Do Long Distance Trains Really Lose Money

[henryj]

I read Fred Frailey’s Trains Magazine article on “What’s Next for Amtrak” with great interest. But what struck me most was the comment that the 15 LD train were expected to lose $530 million in 2012. I am a retired CPA so I like to work with numbers and I just found this hard to believe. So I went back and re-read Amtrak’s Performance Report for FY2011 and their PIP Plan reviews on certain LD trains and routes. What I found was shocking and eye opening. I used Amtrak’s own published numbers or extrapolated them to the trains that have not been reviewed yet. I wanted to see if these trains even covered their own operating expenses, that is fuel, rent to private RR($5 a train mile by the way, not free), switching, maintenance, labor T&E, labor OBS, commissary and station costs. What I found is that the trains we often associate with being a big success are in fact the biggest losers. The CZ, SWC, Empire Builder, Crescent and Starlight in that order. The much maligned Sunset actually breaks even or even makes a small profit. The Silvers to Florida also did well with the Star loosing, but the Meteor more than making up for it. I did not evaluate the Auto Train. So altogether, the 14 LD trains lost some $97 million in operating loses, but then Amtrak layers on something like $486 million in ‘overhead’ costs. Does this smell like a skunk in the woodshed to you? Is Amtrak using these 14 LD trains to shed it’s bloated overhead so as to make the NEC look better and make competitive bids to run state supported trains? That’s how it looks to me. Wouldn’t these trains be better off run by a lean mean organization that really wants to run them and make them a success. One other thing I noticed is that Amtrak’s fares for these trains are often less than the Greyhound bus fare and the bus takes longer, requires multiple transfers, offers no diner nor sightseer lounge, has cramped seating, etc. I did some tests, and just a simple increase in coach fares to at least above Greyhound resulted in a break even result for most trains. Add to that some increased capacity and the LD trains do ok. So what do you think about this?

 

[yarrow]

interesting analysis, henry. i am usually a fred frailey fan but i found the article you reference just a jumble of numbers and comments about it being cheaper for amtrak to transport sl pax by air just silly. look forward to the comments of those on the board who have made a study of the costs of ld trains

 

[Trogdor]

Where are you getting your numbers? Particularly, you need to justify your claim that the Sunset breaks even.

 

[Anderson]

Henry,

Could you give an example of a "corrected" set of numbers to look at (either for one train or in a bit more detail for the LD trains as a whole)? It's rather hard for me to evaluate your question without having your calculations out front to look at.

Also, I do have to wonder how much business "the dog" is really doing on a lot of long hauls. Filling a bus and filling a train are two far different propositions.

 

[henryj]

Henry,

Could you give an example of a "corrected" set of numbers to look at (either for one train or in a bit more detail for the LD trains as a whole)? It's rather hard for me to evaluate your question without having your calculations out front to look at.

Also, I do have to wonder how much business "the dog" is really doing on a lot of long hauls. Filling a bus and filling a train are two far different propositions.

I don't know what the 'dog' does, I just used it for comparison. Most of the LD trains appear to sell out. The Sunset numbers come from the FY2011 Performance Report and the PIP report. Revenues $12.7m. Fuel $1.9m, rent $3.1m, switching $.8(I added this), maintenance $1.9m, labor OBS $.3, labor T&E $3.2, commisary $1.4 and station costs $.1. Result, breaks even. Amtrak then loads up it's bloated overhead charge of almost $40m. I don't make this stuff up. If it's wrong send Amtrak a letter.

 

[Ryan]

Why exactly do you think that Amtrak's overhead charges are "bloated"?

How do you think that the overhead costs should be allocated?

 

[Anderson]

Hmm...you know, Henry, you seem to be looking at a variation on the "Avoidable Cost Recovery" metric that showed up in a few PIPs (for example, ACR for the Cap was 78% vs. 48% cited in the chart indicating the metrics being used to split the trains up by category. In light of this, a breakout of all LD trains by Amtrak's ACR metric would certainly be interesting.

 

[xyzzy]

Allocation of overhead costs to specific routes or even groups of routes (NEC vs LD vs non-NEC corridors) has been a continuous argument since 1971. I think of it this way. Take away the LD trains, and the overhead that remains would make the NEC and non-NEC corridors look even worse financially. Take away the NEC, and the overhead that remains would make the LD and non-NEC corridor trains look even worse. Etc.

