Idea: Western LD trains split up into multiple shorter trains?

[Texan Eagle]

I am totally in agreement with those who pointed out that this arrangement will be terrible for end to end passengers since they will have to change trains twice, but as I asked in my original post, how many people do travel that long anyway, except us railfans who do these trips as a vacation in itself? Look at the statistics Paulus provided-

"10.5% of passengers, nearly half of whom are coach, ride the Empire Builder in excess of 2000 miles. 28.5% ride less than 300 miles/"

So, the inconvenience of changing trains twice is probably going to exist for very small fraction of total riders.

I did not present my line of thinking behind this idea, so here goes- I love the LD trains as they are now as much as all of you, but looking at the state these trains are in, with all the cuts coming in, there will come a time soon when Amtrak will have to decide if thy want to project themselves as a luxury land-cruise experience, or a not-widespread but functional transportation service. If former, keep the LD trains as they are now, complete with the wine and cheese, flowers, champagne and what not, to attract the high paying long distance passengers who come for the experience more than as a practical mode of transport. If latter, like it or not, we have to think out of the box ways to make Amtrak survive, and one of them is to cut them into more manageable bits and pieces.

An 8 to 16 hour train route is easier to manage than a 45 hour mammoth in terms of scheduling, with padding and recovery times etc, logistics of food and catering supplies for such a long journey, planning for contingencies and so on. Also, since Amtrak heavily relies on third parties (host railroads), having trains running smaller segments is also better for passengers, since for example a derailment in North Dakota will delay the trip for Montana to Washington bound passengers if there is only one LD train running the entire length, but with three segments,the only ones affected are the ones traveling that segment only (and the few connecting ones, which are, as per statistics above just about 10%.. can't please everyone sorry!).

The Hi-Line may not be the ideal example for such an experiment, but a route like CZ that has major cities along the way can be run as multiple corridor services with lesser amenities than current LD, but still much more convenient than taking cross-country buses. With the state of the economy and government attitude towards rail in the US, sorry to say but we might have to live with this reality coming sooner or later.

California Zephyr could possibly run in segmented avatar as follows-

Zephyr One:

CHI 3.00pm

DEN 8.00am

DEN 1.00pm

CHI 6.00am

Zephyr Two:

DEN 3.00pm

SLC 6.00am

SLC 3.30pm

DEN 6.30am

Zephyr Three:

SLC 2.00pm

EMY 7.00am

EMY 5.00pm

SLC 10.00am

Notice something? These are all now single overnight trains, the type of services very popular in other parts of the world, and can be popular between these city pairs if advertised correctly. If you are really motivated enough to travel transcontinental by train you can still do it in 3 days, giving you nice little breaks in Denver and Salt Lake City to enjoy the cities along the way.

Another thing, since these are all just single overnighters, you can combine Dining Car and SSL into one. Yes, I know we all love to have a full Dining Car, but if the choice comes to having a full Diner vs having no LD trains at all, what would you choose? I would certainly vote for SSL with an improved Cafe that will actually serve you at those tables on the Upper level (and lower level too).

Does this make more sense now? Or does someone still need Advil? :wacko:

 

[neroden]

If we could justify the maintenance bases at SLC and Denver, it might be worth considering. We can't justify them without additional trains per day, unfortunately.

I would love to see a network of corridor services emanating out of Denver (to ski areas, Pueblo, Fort Collins, etc.), and the moment we see this... then it starts being worth considering.

(* Of course, at that point it would become clear that you want to run the Denver-Salt Lake route via Wyoming, separate from the Denver-Ski Areas route.)

 

[Paulus]

Also for consideration:

With full occupancy, at average fare yields, per mile a coach car earns more gross revenue than a sleeper car on the following trains: Silver Star, Silver Meteor, City of New Orleans, and Crescent.

Lies, damned lies, and statistics. Occupancy rates are different, of course (and, specifically, they're usually lower in coach). Also, you can't multiply the "average fare yield" by the occupancy rate to get the gross revenue.

Abusing statistics is fun! Anyhow, while it's not a "This is exactly how Amtrak trains fill out," it is a useful tool for consideration as a first element.

Worse for the spreader of these bad statistics, demand is different for different trips: in general, the sleeper trips are trips which would not be taken in coach; replacing the sleeper with a coach would just lose the income. The coach occupancy rate would then be lower and so would the average fare yield in coach, but hey, you would have replaced the "less profitable sleepers" with "more profitable coaches", right? Wrong.

That's really not necessarily true. Note that, as a percentage of all trips, the best that sleepers do is somewhat over 50% of all trips for particular segments. So for every sleeper that you fill from Chicago to Seattle, you have to keep in mind that you also have about half to two thirds of a coach car filled with passengers doing the exact same thing. It's quite possible that at least some of the current sleeper passengers would willingly take coach if it was all that was offered (and note that I do not suggest that only coach be offered; I believe that every train should be offering a business class service and that this would be far more valuable than sleeper accommodations). And of course there is the question of whether there truly is net income from sleeper passengers.

Paulus, you obviously didn't read my essays in the other thread: I can't really blame you, since they were so long.

And I've got a lot of reading for class. This was not your week for getting me to read long essays I'm afraid

The key point: the problem is not amenities, the problem is route segments.

The Crescent suffers from very low demand south of Atlanta, which is undoubtedly distorting the numbers. It's inadvisable to be running so many sleeping cars south of Atlanta, because they're running empty or selling cheap. Then again, it seems to be inadvisable to be running so many coaches south of Atlanta, too, since they're also running empty or selling cheap.

This is completely screwing up the accounting there. With a proper Atlanta station, one could run the full Crescent to Atlanta and a continuing train (with very few cars) onward to New Orleans, and that would actually make sense. And the numbers for the train *north of Atlanta* would show that the sleepers get higher revenue per car than coaches. Or with much faster tracks south of Atlanta, people might actually ride the train south of Atlanta.

