streets make a profit? How about sewers?

The value of those resources is immeasurable - they enable trillions of dollars of economic benefit in the US alone.

[bombcar]

No, I'm saying it's hard for people to see the externalities and so they're inclined to ignore them. And in the US at least, sewers are "farebox positive".

Instead of trying to nail on rail to cities it should be part of comprehensive travel planning that includes roads, etc. But selling it alone gets things like the California High Speed rail which hasn't sped anywhere, and dampens further similar projects.

[closeparen]

Yes, the parcels along those streets pay property tax. Municipalities that supply infrastructure to vacant or low-value used are in trouble.

[oneoff786]

It’s not, and random commenters don’t mean anything.

Amtrak is a state owned enterprise. It is for profit, but it’s understood that it’s an economic multiplier.

[missedthecue]

The word "profit" has turned into a pejorative in the modern lexicon for some reason, but don't think it's unreasonable that such services should be self sufficient

[lkbm]

If all competitors were too, and we internalized all externalities, sure. But personal car travel is currently heavily subsidized (no, the fuel tax does not cover road costs), and has some serious negative externalities (both from air pollution, and traffic accidents being the leading cause of death for people under 30).

We demand that public transit be self-sufficient, while subsidizing private personal transportation. The market is a great "figure out the most efficient solution" mechanism, but not if you skew it in favor of one particular solution as we're doing now.

[missedthecue]

The conclusion here should be that gas taxes must be increased, not that we should continue to pump infinite money into the industrial sized furnace that is Amtrak.

[lkbm]

I wouldn't go with just the gas tax, but yes, ideally we would internalize all externalities via taxation.

I'm certainly in favor of that, but my point is not "we should subsidize everything equally" so much as "pay attention to one-sided demands for self-sufficiency".

(To the extent that it's viable, I think "equal" subsidies would lead to a better outcome and uneven subsidues, since it would allow the market to sort out the most efficient way to meet people's desires, but I'm not sure that's even remotely possible.)

[coredog64]

> But personal car travel is currently heavily subsidized (no, the fuel tax does not cover road costs)

If only there were some other way to collect funds for roads. One idea could be that governments require some kind of annual "license" that they charge you for. Alternatively, since private automobiles involve a large capital purchase, maybe we could levy some kind of fee or tax on the purchase to cover annual road maintenance.

[lkbm]

We could, but as they currently stand, but I've yet to see any analysis suggesting those come close to it covering the difference.

Most things do just look at parking and gas tax, but licensing fees are negligible compared to gas tax. The car sales tax might be a big source to make a difference, though. A few states don't have a sales tax on cars, but most do, and that may outstrip gas tax revenue if people buy frequently enough.

[sokoloff]

We pretty obviously don’t demand that. Amtrak has lost money every year since it was created in 1971.

https://www.businessinsider.com/history-of-amtrak-train-rail...

[lkbm]

We don't legally mandate it, but every discussion about Amtrak and public transit involves people insisting it should cover its own costs while ignoring the fact that the alternatives don't.

[jdasdf]

Profit is not a sin. On the contrary it is a 1 to 1 match with the good that it is providing its users.

This idea that services shouldn't turn a profit is a massive problem.

[jodrellblank]

Coca Cola provides more good to people than water, penicillin, insulin, cabbage, schools, libraries or parkland, because Coca Cola is more profitable than water, penicillin, insulin, cabbage, school, libraries and parkland?

[iso1210]

My understanding is insulin in the US is sold at ridiculous prices for profit, where as in the rest of the world it's just something that people who need it have it, like water, cabbage and schools.

[GraemeMeyer]

Nobody reasonable is saying that profit making is a sin, the argument is that only allowing/focusing on direct profit making services is shortsighted and misses opportunities to implement services that have indirect benefits

[gatlin]

Zero-sum profit certainly requires taking from someone else to succeed. This doesn't make it immoral as such but it creates malevolent incentives.

[Macha]

Modern airport security is fatal for overland airline flights of short durations. An hour by train? You could drive or train that in the time between when you're told you should be at the airport to the time the plane actually takes off

[ghaff]

Yes. New Orleans to Texas cities like Houston can make sense but not Mobile. Also public transit isn't great in either New Orleans or Mobile. It's not much more than a two hour drive. People will just get in a car.

[rascul]

> Yes. New Orleans to Texas cities like Houston can make sense but not Mobile.

It made sense until Katrina destroyed a bunch of the infrastructure and Amtrak had to stop running. The fight is to restore service, not create a new service.

