PRRIA 2015

[fulham]

Yesterday the House Transportation and Infrastructure Committee introduced the Amtrak reauthorization bill, PRRIA 2015.

For FY 2016 the following amounts are in the bill:

Long Distance: $973 million

NEC: 439 million

Infrastructure: 300 million

There is also a note that Amtrak can borrow up to $14 billion through the FRA RRIF loan program for the NEC.

The food/beverage loss provision is still in the bill, but overall I think, given the current situation in Washington, this is not too bad. Boardman said last year in a Railway Age article that Amtrak needed $618 million for operations and $295 in capital for the LD network.

Since operationally the NEC is self-sufficient, the $439 million should be able to go to further rebuilding of the NEC, and the $14 billion RRIF loan program hopefully will be used for the B&P tunnels, Portal Bridge, etc.

Don't know what Amtrak was exactly asking for but it looks like this it what they are going to get.

 

[neroden]

If the food & beverage provision stays in for the full five years, I think most of the Western dining cars are dead. (The Empire Builder, which has the highest patronage, and the Coast Starlight, which probably has the second highest, might survive.) The recent menu cuts have very clearly caused sharp drops in patronage, and presumably revenue; I expect the revenue drops have exceeded the cuts in costs.

I rode the SWC recently, both ways. I've taken it at this time of year before, and I have never seen so few people in the dining car. Breakfast demand was holding up strong, but lunch and dinner were really empty. I've never seen so many sleeper passengers skipping their free dining car meals, and coach demand was far lower than I remember from even two years ago.

I certainly assisted with this process, as I skipped four dining car meals (due to lack of ingredients information leading to lack of selections -- basically I could have salad or steak, because I couldn't find out what was in anything else) and never had dessert (due to lack of selections). Although I was in sleeper, this means less revenue transferred from the sleeping car account to the dining car account. Even the people who had the meals were complaining about them -- except breakfast. Breakfast is still OK.

The losses on these cars were already too large to cover with the cafe car profits, and now they're going to be significantly larger. And people who are less hostile to car and air travel than I am will stop riding. Meanwhile, Amtrak will have to deal with more carry-on luggage from all the people bringing their own food! I feel sorry for the chefs, who are clearly doing the best they can with what they're given, but they can't do much with it.

The Eastern (single-level) dining cars might survive because they've got much better patronage to start with, and that patronage is spending substantially more money to start with (higher ticket prices per mile), and there's less staffing required (shorter runs), and the cafe cars are mobbed to the point of overcrowding and running out of stock. The Eastern dining cars can probably be brought to a level where their subsidy can be covered by cafe car profits. The Eastern trains are profitable anyway (before arbitrary overhead allocation); their profits can be assigned to cover dining car costs, thus meaning that *Congress* doesn't subsidize them, which is the actual law. The Western trains still lose money on operations even before overhead, so they can't do that.

We might hope to have this provision deleted by the Senate, because it's stupid. However, the Eastern dining cars could probably survive it; the Western cars cannot. I guess Amtrak will just hope that the provision is removed before the 5-year deadline. And I hope the F&B department stops shooting themselves in the revenue side of the ledger, which they are doing now.

----

With regard to the rest of the system (apart from F&B) the situation is OK. The "National Network" currently costs $21.8 million/year to operate... plus $507.7 million in overhead. If the amount in PRRIA 2015 is actually *appropriated*, Amtrak might be able to order some more much-needed Viewliners.

 

[Andrew]

What about new hudson tunnels?

 

[MikefromCrete]

If the food & beverage provision stays in for the full five years, I think most of the Western dining cars are dead. (The Empire Builder, which has the highest patronage, and the Coast Starlight, which probably has the second highest, might survive.) The recent menu cuts have very clearly caused sharp drops in patronage, and presumably revenue; I expect the revenue drops have exceeded the cuts in costs.

I rode the SWC recently, both ways. I've taken it at this time of year before, and I have never seen so few people in the dining car. Breakfast demand was holding up strong, but lunch and dinner were really empty. I've never seen so many sleeper passengers skipping their free dining car meals, and coach demand was far lower than I remember from even two years ago.

I certainly assisted with this process, as I skipped four dining car meals (due to lack of ingredients information leading to lack of selections -- basically I could have salad or steak, because I couldn't find out what was in anything else) and never had dessert (due to lack of selections). Although I was in sleeper, this means less revenue transferred from the sleeping car account to the dining car account. Even the people who had the meals were complaining about them -- except breakfast. Breakfast is still OK.