30% of the train's ridership is to or from stations south of Atlanta which is in line with the percentage of the route that that segment represents. As for much faster tracks, it averages 43 mph there; New York to Atlanta is an average of 47mph. Speeding up to the average is a hair under an hour between Atlanta and NOLA, it's probably not a significant factor in ridership demand.

The Silver Star suffers from a similar problem to the Crescent, in that nobody's going to buy a roomette from Miami to Tampa. It really needs to terminate in Tampa and connect to a day train to Miami -- the problem there is the lack of a maintenance base in Tampa. There's a similar, smaller problem selling roomettes from Orlando to Miami on the Silver Meteor.

Further, the Silver Star and Silver Meteor have both been badly hurt by trackwork in Florida and by weather disruptions which prevented the trains from running their entire route -- this *particularly* hurts ridership (and the ability to charge high prices) on the longer trips, which are the ones taken by sleeper.

It does not surprise me that those three trains *currently* have sleepers underperforming relative to coaches, but I believe this is entirely a matter of special conditions due to bad trip segments.

I will say that the CONO does seem to be underperforming on sleeper revenue, without obvious problem segments or disruption, and it would be nice to figure out why. I think I know, however. The big markets along the line are:

- Chicago to Carbondale and points north: mostly take Illini/Saluki corridor trains

- New Orleans to Memphis and points south: it's a day train here, nobody will buy sleepers

- Chicago to Memphis, Jackson, and New Orleans: this is it for the sleeper demand, basically three city-pairs. And none of the connections make any sense at New Orleans, so anyone doing this is actually travelling to or from one of those three cities.

What gets me is that you acknowledge the inability to sell sleeper accommodations for daylight corridors while ignoring that the vast majority of trips are precisely that: Daylight trips. And yes, it's all about the corridors and the principle problem with long distance trains: Night runs. Trains live or die on coach, not sleepers.

The CONO as a whole is a particularly short train and has relatively poor ridership, showing low demand overall. The speeds on the route are slower than driving but not outrageously slower. I must attribute the problems primarily to low population along the route. Secondary problems may include poor connecting transit, low awareness of trains in the region, and a lack of social/cultural links between Mississippi/Louisiana and Chicago.

Nope. Adjust for the length of the route and it's one of the better performing trains in the long distance network. That includes for sleepers interestingly enough.

The LSL has a similar overlap with a "corridor" at its east end, but the sleeper demand includes Chicago to Buffalo/Rochester/Syracuse/Albany/Boston/NY, which is a lot more demand. Especially given that there are closer cultural links between Chicago and the Northeast than between Chicago and Louisiana/Mississippi, and there is actual same-day connecting service in multiple directions at Chicago, New York, Boston, and even Schenectady and Buffalo.

Hooray for network effects. Understand them. Use them.

The Lake Shore Limited is absolutely dominated by coach passengers and their demand; it really should have its sleepers and diners stripped and be retimed for a daylight run through Ohio in order to exploit that to its full. It's actually quite weird how its coach fares are on the lower end of things compared to the other LDTs when its coach demand far outstrips the others.

And of course the trouble with the idea of just running them as day trains is that they run too long.

So cut them down to appropriate lengths.

Notice something? These are all now single overnight trains, the type of services very popular in other parts of the world, and can be popular between these city pairs if advertised correctly.

Err, no. What's very popular are same day trips of only a few hours with multiple frequencies per day. Night trains are a dying breed in the rest of the world.

 

[Texan Eagle]

Another Idea! (And no, I am not drunk in case anyone is wondering :giggle: )

While we are at the idea of single overnight trains instead of one that takes 2/3 days, would it work to create an intermediate class of sleeping accommodation for single travelers (students, cost-cutting business travelers on a budget) that costs less than present day roomettes? This topic has been discussed to death but here is one idea that can be implemented without requiring ordering new cars or refurbishing existing ones- what if, in one of the sleeper cars, all the roomettes are left open (seal the roomette doors in open position) and sell the upper and lower berths separately to single travelers at prices lower than roomette. There are valid concerns about privacy if sharing a closeable roomette with a stranger but if it can't be closed, now its a bit more open and relatively less "creepy". These days a lot of youngsters travel solo and use services like AirBnb, couchsurfing and hostels.. and as you might have observed, students and youngsters use LD Amtrak trains in a substantial number. This demographic won't mind sharing an open roomettes with another passenger if it means overnight sleeping facility at lower cost. Pricing can be straight up cost of roomette for 2 passengers divided by 2.

For example, CHI-DEN for 19 March,

Coach - $ 116

Roomette for 2 passengers - $ 568

Roomette if traveling solo - $ 452

New class: shared roomette berth for solo traveler - $ 284 (568/2)

$116 to $452 is quite a big jump for one overnight on train, even after considering one night hotel cost + meals.. but $116 to $284, now that is a bit more lucrative to consider.

 

[Anderson]

I am totally in agreement with those who pointed out that this arrangement will be terrible for end to end passengers since they will have to change trains twice, but as I asked in my original post, how many people do travel that long anyway, except us railfans who do these trips as a vacation in itself? Look at the statistics Paulus provided-

"10.5% of passengers, nearly half of whom are coach, ride the Empire Builder in excess of 2000 miles. 28.5% ride less than 300 miles/"

So, the inconvenience of changing trains twice is probably going to exist for very small fraction of total riders.

I did not present my line of thinking behind this idea, so here goes- I love the LD trains as they are now as much as all of you, but looking at the state these trains are in, with all the cuts coming in, there will come a time soon when Amtrak will have to decide if thy want to project themselves as a luxury land-cruise experience, or a not-widespread but functional transportation service. If former, keep the LD trains as they are now, complete with the wine and cheese, flowers, champagne and what not, to attract the high paying long distance passengers who come for the experience more than as a practical mode of transport. If latter, like it or not, we have to think out of the box ways to make Amtrak survive, and one of them is to cut them into more manageable bits and pieces.