[ghaff]

My comment was in reference to flying. Although I don't know what the pre-Katrina train traffic looked like on that route, it's not unreasonable to have a train connecting the Gulf coastal communities in that area. Though of course you're competing with a not terribly long drive.

[rascul]

> My comment was in reference to flying

Oops, I missed that.

> Although I don't know what the pre-Katrina traffic looked like on that route

I don't know either. I didn't live in the area at the time, and I wasn't easily able to find any numbers.

> Though of course you're competing with a not terribly long drive.

Very true. I've driven part of it a number of times. I-10 can get bad with traffic in some areas at some times of the day, but I suspect the total driving time from New Orleans to Mobile to be around 2.5 hours normally. I'm certainly not authoritative, though.

[inamberclad]

Went to New Zealand. My $30 ticket plane ticket from Christchurch to Hokitika was probably subsidized, but the sheer ease and accessibility of the flight was remarkable. I just got on the plane. No security, hardly a briefing. Walked out onto the ramp and got on. The plane was a Q-400, the size of which would make TSA security mandatory in the US. I know airlines like Surf Air in California were trying to specialize in small-plane (Pilatus PC-12, which is less than 12,500 lbs) but I haven't seen them grow much.

[sueders101]

If you’re looking for an example of a successful small plane airline in the United States: Cape Air.

[hedora]

By this logic, we should also stop allowing passenger cars on the freeways between those cities, and let shipping companies pay for road maintenance if they care.

[sokoloff]

When considering whether to run 4 or 20 round trips per day, it seems like the fundamental economics ought to come into play at some point. Otherwise, you eventually run out of other people’s money.

As a reference, there are only 7 Acela trains per day (or at least on Monday 4/25) from Boston to New York, two cities with substantially higher population and apparent demand (as evidenced by the 59 non-stop flights from BOS to any of the NYC-3 airports on Monday 4/25)

[Macha]

If you can only fill 7 trains per day between cities with metro areas of 5 million and 20 million, there is something else wrong with your network beyond just how much people like to take trains or not.

We manage to do that between a town of 20,000 and a city of 1 million for comparison. Or if you feel commuter routes are different enough to not count, a city of 60,000 and a city of 1 million with similar travel time as google maps quotes me for Boston to New York.

[iso1210]

It's crazy, there's currenlty only 1,000 seats an hour between Lonodn (10m) and Manchester (3m), having dropped from 3tph for covid. My experience recently is those trains are taking at least 600 people per hour now.

In my experience over the last few weeks there's barely been an empty seat.

I took the London-Paris train last week, absolutely rammed, there's 13x 900 seat trains a day at the moment, and that has all the nonsense of eurostar (airport style security, passport checks etc).

[sokoloff]

Note that none of those are “full”. The most crowded of those 7 Acelas is showing ~50% full with the others split between the <20% and ~40% categories.

And this is on the Northeast Acela, the crown jewel of the Amtrak network and between two cities with generally functioning public transit once you arrive. Most US city-pairs would be worse.

[ghaff]

There are a lot of factors.

- There are about double that number of trains if you count the Regional (which you should) so ~hourly trains.

- A lot of people still fly. Especially if you live in Boston proper, flying means you can easily make a morning meeting without flying down the night before (which people with families etc. may prefer not to do)

- Especially if you're south/west of Boston or in New York's Connecticut suburbs, it's probably cheaper/faster to just drive, something I really try to avoid when it comes to NYC but nonetheless for me taking Amtrak actually involves me driving for an hour in the wrong direction/

[bombcar]

I suspect LA-San Diego might be better (Pacific Surfliner is an Amtrak cooperative with the LOSSAN corridor). Those trains end up really full (though off-peak ones are comfortable; the rush hour trains are standing-room only).

The only way to get consistent usage is to commuter rail; commuters travel five times a day in both directions vs "travelers/vacationers" which may travel once a month or less.

[ghaff]

It's also not really correct. That's just the Acela. There are about the same number of regional trains (which are almost as fast; I generally don't even take the Acela unless someone else is paying). For people to the south of the Boston metro, there's also the option of taking Metro North from New Haven.

The Northeast Corridor service is very popular. In fact, I believe Amtrak has plans to expand it given that it's pretty much the only place in the country Amtrak doesn't lose money.

[bombcar]

It can be a lot easier to fill a train if scheduled right between a 20k city and a 1 million one - as the people in the 20k have real reasons to not live in the 1m and still commute there. But if you're looking at 5m vs 20m the "city" experience will be similar so ... why not move to the city you work in?

