Fred Frailey Blog

Amtrak's $100-million trains

  • Comments 39

I'm old enough to not be frightened by big numbers. But I have to admit catching my breath when I saw what it costs Amtrak to operate three of its most popular and endearing trains. We're talking nine digits.

The National Railroad Passenger Corp., dependent upon taxpayer largess, lives a perilous, hand-to-mouth existence. Perhaps that's why long-term changes in ridership, revenues, and yes, deficits, get so little attention. Why worry about yesterday, in other words, when there may be no tomorrow? 

In this and following blogs, I'm going to examine these trends, starting today with the long-distance trains. Subsequent pieces will look at short-distance trains and the Northeast Corridor. 

So which are those $100 million trains? Leading the parade is the Chicago-Seattle/Portland Empire Builder, whose fully allocated costs during 2010 came to a startling $123 million. If you're having trouble imagining how one pair a trains can consume so much cash, so am I. What's helpful to know is the actual operating cost. Amtrak no longer publishes the direct costs of running individual trains. But it did through 2009 (all years are fiscal, ending Sept. 30), and then the Builder bore crew costs of $69 million, plus another $23 million paid to host railroads and for fuel, locomotive and car maintenance, reservations, stations, and the like. Finally, the train's share of corporate overhead and other shared costs came to $18 million. At least, this is how I interpret the rather byzantine footnotes to the 2009 financial report.

So balance the Empire Builder's $123 million in 2010 costs against the $62 million in revenue it generated. Passengers paid for 51 percent of what it took to keep this train going, in other words. At that, it did better than the typical long distance train, whose revenue constituted just 46 percent of its costs, a percentage that's unchanged from 2003.

The other $100 million trains are the Southwest Chief (Chicago-Los Angeles), $107 million, and California Zephyr (Chicago-Oakland), $103 million. In the past eight years, the cost of operating the Builder rose 38 percent, but that of the Zephyr just 14 percent. The Chief saw its costs fall from $123 million in 2003, but then so did revenue, from $69 million in 2003 to $62 million in 2010.

Some other ways of looking at operating results:

Ridership. The biggest gains from 2003 through 2010 came from the Capitol Limited (42 percent), Lake Shore Limited (37 percent), Texas Eagle (34 percent), Empire Builder (28 percent), City of New Orleans (26 percent), Southwest Chief (25 percent), and Auto Train (22 percent). Please don't ask me how the Southwest Chief can handle 25 percent more people on falling revenue and expenses. At the other end come the Silver Star/Silver Meteor from New York to Miami (up 3 percent), Coast Starlight from Los Angeles to Seattle (no change), and Sunset Limited from Los Angeles to New Orleans (down 12 percent).

Losses.  I wish this came under the heading of "profits and losses," but no long-distance train comes close to closing the gap between revenues and costs. Well, I take that back. The Auto Train recovered 74 percent of its costs in 2010, thanks those 15-17 Superliner cars it trundles daily between Northern Virginia and Central Florida, followed by dozens of auto carriers that require no care and feeding en route. The Empire Builder, Palmetto, Capitol Limited, Silver Meteor, California Zephyr, and Texas Eagle managed cost recovery of 51 to 46 percent, in descending order. Worst of all are two tri-weekly trains, the Sunset Limited (22 percent) and New York-Chicago Cardinal (30 percent).  

Finally, the I'm Speechless Award goes to the Silver Meteor and Silver Star. Together, their losses in 2010 came to almost $100 million. 

By now you're probably napping between paragraphs. So let me leave you with some top-line and bottom-line numbers for the long-distance fleet, for both 2003 and 2010. 

Route Ridership change 2003-2010 Revenue change 2003-2010 Fully Allocated Loss 2003 Fully Allocated Loss 2010
Auto Train 22% 26% $12 mil. $21 mil
California Zephyr 17% 13% $49 mil. $56 mil
Capitol Limited 42% 15% $21 mil. $22 mil
Cardinal 20% 47% $11 mil. $17 mil
City of New Orleans 26% 33% $19 mil. $23 mil
Coast Starlight 0% 21% $36 mil. $50 mil
Crescent 17% 17% $38 mil. $44 mil
Empire Builder 28% 28% $44 mil. $61 mil
Lake Shore Limited 37% 15% $37 mil. $38 mil
Silver Star/Meteor 3% 27% $57 mil. $92 mil
Southwest Chief 25% -19% $69 mil. $62 mil
Sunset Limited -12% -28% $30 mil. $38 mil
Texas Eagle 34% 26% $29 mil. $29 mil
  • Since I think the future of rail travel is in the 100 to 300 mile corridors let's just can the Cardinal. Replace it with two daily round trip Chicago-Indianapolis-Cincinnati trains. Getting to these major cities in the day instead of the middle of the night should increase ridership.

  • Fred:

    Do the 2003 numbers include headend traffic and how correct do you think David Gunn was to eliminate it?   The fact that all the losses increased despite material revenue increases makes me wonder.

    Bill McDonald

  • I would like to know the source of these figures.

  • Go to, click on Media at the bottom of the home page, then Reports & Documents and scroll down to Monthly Performance Reports. Select September 2010 for fiscal year totals. I got reports for prior fiscal years using the Wayback Machine, an Internet archive.  September 2004 which also had fiscal 2003 numbers is as far back as I could go. Took most of a day to do all this and massage the numbers in Excel.

    Fred Frailey

  • Yes, 2003 numbers did include mail, which explains why the Southwest Chief's revenue went down noticeably. So did its loss! So Gunn was probably right that it was not making the company any money. Those 30 car SW Chief trains sure were a hoot to look at, though.