An 8 to 16 hour train route is easier to manage than a 45 hour mammoth in terms of scheduling, with padding and recovery times etc, logistics of food and catering supplies for such a long journey, planning for contingencies and so on. Also, since Amtrak heavily relies on third parties (host railroads), having trains running smaller segments is also better for passengers, since for example a derailment in North Dakota will delay the trip for Montana to Washington bound passengers if there is only one LD train running the entire length, but with three segments,the only ones affected are the ones traveling that segment only (and the few connecting ones, which are, as per statistics above just about 10%.. can't please everyone sorry!).

The Hi-Line may not be the ideal example for such an experiment, but a route like CZ that has major cities along the way can be run as multiple corridor services with lesser amenities than current LD, but still much more convenient than taking cross-country buses. With the state of the economy and government attitude towards rail in the US, sorry to say but we might have to live with this reality coming sooner or later.

California Zephyr could possibly run in segmented avatar as follows-

Zephyr One:

CHI 3.00pm

DEN 8.00am

DEN 1.00pm

CHI 6.00am

Zephyr Two:

DEN 3.00pm

SLC 6.00am

SLC 3.30pm

DEN 6.30am

Zephyr Three:

SLC 2.00pm

EMY 7.00am

EMY 5.00pm

SLC 10.00am

Notice something? These are all now single overnight trains, the type of services very popular in other parts of the world, and can be popular between these city pairs if advertised correctly. If you are really motivated enough to travel transcontinental by train you can still do it in 3 days, giving you nice little breaks in Denver and Salt Lake City to enjoy the cities along the way.

Another thing, since these are all just single overnighters, you can combine Dining Car and SSL into one. Yes, I know we all love to have a full Dining Car, but if the choice comes to having a full Diner vs having no LD trains at all, what would you choose? I would certainly vote for SSL with an improved Cafe that will actually serve you at those tables on the Upper level (and lower level too).

Does this make more sense now? Or does someone still need Advil? :wacko:

Looking at your proposal...to be honest, if we had to split the Zephyr, I'd do it at DEN only (owing to the handoff between UP and BNSF, which doesn't always go well)...but then again, any proposal here that really makes sense is probably going to put extra trains into the mix around Denver.

Of course, a bit of heresy comes to mind...I can't help but wonder if it wouldn't make sense to make the Zephyr UP all the way. If I'm not mistaken, UP's mainline condition is at least as good as the ex-CB&Q under BNSF. Of course, I think there's a track gap between Union, CO and Denver, CO.

 

[Trogdor]

"10.5% of passengers, nearly half of whom are coach, ride the Empire Builder in excess of 2000 miles. 28.5% ride less than 300 miles/"

So, the inconvenience of changing trains twice is probably going to exist for very small fraction of total riders.

The problem with your logic is that it doesn't recognize that many of those shorter-distance passengers are still crossing what is, essentially, an arbitrary boundary where they now have to change trains.

Seattle to Whitefish, for example, is less than 600 miles, but under your proposal, they either have to make a middle-of-the-night connection (westbound), or they're spending the night in Spokane. You've effectively killed that (fairly short-distance) market.

Where are the folks going to/from Williston, Havre, Shelby, Whitefish, etc. going? While I don't have specific stats, I'd wager that a good percentage of them are going east of Minot or west of Spokane.

Point is, it doesn't matter how many people are going end-to-end. Each additional stop represents additional potential city pairs that are now no longer served (except through less convenient connections). If Spokane or Minot were rail hubs, with multiple connecting trains going all over the place (as Chicago is), and if they were also popular destinations in and of themselves (as Chicago, Portland and Seattle are) then it might make sense to terminate trains there, because the value of the connections would outweigh the inconvenience of making folks transfer.

But if all you're doing is kicking people out at Minot and Spokane, then you might as well not even bother running the Minot-Spokane train, because, seriously, where is the local O&D demand along that route? There isn't any.

I did not present my line of thinking behind this idea, so here goes- I love the LD trains as they are now as much as all of you, but looking at the state these trains are in, with all the cuts coming in, there will come a time soon when Amtrak will have to decide if thy want to project themselves as a luxury land-cruise experience, or a not-widespread but functional transportation service.

"All the cuts coming" represent, besides the loss of wine and cheese tasting, things that most passengers would neither notice nor care about. Believe it or not, 10 years ago the Empire Builder didn't have wine and cheese tasting on board, nor the amenity kits. Even with the loss of the free newspaper and the lack of flowers on the table in the dining car, the long-distance trains are still in better shape today than they were during the George Warrington years.

George Warrington, before thankfully resigning to go over to NJT, was actually a few weeks away from putting 180-day discontinuance notices on every long-distance train in the system. Then David Gunn took over and found out that Amtrak was actually just weeks away from totally running out of cash and shutting down (the whole company, including NEC, every station Amtrak owned/operated/dispatched; meaning even commuter services in New York and Chicago would be disrupted). Amtrak had to go to a Republican-controlled congress, with an anti-Amtrak president, and perhaps the worst and most anti-Amtrak DOT secretary in history, and beg for emergency money just to keep running.

Amtrak is nowhere near that position today. They are a million times better. The LD trains aren't under any specific threat that hasn't existed every year since 1971.

The one thing Amtrak really can't do and expect to continue to receive federal funding is "project themselves as a luxury land-cruise experience."

If former, keep the LD trains as they are now, complete with the wine and cheese, flowers, champagne and what not, to attract the high paying long distance passengers who come for the experience more than as a practical mode of transport. If latter, like it or not, we have to think out of the box ways to make Amtrak survive, and one of them is to cut them into more manageable bits and pieces.

No, it's not cutting them into "more manageable bits and pieces." It's cutting them into totally useless bits and pieces that will just die off fairly quickly.

An 8 to 16 hour train route is easier to manage than a 45 hour mammoth in terms of scheduling, with padding and recovery times etc, logistics of food and catering supplies for such a long journey, planning for contingencies and so on.

Having actually scheduled trains, I don't necessarily agree with the premise that an 8-to-16-hour route is easier to manage than a 45-hour trip, as long as the assumption is that those 8-to-16-hour trips all have to connect to each other. If they don't, then fine, but that just further fragments the service to the point where the orphaned middle segments will die off.