The losses on these cars were already too large to cover with the cafe car profits, and now they're going to be significantly larger. And people who are less hostile to car and air travel than I am will stop riding. Meanwhile, Amtrak will have to deal with more carry-on luggage from all the people bringing their own food! I feel sorry for the chefs, who are clearly doing the best they can with what they're given, but they can't do much with it.

The Eastern (single-level) dining cars might survive because they've got much better patronage to start with, and that patronage is spending substantially more money to start with (higher ticket prices per mile), and there's less staffing required (shorter runs), and the cafe cars are mobbed to the point of overcrowding and running out of stock. The Eastern dining cars can probably be brought to a level where their subsidy can be covered by cafe car profits. The Eastern trains are profitable anyway (before arbitrary overhead allocation); their profits can be assigned to cover dining car costs, thus meaning that *Congress* doesn't subsidize them, which is the actual law. The Western trains still lose money on operations even before overhead, so they can't do that.

We might hope to have this provision deleted by the Senate, because it's stupid. However, the Eastern dining cars could probably survive it; the Western cars cannot. I guess Amtrak will just hope that the provision is removed before the 5-year deadline. And I hope the F&B department stops shooting themselves in the revenue side of the ledger, which they are doing now.

----

With regard to the rest of the system (apart from F&B) the situation is OK. The "National Network" currently costs $21.8 million/year to operate... plus $507.7 million in overhead. If the amount in PRRIA 2015 is actually *appropriated*, Amtrak might be able to order some more much-needed Viewliners.

By skipping the meals you probably helped improve the diners' bottom line. I don't see diners being removed from the western trains. What are you going to do? Start stopping at stations along the way for meals? I really don't think the eastern LD trains are "profitable" by any measure.

 

[inthegauge]

Lose dining cars = lose trains

Its all part of the Republican plan. And Big A, gullible as ever, is giving into the demands of politicians who couldn't run a train around a Christmas Tree.

 

[PRR 60]

Lose dining cars = lose trains

Its all part of the Republican plan. And Big A, gullible as ever, is giving into the demands of politicians who couldn't run a train around a Christmas Tree.

This bill has sponsors from both sides of the aisle. It is considered bipartisan, or at least as bipartisan as things get these days.

Note that this is an "authorization," not an "appropriation." The authorization is kind of like a plan that the annual appropriation should follow. Whether it does to not is another story. The funding plan in the last authorization was not followed by even one annual appropriation.

 

[Blackwolf]

Received the weekly newsletter from MHSRA by email today, with an interesting take on PRRIA and Amtrak food service.

[SIZE=medium]The T&I bill calls for Amtrak to cease food service unless it can cover its costs in the next five years. At the same time, the bill stipulates that no personnel can be laid off to reduce costs, even though labor costs make up the largest part of the dining service budget.

At first it appears Congress has almost certainly ended food service. That is until you see their ‘out’. The bill also allows Amtrak to designate a part of ticket revenue to cover any deficits associate with food service. In effect, Amtrak will continue to fund on-board dining the way it always has.

The smoke and mirrors around food service is designed to create a distraction from the real issue: the continued starvation diet for Amtrak and passenger rail as a whole.[/SIZE]

Looks like Business As Usual for Amtrak when it comes to the diners. Meals are not going anywhere, and this might lead some evidence toward Boardman's "Diners breaking even" quest if the company is able to continue including ticket revenue to off-set F&B losses, albeit with new permission to do so and an additional layer of hidden accounting to make it happen.

 

[Bus Nut]

You would think so ... but that's not how the economics of LD trains work.

Terrible food will discourage repeat patronage, which causes losses to increase because LDs don't make profits ... strike one.

Every meal a high-paying sleeper section patron eats is credited to the diner, reducing losses attributed to that part of the business--remember, no diner on LD, no LD, it's that simple, that's your boundary condition--so sleeper patrons skipping meals is horrible for the short term and long term viability of these trains ... strike two.

Less sleeper patrons eating meals means pressure to increase the already high nominal prices to fix the accounting, driving coach patrons away from the expensive fare that's not worth the price, in other words you hit a pricing death spiral ... strike three.

Congress is meddling because "$10 cokes" gets you on TV for 15 seconds. And it takes longer than 15 seconds to explain things like cross-subsidization, cost-death spiral, the difficulties of attributing costs, revenue, and business value to different parts of the customer service experience ... just for starters.