[Macha]

It's almost like I gave a second example that's not a commuter town for this exact reason.

[iso1210]

People travel for reasons other than commuting.

[bombcar]

Certainly, but if you look at the percentages, the easiest way to get to consistent usage is to use commuter. "Business" can be considered as a superset of commuter, short irregular business trips, but that's harder to build an entire line on (it can certainly be an upgrade to an existing line).

The second easiest is feeder - for example if the line between cities includes the airport, etc.

[saalweachter]

It also becomes tricky because adding more trains (or more train cars) at different times does different things.

A train every hour is convenient for some purposes, but for others you just want a bunch of trains early in the morning and later in the evening.

[bombcar]

For commuter rail inside a city it's nice to know "the trains run every X minutes so I can always get one" - between cities it can be more clumped around commuter times.

You also run into trainset issues where you want to run 5 trains in one direction and 5 back in the evening, which will require 5 trainsets, but if you run them back and forth you could do more trips with less trainsets, but some would be running off-peak (and in the worst case, nearly empty, but getting into position).

[hedora]

Those two stations are 215 miles apart by car, and the trip takes 4 hours by train. The train is only averaging 53mph.

I don't think its popularity is a reasonable predictor of demand for a modern train that would be 3-4 times faster.

I'd wager if all the other competitors were also running with tech typical of the 1920's, the train would be more popular.

[bombcar]

At some point even if your train is infinitely fast, you can't get above a certain average speed, dependent on the distance between stations.

Fun fact: technically the boring large Pacific Surfliner trains could be "high speed rail" since they could get to 120 MPH through Camp Pendleton if the line had PTC and was signaled correctly. As it is hits 80-90 through there, but it soon has to slow down for a stop.

To do high-speed rail right you basically need four tracks - a slower local service that stops at every stop, and a faster high-speed express service that only stops rarely.

[iso1210]

Manchester-London is 180 miles and takes 2h20, which isn't great, certainly not high speed, but that's being rectified. Pre covid it was 3 trains an hour for the majority of the day.

[rascul]

There's a lot of people and a number of cities along the Mississippi Gulf Coast where the train could also stop.

[sokoloff]

Those intermediate stops must be severely limited if the train is to cover the 140 miles in an hour total as hypothesized above.

[rascul]

It's part of the plan

https://www.southernrailcommission.org/new-orleans-to-mobile

> To initiate new daily passenger rail service between New Orleans and Mobile with two round trips each day, morning and evening, with stops in Bay St. Louis, Gulfport, Biloxi, and Pascagoula offering business-friendly service.

[ghaff]

So basically the plan would seem to be to service New Orleans, Mobile, and the Mississippi coast with a relatively leisurely train trip as an alternative to driving or bus (given there is apparently no real air).

[rascul]

So it would seem. I did check just now and find no direct flights from New Orleans (MSY) to Mobile (MOB). The shortest flights (about four hours) appear to stop in either Houston or Atlanta. I have no idea what bus service there might be.

[bombcar]

And one of the huge values of a train is that it CAN stop at those intermediate stops. A high speed few stop rail from the big city near me to the next one is meh as by the time I've driven into the city I might as well just keep driving to my destination.

But if a train, even a slower one, stopped in my smaller town or the next one over it becomes much more interesting.

[sokoloff]

If you had the high-speed train stop only at Gulfport (xor Biloxi) and maybe at Slidell, with regional service to pick up the small towns in between and take pax to the larger stations where they’d change to the faster train, you may be able to preserve a competitive Mobile to New Orleans time and still catch the intermediate city travelers. (This is a limited version of the airline hub-and-spoke to allow the trains to be faster door-to-door than driving, because if you can’t be better than driving in some obvious and personal way, many people will quite reasonably just drive the 140 miles on their own schedule and terms.)

[bombcar]

That's the hardest part - people will drive 6+ hours even when there are faster/better options if they need a car at the other end.

So transit between cites doesn't work as well until the endpoints are adequately transited themselves (or your destination is something like an airport where you can't bring your car anyway).

[ghaff]

At one point I was doing quite a bit of driving between New Orleans and a worksite in Pascagoula (just across the MS/AL state line from Mobile). This wouldn't have been very interesting at all given Pascagoula was very spread out and I absolutely needed a car once I was there. And it's a pretty straightforward ~2.5 hour drive.

Taking the train to NYC by contrast a car is, in general, actively a negative thing once you arrive.