If they do have to connect, then planning for contingencies actually gets worse because now you have to have staff on hand (other than the couple of ticket/baggage agents) at Minot and Spokane to handle misconnects when trains are late, whereas previously, there was no connection, the passenger just stayed on the train and kept going to their destination.

So, if the one or two hundred or so folks traveling past Minot happen to misconnect, what hotel are you going to put them in that has 100-200 rooms available? What transportation are you going to have available to get them to those hotels?

Catering might appear easier, until you realize that you now have to set up a commissary in two new cities. That's an extra cost. Maybe it will be offset by extra revenue from trains being less likely to run out of stuff at the end of the trip (extra revenue which will only materialize if the ridership is there to support it, but it will likely be less anyway, thus depressing food and beverage revenue).

Also, since Amtrak heavily relies on third parties (host railroads), having trains running smaller segments is also better for passengers, since for example a derailment in North Dakota will delay the trip for Montana to Washington bound passengers if there is only one LD train running the entire length, but with three segments,the only ones affected are the ones traveling that segment only (and the few connecting ones, which are, as per statistics above just about 10%.. can't please everyone sorry!).

Already noted that the number of connecting passengers will be above 10%, since everyone traveling from anywhere east of Minot (even Rugby, the next stop) to anywhere west of Minot (such as Stanley, the next stop) is now a connecting passenger. Likewise those traveling through Spokane.

The Hi-Line may not be the ideal example for such an experiment, but a route like CZ that has major cities along the way can be run as multiple corridor services with lesser amenities than current LD, but still much more convenient than taking cross-country buses. With the state of the economy and government attitude towards rail in the US, sorry to say but we might have to live with this reality coming sooner or later.

California Zephyr could possibly run in segmented avatar as follows-

Zephyr One:

CHI 3.00pm

DEN 8.00am

DEN 1.00pm

CHI 6.00am

Zephyr Two:

DEN 3.00pm

SLC 6.00am

SLC 3.30pm

DEN 6.30am

Zephyr Three:

SLC 2.00pm

EMY 7.00am

EMY 5.00pm

SLC 10.00am

Traveling through the most scenic areas at night. That makes sense.

but if the choice comes to having a full Diner vs having no LD trains at all, what would you choose?

False choice. Thankfully, we're not to that point yet.

Does this make more sense now? Or does someone still need Advil? :wacko:

It didn't make sense when you suggested the idea for the Builder, and it makes even less sense on the Zephyr, given that there isn't much wrong with that route operationally, and you've turned a two-night trip into a three-night trip for those that do want to travel the full way, and haven't really made service any better for those who retain the one-seat ride.

 

[neroden]

And of course the trouble with the idea of just running them as day trains is that they run too long.

So cut them down to appropriate lengths.

If you do that, you just threw away all the major markets. And I do mean *all* of them.

You can't "cut them down to appropriate lengths". What you have to do is SPEED THEM UP to appropriate runtimes. Which is what I've hammered at for a very long time. It's really about running the trains fast enough to run them as day trains.

 

[neroden]

30% of the train's ridership is to or from stations south of Atlanta which is in line with the percentage of the route that that segment represents.

Again, that's an abuse of statistics. The section north of Atlanta has high fares and sells out; it's capacity-limited. The section south of Atlanta has fares which barely add to the fares to Atlanta, and it's hard to keep it full; it's limited by ability to charge fares.

Obviously Amtrak is managing fare policy to try to keep the train loaded, because *that's what you do*.

Amtrak actually studied the occupancy and wanted to cut off cars at Atlanta. That's not a symmetrical loading. Amtrak routinely agrees to shut the Atlanta-New Orleans service down 4 days a week for two months every year! Amtrak doesn't agree to do this north of Atlanta.

As for much faster tracks, it averages 43 mph there; New York to Atlanta is an average of 47mph.

Compare road timings. This is actually a significant difference.

 

[neroden]

Of course, a bit of heresy comes to mind...I can't help but wonder if it wouldn't make sense to make the Zephyr UP all the way. If I'm not mistaken, UP's mainline condition is at least as good as the ex-CB&Q under BNSF. Of course, I think there's a track gap between Union, CO and Denver, CO.

I don't know the speeds of the UP line through Nebraska and Iowa, but I think it crosses several bits controlled by other dispatchers.
The UP line through Kansas, on the other hand... I also don't know the speed of the line, but it's definitely all-UP all the way to Kansas City.

George Warrington, before thankfully resigning to go over to NJT, was actually a few weeks away from putting 180-day discontinuance notices on every long-distance train in the system.

Then David Gunn took over and found out that Amtrak was actually just weeks away from totally running out of cash and shutting down (the whole company, including NEC, every station Amtrak owned/operated/dispatched; meaning even commuter services in New York and Chicago would be disrupted). Amtrak had to go to a Republican-controlled congress, with an anti-Amtrak president, and perhaps the worst and most anti-Amtrak DOT secretary in history, and beg for emergency money just to keep running.

Apparently Amtrak actually ran out of cash under Downs, though the DOT secretary was much friendlier.

(A company can usually keep running for days after running out of cash; payroll and power bills are intermittent.) So... hmmm.

There was definitely a lot of bad management after Claytor and before Gunn, with not one but two cash crises. Warrington had been put in an impossible position by Congress, of course. I don't quite understand how things collapsed so fast under Downs after Claytor left, though.

Having actually scheduled trains, I don't necessarily agree with the premise that an 8-to-16-hour route is easier to manage than a 45-hour trip, as long as the assumption is that those 8-to-16-hour trips all have to connect to each other. If they don't, then fine, but that just further fragments the service to the point where the orphaned middle segments will die off.

In regards to this, I would raise the question of how many of the passengers are crossing the connecting point.

If you split a service into two routes at a point where more than half the train turns over (Denver?) then it may make more sense to split it into two runs for improved OTP, with a planned connection but not a "we hold the train for you" connection, and allow the connection to fail sometimes. This is because you decrease the experienced result for less than 1/3 of your passengers but improve the experienced on-time performance for more than 1/3 of your passengers. (The other 1/3 are doomed in any case because they're on the late train. Think about it.)