Undoubtedly some LD routes have better cost recovery, on a direct basis, than others. It doesn't surprise me that the SIlvers look relatively good because most of the stations they serve from Savannah north are served by multiple trains.

But at the end of the day, no LD route earns enough to cover the depreciation and cost of capital to replace its equipment every 20-30 years.

 

[jis]

Why exactly do you think that Amtrak's overhead charges are "bloated"?

How do you think that the overhead costs should be allocated?

I also do not understand how without knowing what is or is not included in the overhead one can come to any conclusion about whether it is bloated or not. So I discount the term "bloated" and note that there is an overhead added the content of which we do not know.

So IMHO Henry is yet to make a cogent case supporting his contention, and that's OK. Given the task of making that case I could not come up with the numbers to make a case this way or that given that the numbers are not really available to us.

 

[henryj]

Hmm...you know, Henry, you seem to be looking at a variation on the "Avoidable Cost Recovery" metric that showed up in a few PIPs (for example, ACR for the Cap was 78% vs. 48% cited in the chart indicating the metrics being used to split the trains up by category. In light of this, a breakout of all LD trains by Amtrak's ACR metric would certainly be interesting.

Look at the CZ for comparison. They also have info for the Cardinal and Capitol. The CZ's operating loss is something like 28 or 29 million. Amtrak then adds something like $36m in overhead including insurance, sales & marketing, commissions, passenger inconvenience, police and safety and finally G&A. The biggest operating costs for the CZ are fuel, maintenance and labor, probably because it takes the longest time to make it's run. Anyway, it's an interesting exercise.

 

[Lawdude]

The first thing you need to know is that the private RR's did everything they could to get rid of LD trains prior to Amtrak's formation. If LD trains made profits, or even lost a bit but covered most of their costs, that would not have happened. Thus, you should be very careful about any argument that would imply that the railroads did not understand their own business.

The second thing is that it isn't as though Amtrak is the only LD passenger railroad in the developed world. Other rich countries also have LD trains and they lose money or require big subsidies too.

The third thing is that it would be silly to believe that the growth of interstate highways and cheap air travel would not have impacted the profitability of LD rail. Yes, it is true that there are still lots of point to point travelers on the lines, but LD lines have sleeping cars and full service diners and lounges that serve the interests of wealthier passengers traveling longer distances, and those services do not recover their costs.

So there's plenty of circumstantial evidence that long distance routes are money losers. A claim that despite all that, they really don't bears a strong burden of proof.

Long term, all arguments for Amtrak go back to the public service it provides. It is never going to make money, though of course it needs to control its losses to the best extent possible. The LD trains provide lots of point to point transportation on routes and through town with few other options. It's sort of like Essential Air Service, which doesn't make a profit either.

 

[henryj]

Why exactly do you think that Amtrak's overhead charges are "bloated"?

How do you think that the overhead costs should be allocated?

Ryan, all I can do for you is quote from Frailey's article...."For example, in transportation, engineers and conductors report to road foremen of engines, who report to assistant superintendents, who report to superintendents, who report to general superintendents, who report to general managers, who report to the VP operations, who reports to Boardman. That's eight layers, which Boardman thinks is too many."

But hey, that may be how they do it in Washington, DC.

As for how overhead costs should be allocated, I have no idea unless you want to get me access to Amtrak's accounting system. Amtrak has some 20,000 employees 85% of which are covered by labor agreements. Only around 3,000 are managers. I have no idea how they come up with 'overhead', but $486million for 14 trains seems a bit much.

 

[Ryan]

I'm not interested in Fred's opinion, I'm interested in yours.

How many layers of management should there be in between the bottom and the top? If $486 million in overhead is a bit much for Amtrak, how much should that be? How should it be allocated between departments? What are they spending money on now that should be cut?

If you're going to criticize how Amtrak allocates overhead amongst it's trains, don't you think that you should know 1)How it is currently split up and 2)What a better method would look like?

You can play accountant and cut up the numbers all you want. It's interesting information, and part of the story needed to answer the questions that I've posed to you. But it seems to me that you should have all of the information before you start making value judgements, and if you're going to say that Amtrak is doing poorly, you should have concrete ideas for how they can do better.

 

[Guest]

But what struck me most was the comment that the 15 LD One other thing I noticed is that Amtrak’s fares for these trains are often less than the Greyhound bus fare and the bus takes longer, requires multiple transfers, offers no diner nor sightseer lounge, has cramped seating, etc. I did some tests, and just a simple increase in coach fares to at least above Greyhound resulted in a break even result for most trains. Add to that some increased capacity and the LD trains do ok. So what do you think about this?