    Fred Frailey

  • As a European who has travelled many times on Amtrak long distance I am not surprised at the disappointing financial returns of these trains. On most long distance routes Amtrak does not provide a reliable or frequent or expeditious service for any point to point journey along the route. They cannot do this because the tracks are controlled by freight railroads running  their giant lumbering slow trains on tracks  that has been pared to the bone.

    Congress has condemned Amtrak to providing a minimal passenger railway on life support since 1971 and the result is the worst of all worlds. The promise of significant investment in conventional passenger rail now appears to have evaporated at the first sign of opposition from the hard core Republicans. This really is a tenth rate nation.

  • Well sir, you got me into research mode. In  July/August 1974 the one way coach fare between Chicago and Los Angeles was $113.50; corrected to 2011 dollars this would be $514.47 Today's fare is one-half to one-third the inflation corrected fare. No wonder their losses are growing.

  • This is fascinating and there is no common thread - yield up on some eg Coast Starlight,Cardinal,Silver Service, steady on EB and Crescent and down on Capitol Ltd and LSL.

    I guess the Coast Srlight had a poor 2010 as it was annulled north of SF for over a month and the New Orleans trains will have been affected by Katrina and its aftermath.

    The most promising results are perhaps on Capitol and LSL and are of course the focus of the initial investment in extra sleepers and new diners.

    There were action plans published last year for Capitol (bascally put Pennsylvannian onto it west of Pittsburgh), Cardinal (run daily E of Indiannapolis, Texas Eagle - run to LA daily - they are all on the FRA site at Amtrak.

    We all know that UP would not play with the daily TE/SL without more infrastructure but where do the others stand which broadly improved rolling stock utilisation and therefore productivity to varying degrees.

    I guess that the Florida trains are but shadows of their 1990's up to 20 vehicle length but presume that some of the new sleepers will be added to these formations.

    Recently read Fred's book on the Last of the Great trains which showed that different RR's in the 1960's used different costing methods to justify perpetuation of - but ultimately exit from - passenger services. It is amazing that one organisation Amtrak manages to get so many different answers on revenue,volume,yield and cost trends in my view.

  • It's cheaper to have 8 people take a Limo to New York City from Upstate New York  - and we get free champaign.

  • Just going by ridership numbers doesn't tell us enough.  A hundred passengers taking the Empire Builder from Winona to Red Wing counts just as much as 100 going from Milwaukee to Grand Forks or from St. Paul to Seattle.  A very high ridership count might easily result in less revenue if most of the passengers are just going short distances.  That said however, there are clearly serious problems revenue wise.

  • Some trains are always sold out. The coast starlight has only two trainsets and runs just once a day when this route could handle easily 10 trains a day. Not everyone needs to ride the Zephyr from Emeryville to Chicago. But a lot might ride from Sacramento to Reno, Sacramento to Denver, Salt Lake to Denver, etc. Again only two trainsets and no reserves. They could easily sell out 5 trains each way every day. The strategy has been to starve the passenger trains since there is no room for them on the freight railroads' trackage. Thus poor service, inadequate service and antiquated equipment. The repubs would like to remove Amtrak altogether.

  • I would like to know how much the federal government subsides air and highway transportation systems. I would guess that taxes cover only about 10% of highway construction. I wouldn't want to even guess how much the cost of moving a passenger by air is paid by tax dollars. That should include airport construction and everything else to and including the cost of the air traffic control system. To be fair, include deprecation of the entire interstate highway system and all the above in the air system.

  • Someone needs to be able to take all of Amtrak's raw data and produce a clear accurate cost breakdown.  I am curious about all of the corporate overhead and other expenses that get loaded onto the LD trains.  Many people over the years (such as Andrew Selden) have argued the Amtrak has "hidden" all kinds of NE Corridor costs onto trains all over the system to hide their true cost and make the NE Corridor look good.

  • Champaign I would expect to be free...  Champagne I will pay for...  The sparkling wine served complimentary on the Coast Starlight 10.03.2010 was unquaffable, the Pinot Grigio served me the following afternoon was "corked..."  The Dining Car staff thought I was crazy when I refused the bottle, and refused to pay for it, though I ordered and paid for another...  It seems they did not know the term, nor could they verify may concern because of "Rule G" did not allow them...  They (the staff) could only top my hamburger with processed American Cheese, even after I requested Gorgonzola -- and even though my wife was served a pear and Gorgonzola salad...  First Class $$$ we paid, one train a day -- we as taxpayers should not have to pay for more of the same...

  • This does seem to be a manifestation of the 40 year question, what does Amtrak want to be when it grows up? After checking coach fares between numerous city pairs today’s fare is one-half or less that that of 35 years ago when corrected for inflation. (Remarkably, sleeper fares are about the same.) The result of this has, in effect, rendered Amtrak’s trains (long distance in particular) into Greyhound busses on rails. Should the American taxpayer subsidize a low cost method of travel? A plausible argument could be made for such. Should the American taxpayer subsidize a first class method of travel? Does Amtrak really provide a first class service? Is the provision of a bed enough to be classified as “first class?” If so I’ll try to remember that the next time Motel 6 leaves the light on for me.

    Ultimately there are a lot of unanswered questions. Make no mistake the questions have been asked but the answers were either incomplete or implausible. The face of transportation is changing so it shall be interesting to see how Amtrak changes (or is changed) as a result.

Amtrak's $100-million trains