Just like people on boards like these get glib and blame LD losses on "labor agreements" as if Congress hadn't meddled in both equipment and staffing over the years making the whole personnel situation much more complex than running a McDonald's.

 

[Bus Nut]

Note that this is an "authorization," not an "appropriation." The authorization is kind of like a plan that the annual appropriation should follow. Whether it does to not is another story. The funding plan in the last authorization was not followed by even one annual appropriation.

Ain't that the truth. Part of Amtrak's "malaise" 15 years ago was the fact that Congress never appropriated the funds authorized, after Amtrak had committed to sinking billions in shmancy problem-plagued trainsets built by a VERY politically connected company, which put Amtrak in debt and forced them to burn cash on debt service every year. Amtrak's very lucky that demand for public ground transportation services has grown so dramatically.

 

[Train2104]

House committee passes bill, NEC cross-subsidization prohibited.

http://www.nj.com/politics/index.ssf/2015/02/whats_raised_in_amtraks_northeast_corridor_should.html

 

[Train2104]

There's also a pet pilot requirement, paid for by pet fees:

http://www.huffingtonpost.com/2015/02/12/dogs-on-trains_n_6672722.html

https://www.congress.gov/114/bills/hr749/BILLS-114hr749ih.pdf

 

[Anderson]

You know, I've honestly wondered for some time why we even needed a new authorization bill. Can anyone answer that one?

 

[neroden]

By skipping the meals you probably helped improve the diners' bottom line.

I happen to know how it works. I made the diners' bottom line worse and I made the sleepers' bottom line better.

I really don't think the eastern LD trains are "profitable" by any measure.

Most of the eastern LD trains are profitable in the following strict sense: cancelling one of them would cause Amtrak's bottom line to get worse. They are a benefit to the bottom line.

There's really no question about this. Boardman made it quite clear what the situation was regarding the 2012 financials on page 11 of this presenation:

http://www.amtrak.com/ccurl/778/373/Amtrak-Covers-88-Percent-of-Operating-Costs-ATK-13-022.pdf

The standard monthly presentations of Amtrak numbers misleadingly allocate system-wide overhead (like the cost of the reservations system) to individual trains. This presentation does not make that mistake. You will also notice that in the presentation, Amtrak's scenario map for Congress not funding long-distance trains... is to keep the eastern trains. This is because they were either profitable, or expected to be in the near future.

(Long discussion at http://discuss.amtraktrains.com/index.php?/topic/54279-boardman-testimony-5-mar-2013/ )

Since then, revenues have risen substantially for nearly every train, Empire Builder excepted. And direct costs have dropped noticeably for nearly every train as well. (Unfortunately, overhead has risen.) So every train is doing better financially. However, the relative positions of the trains on the chart has mostly not changed (with the exception of the Empire Builder, which is a special case due to the disruptions). The result is that the eastern trains are mostly profitable now.

In 2012, the Palmetto and Silver Meteor were profitable while the Auto Train and LSL were roughly break-even; in 2014, it appears (from some very careful calculation) that the Palmetto, Silver Star, Silver Meteor, and Lake Shore Limited are all profitable, and the Auto Train is substantially profitable (almost enough to cover the overhead arbitrarily allocated to it).

The Crescent, Cardinal, and Capitol Limited all improved as well, but it seems that they still continue to be "in the red" on direct costs, though in the case of the Capitol Limited it may be generating enough connecting revenue on the Florida trains that it is actually beneficial to Amtrak's bottom line. (This is very hard to figure since Amtrak never releases connecting revenue numbers.)

The Coast Starlight and Texas Eagle have improved substantially financially speaking, but are still "in the red", and by more than any of the eastern trains. The CONO, unusually, appears to actually be doing worse financially in 2014 than in 2012.

The Empire Builder has had a collapse in revenue but it appears that costs have dropped faster -- perhaps this is BNSF paying compensation to Amtrak for recent delays, so I hesitate to analyze it further.

The other Western transcons have also improved financially, and by a lot -- but they were *much, much* more expensive to start with. Each of the CZ, SWC, and SL required over $16 million subsidy on a direct cost basis even in 2014 by my estimation, still being the three most expensive by a large margin.

 

[neroden]

House committee passes bill, NEC cross-subsidization prohibited.

