Passenger

Amtrak's key performance numbers for FY 2008 show many improvements over FY 2007.

Consolidated revenues increased 14.2 per cent.  Passenger related revenues increased 13.1 per cent, whilst commuter revenues increased 10.3 per cent, other revenues increased 12.7 per cent, and state and capital payments increased 1,258 per cent.   

Ticket revenues increased 14.7 per cent for the system, i.e. 14.6 per cent for the NEC, 17.5 per cent for State Supported and Other Short Distance Corridor trains (SS&SD), and 10.3 per cent for Long Distance trains (LD).

The number of passengers - 28.7 million - increased 11.1 per cent.  Passenger miles increased 8.95 per cent and seat miles increased 1.9 per cent.  The load factor increased from 48.9 per cent to 52.3 per cent, an increase of 3.38 percentage points.  This suggests that the system was more efficient in FY 2008.  However, the notion that it was standing room only on Amtrak for most of the year is not supported by the numbers, although there were surely heavy loads during peak travel periods, especially whilst gasoline was approximately $4 per gallon.

The NEC carried 38 per cent of Amtrak's passengers compared to 38.8 per cent in FY 2007.  The SS&SD trains carried 47.5 per cent compared to 46.4 last year, whilst the LD trains hoisted 14.5 per cent of the system passengers compared to 14.8 in FY 2007.

The NEC operating profit increased 43.4 per cent to $369 million before interest and depreciation.  The operating loss for the SS&SD trains decreased 3.6 per cent to $117.5 million, but the LD operating loss increased 9.8 per cent to $481.8 million. 

The LD trains brought in 22.5 per cent of Amtrak's FY 2008 revenues but accounted for 47.5 per cent of Total Attributed Costs before interest and depreciation.  While carrying just 14.5 per cent of Amtrak's passengers, they accounted for 209 per cent of Amtrak's operating loss before interest and depreciation.  How did they do this?  They ate all the operating profit on the NEC ($369 million) and contributed another $112.8 million to the operating loss.    This is the side of the LD train argument that its proponents, including NARP, never mention whilst reporting increases in revenues and number of riders.       

The net system loss from consolidated operations decreased from $1.12 billion in FY 2007 to $1.01 billion, a decrease of 9.6 per cent.  The increase in revenues was offset partially by significant increases in fuel, power, and utilities, casualty and other claims, depreciation, and miscellaneous items.

On a passenger per mile basis, the NEC trains contributed 20.7 cents before interest and depreciation, whilst the SS&SD trains lost 6.6 cents, and the LD trains lost 18.5 cents.  The Acela was the biggest contributor in the NEC at 34.9 cents.  The Hoosier State was the biggest SS&SD loser at 59.4 cents, whilst the Sunset Limited turned in the worst LD performance at 47.5 cents. 

A passenger traveling from Los Angeles to New Orleans on the Sunset Limited in FY 2008 received an average federal subsidy of $947.63 before interest and depreciation.  A business class seat on AirTrans would have cost approximately $539.  Amtrak could have bought the passenger a business class ticket on AirTrans and saved the taxpayers more than $408.63 per LA to NO passenger by discontinuing the train. 

Four regularly scheduled SS&SD trains covered their operating costs compared to five in FY 2007.  Not surprisingly, none of the LD trains even came close to covering their operating costs.   

End point on-time performance improved to 71.2 per cent during FY 2008.  This is up from 68.6 per cent in FY 2007.  The Acela trains were on time 84.5 per cent, down 3.3 per cent from last year.  Other NEC trains, including the Keystone service, had on-time record of 79.7 per cent.  The SS&SD trains had on-time performance of 68.6 per cent, which was a 3.1 per cent improvement over the prior year, whilst the LD trains had an on-time arrival rate of 54.2 per cent compared to 41.6 per cent.

If Amtrak was managed like a business, its executives would concentrate on the business lines that serve the greatest number of passengers with the smallest losses and the best prospects for breaking even.  These are the NEC and SS&SD trains.  They would drop the LD trains because they have no chance of ever covering their operating costs. 

Sam1, thank you very much for that summary.  I especially appreciate the observation that Amtrak could save so much money by buying Airtrans tickets for passengers going from LA to New Orleans.