To do proper analysis, you would need to compare the fares market by market and have information as to how many people actually purchased tickets for each of those markets rather than just checking fares.

On Amtrak's LD trains, with the exception of the Auto Train (obviously) and maybe the CL, the majority of coach passengers do not travel from endpoint to endpoint, but travel between the various other stations. Perhaps the same could be true for Greyhound.

You can not assume that a "simple increase in coach fares" would not have a negative impact on ridership numbers.

 

[henryj]

I'm not interested in Fred's opinion, I'm interested in yours.

How many layers of management should there be in between the bottom and the top? If $486 million in overhead is a bit much for Amtrak, how much should that be? How should it be allocated between departments? What are they spending money on now that should be cut?

If you're going to criticize how Amtrak allocates overhead amongst it's trains, don't you think that you should know 1)How it is currently split up and 2)What a better method would look like?

You can play accountant and cut up the numbers all you want. It's interesting information, and part of the story needed to answer the questions that I've posed to you. But it seems to me that you should have all of the information before you start making value judgements, and if you're going to say that Amtrak is doing poorly, you should have concrete ideas for how they can do better.

I would like to have all that info, but unfortunately Amtrak doesn't publish it so far as I can tell. Perhaps they should. It seems to be like pulling teeth to get info from Amtrak, so much so that Congress has to mandate it. Now if some of Amtrak's 'bean counters' are also rail fans and post on here maybe they could give us some insite. When I worked for BP I had access and could get anything I wanted out of the system. But I was one of only a very few that could do that and I certainly couldn't make it public. Most of our accountants were hard pressed to just get their little portion done each month and I would imagine Amtrak's are the same. To account for all this stuff and prepare all those nice slick paper reports must take a small army anyway.

 

[lthanlon]

I just priced Amtrak vs. Greyhound for Chicago to Los Angeles on Aug. 14.

Amtrak - Texas Eagle (65 hours) / Southwest Chief (43 hours)

$195

Greyhound (around 48 hours; 2 or 3 transfers depending on route selection)

$139 advance

$185 web only

$206 standard

$229 refundable

 

[Montanan]

It would definitely be interesting to know how Amtrak allocates its overhead costs, but I'm not enough of a conspiracy theorist to think that Amtrak is deliberately shifting all of its overhead to the long-distance trains just to make them look bad. That wouldn't be good for Amtrak's corridor services in the long run, since if the long-distance trains were gone the corridors would need to pick up all of the overhead.

Back in the days before Amtrak, the freight railroads were often accused of inflating the losses incurred by passenger trains, and I think there's some evidence to indicate that actually happened. Regardless, the ICC established different formulas for tracking passenger train costs -- "fully-allocated," where the passenger trains were charged for a share of things like track maintenance; and "avoidable," which were just the above-the-rail costs of running a specific train. Which formula was more appropriate, of course, depended on the ratio of passenger to freight trains on a given line.

 

[Bob Dylan]

I think that Henry is on the right track, the numbers are just "estimates" and guesses since Amtrak seems to keep the real Books more Secret than the Presidents Cell Phone! (It's a Crackberry! :lol: )(aka known as "Creative Accounting!" <_< )

Im reminded of the old Mark Twain saying: "There's three kinds of Lies, Lies, Damn Lies and Statistics!" :wacko: IMO Bean Counters are next to Congress, Medicine Men and Evangelists when it comes to Flim Flams and Smoke and Mirrors!

 

[Trogdor]

Henry,

Could you give an example of a "corrected" set of numbers to look at (either for one train or in a bit more detail for the LD trains as a whole)? It's rather hard for me to evaluate your question without having your calculations out front to look at.

Also, I do have to wonder how much business "the dog" is really doing on a lot of long hauls. Filling a bus and filling a train are two far different propositions.

I don't know what the 'dog' does, I just used it for comparison. Most of the LD trains appear to sell out. The Sunset numbers come from the FY2011 Performance Report and the PIP report. Revenues $12.7m. Fuel $1.9m, rent $3.1m, switching $.8(I added this), maintenance $1.9m, labor OBS $.3, labor T&E $3.2, commisary $1.4 and station costs $.1. Result, breaks even. Amtrak then loads up it's bloated overhead charge of almost $40m. I don't make this stuff up. If it's wrong send Amtrak a letter.