That's the only circumstance under which it makes sense. Service pattern should follow demand pattern.

Similarly, on the other side, if more than half the passengers on one train are changing to the same connecting train, you probably want to find a way to provide direct service...

 

[Ispolkom]

Another Idea! (And no, I am not drunk in case anyone is wondering :giggle: )

While we are at the idea of single overnight trains instead of one that takes 2/3 days, would it work to create an intermediate class of sleeping accommodation for single travelers (students, cost-cutting business travelers on a budget) that costs less than present day roomettes? This topic has been discussed to death but here is one idea that can be implemented without requiring ordering new cars or refurbishing existing ones- what if, in one of the sleeper cars, all the roomettes are left open (seal the roomette doors in open position) and sell the upper and lower berths separately to single travelers at prices lower than roomette. There are valid concerns about privacy if sharing a closeable roomette with a stranger but if it can't be closed, now its a bit more open and relatively less "creepy". These days a lot of youngsters travel solo and use services like AirBnb, couchsurfing and hostels.. and as you might have observed, students and youngsters use LD Amtrak trains in a substantial number. This demographic won't mind sharing an open roomettes with another passenger if it means overnight sleeping facility at lower cost. Pricing can be straight up cost of roomette for 2 passengers divided by 2.

For example, CHI-DEN for 19 March,

Coach - $ 116

Roomette for 2 passengers - $ 568

Roomette if traveling solo - $ 452

New class: shared roomette berth for solo traveler - $ 284 (568/2)

$116 to $452 is quite a big jump for one overnight on train, even after considering one night hotel cost + meals.. but $116 to $284, now that is a bit more lucrative to consider.

MDW-DEN on that date is $89 on Frontier. $89 to $284 is a big jump, considering you'd be sleeping on a 2-foot-wide bunk. I can't see how any price-sensitive customer would ride a sleeper.

 

[jis]

I would imagine that since folks do travel single in roomettes paying $452 there would a considerable number traveling paying $284. The question is will there be enough to make it revenue positive overall.

Hey they could go the the way it is done in some other realms, where you can either buy by the berth or buy rooms. If you buy a room then you can get upto the capacity of the room worth of people to travel in the room. Transport of people in the room is included in the room charge. So in the above scenario if you prefer to use the room all by yourself you still get to pay $582. Or you could buy by the berth and pay half. The logisitcs of managing that when you have one car on the train with 6 roomettes available would be mind boggling. But in these other realms usually there are 10s of Sleeping Cars per train with 50+ or more rooms available to divvy up.

At the end of the day perhaps the focus of this sort of discussion should be on capturing the city pairs that do not have excellent air connections IMHO. Clearly there is something going on that we do not understand or are unable to take into consideration when in spite of Sleepers being clearly way more expensive than alternative air arrangements, they still keep getting sold out.

 

[afigg]

I did not present my line of thinking behind this idea, so here goes- I love the LD trains as they are now as much as all of you, but looking at the state these trains are in, with all the cuts coming in, there will come a time soon when Amtrak will have to decide if thy want to project themselves as a luxury land-cruise experience, or a not-widespread but functional transportation service.

"All the cuts coming" represent, besides the loss of wine and cheese tasting, things that most passengers would neither notice nor care about. Believe it or not, 10 years ago the Empire Builder didn't have wine and cheese tasting on board, nor the amenity kits. Even with the loss of the free newspaper and the lack of flowers on the table in the dining car, the long-distance trains are still in better shape today than they were during the George Warrington years.
George Warrington, before thankfully resigning to go over to NJT, was actually a few weeks away from putting 180-day discontinuance notices on every long-distance train in the system. Then David Gunn took over and found out that Amtrak was actually just weeks away from totally running out of cash and shutting down (the whole company, including NEC, every station Amtrak owned/operated/dispatched; meaning even commuter services in New York and Chicago would be disrupted). Amtrak had to go to a Republican-controlled congress, with an anti-Amtrak president, and perhaps the worst and most anti-Amtrak DOT secretary in history, and beg for emergency money just to keep running.

Amtrak is nowhere near that position today. They are a million times better. The LD trains aren't under any specific threat that hasn't existed every year since 1971.

The one thing Amtrak really can't do and expect to continue to receive federal funding is "project themselves as a luxury land-cruise experience."

...

Beyond dropping some amenities, I'm not aware of any major cuts coming for the LD trains. Despite a dysfunctional Congress with some very hostile members, Amtrak got $340 million in operating subsidy for FY14 which is enough to keep the LD train running (with the state corridor subsidies now kicking in). We will see what happens for FY15 and the re-authorization act.
As of the end of the first quarter of FY14, Amtrak's cash account was in damn good shape compared to earlier times. Amtrak ran up a adjusted operating cash surplus for the 1st quarter, although that was in part due to delayed capital spending. Amtrak adjusted losses will grow over the rest of the fiscal year, with the January and February weather disruptions not helping at all, but if Congress continues to provide $300 to $350 million in operating subsidies, the LD trains will survive. What Amtrak really needs is an increase in annual capital grant funding so it can upgrade the infrastructure, including shared cost projects on the LD routes, and begin to order replacement and expansion rolling stock. Breaking the western LD trains into short segments won't contribute to solving the capital funding shortfall.

 

[Bob Dylan]

Interesting thread, thanks to all the Posters and the Stat Geeks for Crunching the Numbers! Amtrak is very much in need of New Ideas and Better ways to Improve Service without Increasing Costs!

As an old time Passenger Train Rider (I go back to the 1940s) in the US, and having experienced Trains in other Countries, I'm a Big proponent of Slumber Coaches! I do feel that every one made available on Overnight Trains would Sell Out! Other considerations are having Passengers pay for their Meals in the Diner! (Food and Beverage Services, even in Pullmans heyday, have Always Lost Money) Buffet/Cafeteria Cars (that Buffet/Dome Car on the Auto Train in the 1994 Video looks Cool! ), and Contracting out Food Service! ( I'm not Anti-Union, I'm a Life Member of 2 Unions! The Displaced OBS could be moved to other badly needed OBS Positions so no-one would Lose their Job!))