NEC / non-NEC cross-subsidization is explicitly allowed by the bill. Read it. Amtrak is supposed to document exactly what transfers are made.

More interesting: The overhead on NEC and "National Network", currently running at roughly $1.5 billion dollars to what, is supposed to be allocated "proportionally".

Proportionally to what? Ridership? Train-miles? Whichever it is, it's quite clear this isn't being done currently. For example, a third of the Information Technology overhead is currently being dumped on the long-distance trains, and a third on the state-supported trains, which is *grossly* disproportionate. Fixing this sort of misallocation will make the monthly reports look a lot better for the long-distance trains (and a lot worse for the NEC).

 

[neroden]

You know, I've honestly wondered for some time why we even needed a new authorization bill. Can anyone answer that one?

We never need authorization bills. It's just a habit of Congress. A fair number of important laws simply "authorize and appropriate" in one go.

The "authorization bills" were invented when the "federal budget" was invented back in the Harding administration -- and frankly I consider the "federal budget" to be a terrible mistake. It was an early example of doctrinaire attempts to "run the government like a business", which makes no sense. The "authorization bills" were an attempt to get all the usual lawmaking stuff out of the appropriation bills, since the appropriation bills had been put into the "budget".

It doesn't work; lawmaking stuff is still put in the appropriation bills routinely (such as the infamous prohibition on Amtrak discounting more than 50% off peak fare). The authorization bills just create confusion.

 

[AlexandriaVA]

I don't know why many of the posters on here seem to be skeptical of the performance of Eastern trains. I notice that many of them tend to live in Midwestern or Western states, so perhaps that has something to do with it, but understand that the East has greater population density than most of the rest of the country. The area particularly between Alexandria and Boston has very dense and high-population urban centers, something that is very conducive to rail travel.

In other words, you can take a 10 minute walk from Union Station, 30th Street Station, Penn Station, or South Station, and pass by hundreds, if not thousands, of people. Add in the strong mass transit connections to the stations, and you can see why these Eastern trains can do so well. Additionally, most cities in the East are closer together.

It's not a knock at all on Western states...it's just that it's very hard by rail to profitable go long distance between cities with low population density. I don't know if the Western long distance trains will ever be profitable (but I don't believe that public transportation necessarily needs to be profitable).

 

[WoodyinNYC]

...

... not a knock at all on Western states ... it's very hard by rail to

profitably go long distance between cities with low population density.

Our friend Neroden is not optimistic about the Western trains. Take a longer view and see some hope. It is the Western cities that are showing big growth -- Phoenix, Denver, Salt Lake, San Antonio, Las Vegas, Seattle, even Boise for Heaven's sake. So one day the population problem will solve itself. Already, the route of the Sunset Limited -- three trains a week since before Amtrak was formed -- connects Houston, San Antonio, El Paso, Tucson, more or less Phoenix, Palm Springs, and L.A., with most of those places now among the top 20 by population.

Even where no growth is seen or will be seen -- the desert stretches between Salt Lake City and Reno, the mountains and forests between Sacramento (or Redding, CA for sure) and Portland -- Amtrak runs its LD trains at night. Would they lose much less money if Amtrak ran them at night thru more populous routes? So the equipment cost is between parking the trains overnight with zero revenue, or running them at night thru empty areas.

In some places, upgrading to 110-mp corridors like Chicago-St Louis could help some trains enormously. I could see heavy corridor service Chicago-Twin Cities at 110 mph, extending with a few trains at 79 mph to Fargo-Grand Forks. And for the Empire Builder, the stretch Spokane-
Seattle also grow into a corridor. The time savings from upgrading these corridors be smallish -- Chicago-St Louis is looking to cut about 50 minutes after the first Billion investment -- but the increase in reliability and onetime performance should be huge, while some costs like stations and marketing could be shared.

Other potential corridor upgrades on LD routes include Chicago-Denver and Bay Area-Reno; Chicago-Kansas City; Dallas-Ft Worth-Austin-San Antonio; and most of all New Orleans- Houston-San Antonio and Tucson-Phoenix-L.A.

I won't live long enuff to see the Western routes be "profitable" by any simple measure. But losses could decline over the years to more tolerable, or less controversial, levels if Congress doesn't have a panic attack in the meantime. LOL.

 

[wolfspirit]

Proportionally to what? Ridership? Train-miles? Whichever it is, it's quite clear this isn't being done currently. For example, a third of the Information Technology overhead is currently being dumped on the long-distance trains, and a third on the state-supported trains, which is *grossly* disproportionate. Fixing this sort of misallocation will make the monthly reports look a lot better for the long-distance trains (and a lot worse for the NEC).