The numbers you are citing are the incremental changes anticipated with daily service. Those are not the actual total costs of running the train, but rather the additional costs of running seven round-trips per week instead of three (and, for what it's worth, that report cites incremental revenue of $10 million, not $12.7 million).

I can't remember where I've seen this, but I remember reading some report that had a bit more detail, and showed that the total ticket revenue for the Sunset Limited only covers about 2/3 of the cost of T&E + OBS crew salaries, to say nothing of fuel, maintenance, track charges, station costs, etc. And all of that is before you even begin to get into any sort of debate over allocation of other system costs.

I'd think it would be pretty obvious that OBS costs of $0.3 million/year couldn't possibly be the full cost of running the route.

 

[afigg]

I would like to have all that info, but unfortunately Amtrak doesn't publish it so far as I can tell. Perhaps they should. It seems to be like pulling teeth to get info from Amtrak, so much so that Congress has to mandate it. Now if some of Amtrak's 'bean counters' are also rail fans and post on here maybe they could give us some insite. When I worked for BP I had access and could get anything I wanted out of the system. But I was one of only a very few that could do that and I certainly couldn't make it public. Most of our accountants were hard pressed to just get their little portion done each month and I would imagine Amtrak's are the same. To account for all this stuff and prepare all those nice slick paper reports must take a small army anyway.

There is a fair amount of information in the Amtrak monthly reports, the annual financial reports, PIP reports, the state fact sheets, and so on. If you look at the monthly reports prior to the June 2011 monthly report, you will find the route performance reports which provide revenue, total costs, allocated costs for each train service. The financial information may not be presented at the level of detail you are seeking, but there is far more publicly available than you will find for an airline or a Greyhound or a Megabus.

The bottom line is that the LD trains do lose money. The goal for Amtrak is to incrementally improve the cost recovery percentage for each LD train to reduce the losses. How to achieve that is the hard part.

One comment, in a previous post you included the passenger inconvenience line item as an overhead item. I would not consider that as overhead. Putting people up in hotels because of missed connections, placing them on a later train, reimbursing them or providing vouchers because of problems on the train trip is an operating cost for the train, not overhead. Is police and security an overhead cost or a shared cost that has to allocated to each train as maintenance is?

 

[Swadian Hardcore]

I don't know about you guys, but my personal calculation for the operating cost of the LSL each years puts it around $50,000,000-$60,000,000.

 

[Trogdor]

I don't know about you guys, but my personal calculation for the operating cost of the LSL each years puts it around $50,000,000-$60,000,000.

Based on what?

 

[Swadian Hardcore]

I don't know about you guys, but my personal calculation for the operating cost of the LSL each years puts it around $50,000,000-$60,000,000.

Based on what?

I did this a very long time ago, but I think it was based on crew costs, fuel costs, etc.. There were also many other costs but I don't remember them all.

I just have simply: LSL: $57,000,000/year.

That's why I was asking about P42DC fuel consumption back then.

 

[MARC Rider]

On Amtrak's LD trains, with the exception of the Auto Train (obviously) and maybe the CL, the majority of coach passengers do not travel from endpoint to endpoint, but travel between the various other stations. Perhaps the same could be true for Greyhound.

.

I'm sure the same is true for Greyhound, because whenever I've priced a long distance trip, you have to changes buses a lot (and at inconvenient times, too.)

If they had a lot of passsenger demand for longer trips, they'd have through buses.

 

[henryj]

The numbers you are citing are the incremental changes anticipated with daily service. Those are not the actual total costs of running the train, but rather the additional costs of running seven round-trips per week instead of three (and, for what it's worth, that report cites incremental revenue of $10 million, not $12.7 million).

I can't remember where I've seen this, but I remember reading some report that had a bit more detail, and showed that the total ticket revenue for the Sunset Limited only covers about 2/3 of the cost of T&E + OBS crew salaries, to say nothing of fuel, maintenance, track charges, station costs, etc. And all of that is before you even begin to get into any sort of debate over allocation of other system costs.

I'd think it would be pretty obvious that OBS costs of $0.3 million/year couldn't possibly be the full cost of running the route.

The FY2011 performance reports lists the Sunsets revenue as $12.7m and total costs as $52.5m. If the train is only running three times a week and you add 4 more to that then the incremental costs would be more than the current operation. I think the OBS number is just mis-stated. If I rework the numbers then the train may lose $4m a year in avoidable operating costs. Each of those PIP reports is presented differently and the Silver/Crescent one doesn't seem to even break out any costs. Thanks for the catch.