Splitting LD Trains into Sections, as was said, isnt Cost/Benefit Feasible but since Joe Boardman has promised the Budget Hawks in Congress that Amtrak Will Cut Losses on Food and Drink Service and LD Trains, the Suits @ 60 Mass need Ideas and feedback from Customers and Employees! Stopping Wine and Cheese Tastings, Newspapers, Cranberry Juice and a Flower on your Diner Table isn't going to Stop the Millions of Dollars in Losses! Hopefully besides Posting on AU and other Rail Forums,folks will send their Ideas/Suggestions for improvements and changes that wont Increase Costs to Amtrak!

 

[zephyr17]

Nathaniel, just wanted to say thanks for this extremely well-reasoned analysis. You make a compelling case. Kudos!

Bleah. Screwed up my quoting again. Most of that text is mine. Anyway...

Also for consideration:
With full occupancy, at average fare yields, per mile a coach car earns more gross revenue than a sleeper car on the following trains: Silver Star, Silver Meteor, City of New Orleans, and Crescent.

Lies, damned lies, and statistics. Occupancy rates are different, of course (and, specifically, they're usually lower in coach). Also, you can't multiply the "average fare yield" by the occupancy rate to get the gross revenue.

Worse for the spreader of these bad statistics, demand is different for different trips: in general, the sleeper trips are trips which would not be taken in coach; replacing the sleeper with a coach would just lose the income. The coach occupancy rate would then be lower and so would the average fare yield in coach, but hey, you would have replaced the "less profitable sleepers" with "more profitable coaches", right? Wrong.

Paulus, you obviously didn't read my essays in the other thread: I can't really blame you, since they were so long.

The key point: the problem is not amenities, the problem is route segments.

The Crescent suffers from very low demand south of Atlanta, which is undoubtedly distorting the numbers. It's inadvisable to be running so many sleeping cars south of Atlanta, because they're running empty or selling cheap. Then again, it seems to be inadvisable to be running so many coaches south of Atlanta, too, since they're also running empty or selling cheap.

This is completely screwing up the accounting there. With a proper Atlanta station, one could run the full Crescent to Atlanta and a continuing train (with very few cars) onward to New Orleans, and that would actually make sense. And the numbers for the train *north of Atlanta* would show that the sleepers get higher revenue per car than coaches. Or with much faster tracks south of Atlanta, people might actually ride the train south of Atlanta.

The Silver Star suffers from a similar problem to the Crescent, in that nobody's going to buy a roomette from Miami to Tampa. It really needs to terminate in Tampa and connect to a day train to Miami -- the problem there is the lack of a maintenance base in Tampa. There's a similar, smaller problem selling roomettes from Orlando to Miami on the Silver Meteor.

Further, the Silver Star and Silver Meteor have both been badly hurt by trackwork in Florida and by weather disruptions which prevented the trains from running their entire route -- this *particularly* hurts ridership (and the ability to charge high prices) on the longer trips, which are the ones taken by sleeper.

It does not surprise me that those three trains *currently* have sleepers underperforming relative to coaches, but I believe this is entirely a matter of special conditions due to bad trip segments.

I will say that the CONO does seem to be underperforming on sleeper revenue, without obvious problem segments or disruption, and it would be nice to figure out why. I think I know, however. The big markets along the line are:
- Chicago to Carbondale and points north: mostly take Illini/Saluki corridor trains
- New Orleans to Memphis and points south: it's a day train here, nobody will buy sleepers
- Chicago to Memphis, Jackson, and New Orleans: this is it for the sleeper demand, basically three city-pairs. And none of the connections make any sense at New Orleans, so anyone doing this is actually travelling to or from one of those three cities.

The CONO as a whole is a particularly short train and has relatively poor ridership, showing low demand overall. The speeds on the route are slower than driving but not outrageously slower. I must attribute the problems primarily to low population along the route. Secondary problems may include poor connecting transit, low awareness of trains in the region, and a lack of social/cultural links between Mississippi/Louisiana and Chicago.

The LSL has a similar overlap with a "corridor" at its east end, but the sleeper demand includes Chicago to Buffalo/Rochester/Syracuse/Albany/Boston/NY, which is a lot more demand. Especially given that there are closer cultural links between Chicago and the Northeast than between Chicago and Louisiana/Mississippi, and there is actual same-day connecting service in multiple directions at Chicago, New York, Boston, and even Schenectady and Buffalo.

Hooray for network effects. Understand them. Use them.

I don't want to remove any of the remaining transcontinental connections purely for network effect reasons; severing the network is always bad.

They are quite impractical until there is more underlying network for them to connect to, however. The three-a-week services barely even provide a network effect because their scheduling is so impractical.

The trouble with the idea of chopping up the transcontinentals to provide more shorter-distance service is that the tradeoff doesn't actually exist; you wouldn't get more shorter-distance service, which requires negotiations with Class I railroads, state support, track purchases, etc. (That said, if there was a good shorter-distance service all ready to go which *just* needed a bunch of sleepers and observation cars, I'd reconsider. Tell me when that happens or pigs fly; the sticking point is always, always, track improvements and slots.)

The trouble with the idea of converting them to a bunch of connected trains (whether day trains or not) is that it would actually be far worse for the bottom line -- not least due to the new maintenance bases required.

And of course the trouble with the idea of just running them as day trains is that they run too long.

The only reason to chop the routes up would be to get better On Time Performance. Given the nightmare which has been Amtrak OTP, if it did give significantly better OTP on either half, I'd probably say, chop the route in half now. But where would this be the case? This might be true for splitting the Zephyr at Denver (remove the UP / BNSF handoff) or terminating the Coast Starlight at Portland (remove the UP / BNSF handoff) but I don't see why it would help anywhere else -- and both of those would require new maintenance bases.

 

[neroden]

Clearly there is something going on that we do not understand or are unable to take into consideration when in spite of Sleepers being clearly way more expensive than alternative air arrangements, they still keep getting sold out.