I would guess seat miles. Doing either pax-miles or train miles would shift more of the overhead off the NEC. Ridership would swing more toward the state corridors, pax-miles would be more LD. Train miles would also be a shift, evenly balanced between state & LD, but much less on the NEC. (I am using the FY14 budget for the numbers split by department. Pages 80-81.)

 

[Anderson]

I guess another practical question is why NARP, et al push for a reauthorization bill. It would seem that (barring circumstances where a reauthorization would likely be productive) the smart thing would be to simply oppose any sort of reauthorization on the grounds of it being a risky minefield.

 

[Anderson]

Proportionally to what? Ridership? Train-miles? Whichever it is, it's quite clear this isn't being done currently. For example, a third of the Information Technology overhead is currently being dumped on the long-distance trains, and a third on the state-supported trains, which is *grossly* disproportionate. Fixing this sort of misallocation will make the monthly reports look a lot better for the long-distance trains (and a lot worse for the NEC).

I would guess seat miles. Doing either pax-miles or train miles would shift more of the overhead off the NEC. Ridership would swing more toward the state corridors, pax-miles would be more LD. Train miles would also be a shift, evenly balanced between state & LD, but much less on the NEC. (I am using the FY14 budget for the numbers split by department. Pages 80-81.)

I think it is train-miles that are used. I had an interesting discussion on this point with someone recently, and it seems that this was a bad practice that Amtrak inherited.

 

[neroden]

...

... not a knock at all on Western states ... it's very hard by rail to

profitably go long distance between cities with low population density.

Our friend Neroden is not optimistic about the Western trains.

And I'm happy to have a conversation about why. I should note that I am optimistic about bits and pieces of them. (I think the commitment and the demand are there for Washington-Oregon-California and for the Empire Builder route; and the population is there for the Texas Eagle route and for Chicago-Denver and probably Denver-Salt Lake.
I am most pessimistic about the areas where the states and localities refuse to put up any money, and the vast empty deserts.

Even where no growth is seen or will be seen -- the desert stretches

between Salt Lake City and Reno, the mountains and forests between

Sacramento (or Redding, CA for sure) and Portland -- Amtrak runs its

LD trains at night. Would they lose much less money if Amtrak ran them at night thru more populous routes?

Yes. Good examples of this: Pittsburgh, Cleveland, Toledo. Even Spokane and Fargo are examples. Of course, Amtrak can't rearrange the national population to make the train routes more convenient; the western deserts are not going away.

In fact, there's going to be massive droughts coming to the Southwest; models predict bigger droughts than in the last 1000 years, more than the droughts which ended the Anasazi civilization. Phoenix will have to shrink. A lot. It really has no choice. Las Vegas too.

 

[neroden]

Proportionally to what? Ridership? Train-miles? Whichever it is, it's quite clear this isn't being done currently. For example, a third of the Information Technology overhead is currently being dumped on the long-distance trains, and a third on the state-supported trains, which is *grossly* disproportionate. Fixing this sort of misallocation will make the monthly reports look a lot better for the long-distance trains (and a lot worse for the NEC).

I would guess seat miles. Doing either pax-miles or train miles would shift more of the overhead off the NEC. Ridership would swing more toward the state corridors, pax-miles would be more LD. Train miles would also be a shift, evenly balanced between state & LD, but much less on the NEC. (I am using the FY14 budget for the numbers split by department. Pages 80-81.)

I think it is train-miles that are used. I had an interesting discussion on this point with someone recently, and it seems that this was a bad practice that Amtrak inherited.

It certainly isn't all allocated by train-miles. Probably some of it is, but I'm pretty sure the majority of it is not.

First, I find it hard to believe that the NEC runs a mere 1/3 of Amtrak's train-miles.

Also, the calculations of overhead on particular long-distance trains for the years where we've been able to extract them don't match up with train-miles. For FY 2012, the Silver Star was allocated $39 million in overhead, the Silver Meteor $39.3 million (it runs fewer train-miles), the California Zephyr $39.8 million (it runs something like twice as many train-miles), etc.

I suspect that a lot of the allocations are done by ass-pulling. The Information Technology allocation (1/3, 1/3, 1/3) looks especially suspicious.