Air service in the US is so unpleasant these days that a lot of people -- including me -- avoid it if at all possible, and will spend a large premium to avoid it. I don't know how to quantify that factor, but it's enormous.
If air service went back to the days of 1950s Pan Am, I'm sure it would be different. There would still be a large collection of "won't fly" people but not nearly as large as there is now.

 

[Anderson]

I would imagine that since folks do travel single in roomettes paying $452 there would a considerable number traveling paying $284. The question is will there be enough to make it revenue positive overall.

Hey they could go the the way it is done in some other realms, where you can either buy by the berth or buy rooms. If you buy a room then you can get upto the capacity of the room worth of people to travel in the room. Transport of people in the room is included in the room charge. So in the above scenario if you prefer to use the room all by yourself you still get to pay $582. Or you could buy by the berth and pay half. The logisitcs of managing that when you have one car on the train with 6 roomettes available would be mind boggling. But in these other realms usually there are 10s of Sleeping Cars per train with 50+ or more rooms available to divvy up.

At the end of the day perhaps the focus of this sort of discussion should be on capturing the city pairs that do not have excellent air connections IMHO. Clearly there is something going on that we do not understand or are unable to take into consideration when in spite of Sleepers being clearly way more expensive than alternative air arrangements, they still keep getting sold out.

Jis, you largely hit on the head what I was thinking on this front. The key here is for markets that are unreasonable to reach from at least one "well served" city; in the CHI-DEN example, you've got several stops in western IL where you'd have a bit of a drive into CHI. In Colorado, you've got stops up in the mountains that would involve a long (and often unpleasant, at least in the winter) drive into town and then across town to catch a plane. An $89 flight quickly ceases being a bargain when you have to spend $50+ getting to/from the airport, pay for parking, etc.

 

[tomfuller]

I gave a shorter train idea last year for the Empire Builder. I doubt that the host railroad would go for it even if Amtrak asked nicely.

My idea was an EB light that runs from MSP to SPK leaving MSP about 3PM Central every day and arriving in SPK about 5 PM Pacific (28 hours). The consist would be 1 engine, 1 baggage, 1 transdorm sleeper, 1 dining car, 1 lounge car/cafe and 3 coach cars.

There would be about 7 hours for turnaround in Spokane to be hooked in with the current schedule eastbound EB. Depending on how late the eastbound is into MSP there should be a couple hours to turn it for the 3PM departure.

The passengers that get off in SPK at 5-6PM can catch an Amtrak bus to Portland about 5 AM and have a guaranteed connection to the southbound CS at 2:15PM. Of course if the regular EB gets to SPK by 5 AM, they could ride the train to Portland to connect.

 

[Green Maned Lion]

Gosh this is worse micro management then Mica, with about as much understanding. I admit passenger amenities are suffering under Boardman. Now, dunderheads, explain to me on what evidence he is doing a bad job? Amtraks operating loss is less than it's ever been. Trains are running full, whether you can comprehend the why or not, at high fares.

If I ever need a manager, I'd love somebody who improved ROI as well as Boardman has.

 

[June the Coach Rider]

OK, let me get this straight, you want me to take the CZ from EMY to DEN, sit in DEN for 5 hours only to get on a second train to CHI arriving at 6AM only to have to wait in CHI to catch my home train at 9:30PM. It would make me stop riding for sure. I thought the 6 hours in CHI is a long time, Plus being disabled, changing trains just because, well that is not something I would want to do. Like has been said before, if someone wants to ride from Salt Lake to Omaha, they would have to take two trains and wait 5 hours in DEN for a total of 19 hours rather than just 14 hours on the LD train with no transferring needed

 

[neroden]

Looking at flights from Chicago Midway or Chicago O'Hare to Denver, there is no direct first class, and I don't think there's any business class either.

Sure, a sleeper compartment is a lot more expensive than cattle class.

But the sleeper clientele is the clientele who does not want to travel in cattle class. (They probably don't want to drive out to O'Hare or out to DIA either.)

They are able and willing to pay a large premium to avoid cattle class. This clientele, however, is not rich enough to afford chartered private planes. (And nobody is going to add a stop to their route just in order to get a first class ticket -- even if you did, the first class one-way plane ticket from Chicago to Denver would be upwards of $590. I checked!)

This is the core of your sleeper market.

This is why I say that this clientele who requires a certain level of amenities and will pay to avoid travel which lacks those amenities. But if you don't provide that level of amenities, they (we?) simply won't travel, and you'll have lost revenue. (Or, what's worse, they'll travel but provide negative reviews of the service to all their friends and quite possibly online.)

This level of amenities isn't really that high -- it's what you might call "genteel". A bed at night; sit-down meals at mealtimes; sanitary conditions; a place where they can stretch their legs; free water and coffee.

For a long time-duration trip which is running across mealtimes and sleeping time, this means the facilities they might expect to see at a middle-class hotel (maybe a Best Western) or a middle-class restaurant (maybe Denny's). Bringing back china would probably boost the perceptions of Amtrak by this clientele (as well as being more environmentally conscious). (I think this clientele would generally be fine with buffet service if it proved practical, which I have my doubts about.)

Now, *some of this clientele* will tolerate sleeping in a chair for 8 hours... but most won't. *Some of them* will tolerate bringing all their own food... most won't. Most of them will tolerate "cafe" food service for one meal, or even two.... but at the third they'll start wanting something better than reheated junk food. These are people who are used to eating at least one real meal every day and who like their creature comforts. The more the trip smacks of "roughing it", the less interested they are.

But again, these are middle class people, not big corporate CEOs. A few hours in a seat in a crowded coach, *between* meals? No problem. So the appropriate amenities depend on the trip time.

I will also say that these are mostly not going to be frequent travellers. If they were they would be able to get all kinds of deals on bypassing the TSA, special treatment by the airlines, etc., but they're not. Most people do not make intercity trips very often, maybe a few times a year.

And to dispel another myth, they are not generally taking "land cruises" -- a cruise being a trip in a circle for the sake of taking a trip. They have somewhere specific to go. They are not in a huge hurry and have schedule flexibility, but they're trying to get from point A where they live to point B where they have something to do, and then back again -- perhaps for a wedding or a graduation or to go to school or for the holidays or vacation or even on business. They're just trying to do it comfortably.

So just like everyone else, *they care about average speed* and *they care about on-time performance*.

I remember one woman returning to Breckenridge CO (getting off the CZ in Denver) from a visit to her family on the East Coast. I think it was three trains and four buses for the whole one-way trip -- with the trains doing the longer legs. (The amount of bus travel she could tolerate was impressive.) She was a regular Amtrak rider, and had a roomette; she was willing to spend substantial money to not fly. But she immediately commiserated about and commented on the lack of OTP.

Bad on-time-performance has been documented to drive down ticket yields and ridership massively, on the Cascades, the Wolverines, and other routes, as well as in other systems. The same is true for sleeper service. It takes over a year of solidly good OTP for ridership to recover from a history of bad OTP.

Now, "breaking" a train decreases comfort, and increases effective travel time, so it's a reduction in service quality. But OTP is more important to service quality: This is why I say that the only good reason to "break" a train is if it will actually cure the OTP problem for a majority of riders.

If you could get an absolutely reliable on-time Denver Zephyr by breaking the train at Denver, I'd say do it. (Schedule the connection to be quick, but if the connection is missed, people stay overnight like they sometimes do in Chicago.) But I see no evidence that this would actually work, since OTP problems are across the board rather than localized.

 

[afigg]

At the end of the day perhaps the focus of this sort of discussion should be on capturing the city pairs that do not have excellent air connections IMHO. Clearly there is something going on that we do not understand or are unable to take into consideration when in spite of Sleepers being clearly way more expensive than alternative air arrangements, they still keep getting sold out.

Jis, you largely hit on the head what I was thinking on this front. The key here is for markets that are unreasonable to reach from at least one "well served" city; in the CHI-DEN example, you've got several stops in western IL where you'd have a bit of a drive into CHI. In Colorado, you've got stops up in the mountains that would involve a long (and often unpleasant, at least in the winter) drive into town and then across town to catch a plane. An $89 flight quickly ceases being a bargain when you have to spend $50+ getting to/from the airport, pay for parking, etc.

There is a underlying change in the domestic travel patterns at US airports. I saw several weeks ago that United Airlines was dropping Cleveland airport as a hub as a consequence of the merger. The industry has been consolidating to fewer and bigger hub airports which means ever more travelers in the smaller cities have to connect through hubs or sometimes even 2 hubs to get where they want to go. This of course is a trend that has been underway for several decades, but as it continues with industry consolidation, means that air travel options and travel times are getting poorer between the smaller to medium sized cities and places outside of the largest metro regions.
When the 2013 air travel stats are released, should look to see what the airport & air passenger growth and decline patterns are. How much of the domestic air travel market is now confined to and between the largest 30 US airports?

 

[Texan Eagle]

OK, let me get this straight, you want me to take the CZ from EMY to DEN, sit in DEN for 5 hours only to get on a second train to CHI arriving at 6AM only to have to wait in CHI to catch my home train at 9:30PM. It would make me stop riding for sure. I thought the 6 hours in CHI is a long time, Plus being disabled, changing trains just because, well that is not something I would want to do. Like has been said before, if someone wants to ride from Salt Lake to Omaha, they would have to take two trains and wait 5 hours in DEN for a total of 19 hours rather than just 14 hours on the LD train with no transferring needed

As has been repeated multiple times in this thread- people like you who travel super long distance by Amtrak are only about 10% of total passengers, so yes, when you are in minority, you might have to adjust a bit. Why is changing trains in DEN such a big issue while changing trains in CHI is acceptable? Why don't you suggest that all LD trains should run coast-to-coast so that you don't have change trains in CHI? For short distance travelers across train boundaries, even today passengers traveling east of Chicago to west of Chicago have to endure a long layover in CHI. Similar for north of SAS to east or west of SAS on Sunset/Eagle. This is the reality of public transport- it cannot work for everyone from their door to door, if you want to travel long distances, you need to connect or change trains.

 

[Texan Eagle]

Looking at flights from Chicago Midway or Chicago O'Hare to Denver, there is no direct first class, and I don't think there's any business class either.

Says who? Just checked on Kayak for randomly chosen mid-week date March 19- both American and United offer multiple non-stop flights daily with First Class O'Hare to Denver for $394, cheaper than Amtrak's sleeper fare of $452 for the same day. Here is where the per-berth sleeper class I suggested comes into play- for someone who has a morning meeting in Denver (or other way round) usual practice, at least everywhere I have worked so far is to fly out previous evening and stay overnight to avoid problems due to delays if one were to take morning flight straight into the meeting. Now instead of roomette for $452, if you can offer sleeper berth with dinner and breakfast for $284 (see pricing logic from my previous post), you have created a lucrative market for yourself.

Currently: Amtrak $452. First Class flight + a night in hotel = $394+$100 = $494 .. not much difference

With Per-berth pricing: Amtrak $284 vs flight + hotel $494 ... if you can save ~$200, many folks would be willing to start from work a little early previous evening and take the train.

Now before you point out lack of privacy, let me point out that even high-level business folks traveling international Business Class or First Class on flights sleep in open layout flatbed seats where you can actually look at co-passengers sleeping by your side, having an upper berth-lower berth arrangement actually offers more personal space than a flight's high-end cabin.

 

[JoeBas]

I think it's time to subject this argument to its logical reducto ad absurdium base...

From now on, all trains shall only run between adjacent stations. If you wish to go 3 stations down the line, that will be 3 train changes. Most people aren't going that far anyway and won't mind getting out of their seats every couple of hours at 2:30 AM, and if they complain, they should drive or fly. Why we're doing this I have no idea, other than to show that we're thinking outside the box.

 

[Green Maned Lion]

No they took logical thought, put it in a box so reality couldn't get to it, and then retrieved its output later.