 

[Anderson]

At least on the LD side of things, I'm suddenly stuck wondering if re-routing the Zephyr either as the Pioneer or the Desert Wind and keeping some sort of nominal bus/day train connection west of SLC would make sense. Essentially that's what happened back in the late 60s: WP and UP both dumped service on that segment as fast as they could (UP was going 3x weekly and WP was out entirely) while there was daily service both north and south via the City of Everywhere. Reno-Emeryville is a viable market (it could probably go multiple-daily were UP to allow it and swapping a Capitol Corridor train in for the Zephyr's slot probably would not be too much of a hardship).

I guess the big question is that of the three "holes" to choose from west of SLC, which one has the least deficit to work with (to Reno, Las Vegas, or Portland...yes, I know Boise is there, but IIRC the PNW also has a much thinner population than either NorCal or SoCal). Based on Amtrak's results for the Pioneer study (flawed though it was) and the Desert Wind option on the Zephyr, my best guess would be that going to LA is the best option of the three on raw financial performance while trying to manage both north and south would do respectably well, too (especially since if you just split the train in half, more or less, you'd be running "less train" on the emptier segments).

 

[twa904]

Proportionally to what? Ridership? Train-miles? Whichever it is, it's quite clear this isn't being done currently. For example, a third of the Information Technology overhead is currently being dumped on the long-distance trains, and a third on the state-supported trains, which is *grossly* disproportionate. Fixing this sort of misallocation will make the monthly reports look a lot better for the long-distance trains (and a lot worse for the NEC).

I would guess seat miles. Doing either pax-miles or train miles would shift more of the overhead off the NEC. Ridership would swing more toward the state corridors, pax-miles would be more LD. Train miles would also be a shift, evenly balanced between state & LD, but much less on the NEC. (I am using the FY14 budget for the numbers split by department. Pages 80-81.)

I've taken the Information Tech amount of 1.5 Bil and looked up the actual passenger numbers for the Sep 14 monthly reports and see that only about $220 Mil should have been allocated to the LD trains. $714 Mil should go the State Supported Trains and about $565 Mil to the NEC. I've allocated by passenger counts, that the way we did it in the airline industry. So if you took $280 away from expenses for the LD trains, some of these might actually profitable, Right?

 

[neroden]

I've allocated by passenger counts, that the way we did it in the airline industry. So if you took $280 away from expenses for the LD trains, some of these might actually profitable, Right?

By itself, this reallocation would make the Auto Train clearly profitable. The other long-distance trains would still appear "unprofitable" after overhead allocation, but some of them would be looking a lot better than most of the State Supported trains -- maybe on the order of a $2 million/year "loss" after overhead allocation for the Palmetto, or $13 million/year for the LSL; the per-passenger numbers would look pretty good.

Remember, at the moment the trains which are currently profitable before overhead allocation are the Auto Train, Silver Meteor, Palmetto, Silver Star (probably), Lake Shore Limited (probably), and Empire Builder (maybe, but this seems to be due to BNSF paying penalties). The "probably" and "maybe"s on that list don't have very much profit to contribute to overhead. Cancelling them would lose money for Amtrak, because the overhead would just have to be allocated somewhere else -- but they're certainly not gushing cash.

And the Cardinal and Capitol Limited still appear to be losing money before overhead allocation, though it's a very close thing (a daily Cardinal would be profitable, and the CL is probably generating connecting revenues higher than its subsidy).]

----

What all the trains really need is higher volumes of passengers and higher revenue. There have actually been massive positive swings in the direct-loss performance of most of the so-called long-distance trains over a mere two-year period, FY12 - FY14, with improvements ranging from $3 million to $14 million; the laggards are the CONO, Crescent, and CL which didn't seem to improve much at all over that period.

Another couple of years of revenue and volume improvements will make it clear that allocated overhead is the main cost of the "long-distance" trains. Already, the net subsidy of all of them put together, based on direct costs, is about $21.8 million -- plus $507 million in allocated overhead. Anything which can leverage that overhead while producing more direct profits is good for the bottom line -- Broadway Limited? Silver Palm? Silver Meteor on the FEC? Second LSL? More mundanely, longer trains on the profitable routes. And daily Cardinal service, of course; it seems quite clear that this would kick the Cardinal from loss into profit, before overhead is allocated.

Another problem is that overhead seems to be increasing. Quite fast. I hope this is temporary (redoing the IT system should not cost this much *forever*).