However, I'm not ready to accept that your last sentence, "They would drop the LD trains because they have no chance of ever covering their operating costs," should be the end of the story.  What the summary reveals is that it is a great challenge to improve the revenue/cost ratio for LD trains.  From all that I've read over the last year in these forums and in Trains magazine it's not clear to me that we (RRs, industry, science/technology, govt at all levels) have worked the issue extensively enough to conclude that sufficient improvement is not possible at reasonable cost.  Especially, I don't see that any of the major players want to improve passenger rail.  Without the desire, or the vision, we get the business-as-usual approach, which inevitably leads to no innovation and no progress.

So Congress repeatedly funds Amtrak with its losses, but where is the funding for research on technology and business practices to reduce operating costs, to improve service, and to increase revenue?  There are literally dozens of dimensions to these areas, and almost every one of them has the potential for improvement.  And many of the potential improvements would spill over to freight railroading, trucking, electric utilities, and maybe even airlines.  Plus, global warming and other environmental issues.

  

 

I also appreciate the numbers being presented here. I shall point out that comparing Amtrak's Sunset Limited to a point-to-point, two stop aircraft service is the proverbial Apples-to-oranges comparison, in that the air service does not make over two dozen stops on each and every trip as it goes across country. The airplane makes only two stops - one at each end of the route. The numbers are unfavorable as well because of the less than once per day service of certain routes, and the once per day service of other routes. The numbers how that multiple departures per day, while costly, are the way to go towards profitability. Why does the NEC show profit ? In part because there are multiple departures along the Corridor each direction every day. Same with the State supported trains. What Sam's numbers seem to be saying is that the "profits," if there are any to be found, are in the short haul trains. And this seems to contradict the conventional wisdom that the profits are in the long hauls. I also question some of the numbers because it stands to reason that the short distance trains would have a higher capital cost related to operating cost, and it would be more expensive on a per train basis to have only short distance trains serving some geographically limited markets.

CG9602:

I also appreciate the numbers being presented here. I shall point out that comparing Amtrak's Sunset Limited to a point-to-point, two stop aircraft service is the proverbial Apples-to-oranges comparison, in that the air service does not make over two dozen stops on each and every trip as it goes across country. The airplane makes only two stops - one at each end of the route. The numbers are unfavorable as well because of the less than once per day service of certain routes, and the once per day service of other routes. The numbers how that multiple departures per day, while costly, are the way to go towards profitability. Why does the NEC show profit ? In part because there are multiple departures along the Corridor each direction every day. Same with the State supported trains. What Sam's numbers seem to be saying is that the "profits," if there are any to be found, are in the short haul trains. And this seems to contradict the conventional wisdom that the profits are in the long hauls. I also question some of the numbers because it stands to reason that the short distance trains would have a higher capital cost related to operating cost, and it would be more expensive on a per train basis to have only short distance trains serving some geographically limited markets.

 

It would be less expensive to fly people coach class between practically any paired cities along the Sunset route.  The only exception would be between one or two paired communities served by commuter air only.  I showed the business class fare between LA and NO, as an example, to demonstrate the economic dysfunction of the Sunset Limited.  Needless to say, the gap would be even wider if I had used the cost of a coach seat to make the comparison.

Last year I ran three scenarios regarding the allocation of interest, depreciation, and other charges to the long distance trains vs. the short haul trains.  The scenarios assumed that 5, 10, and 15 per cent of these items were assigned to the long distance trains.  No matter how I sliced and diced the numbers, the long haul trains cost more per passenger and seat mile than the short haul trains.  This is true even in the case of the NEC, with its large capital investment, although the cost per passenger mile is much narrower than for other corridors.  However, if one considers that the long haul trains don't pay the full cost of moving them over the hoist railroads, the cost per passenger mile widens.  

Amtrak has been trying to fix the long haul passenger train since 1971.  It has failed.  And it will continue to fail because they long ago outlived their usefulness.  They should be retired, and the resources should be used to enhance existing short haul corridors or establish new ones. 

I also welcome Samantha's examination of the NRPC's numbers, it is just that I look at those numbers, passenger loads, revenue passenger miles, and ask where expansion may be made. Others look at the numbers passenger loads, etc., and say, "Forget about trains; give up." Do the numbers that Samantha1 lists indicate any latent demand for expansions of existing service ? if so, where? Also, does it not seem suspicious that the routes with the highest cost and expense-to-passenger-mile rations are the ones that are presented as least expensive ? One would think that it would be the other way around. Would it not make more sense to lengthen certain routes in order to expand the potential market ? While I disagree with the conclusions, I welcome the work and the numbers as they add some "meat: to the discussion of the "Politics Of Passenger Railroading."

I don't think anyone around here is of the opinion "Forget about trains, give up."

The real issue is that there is one camp in the advocacy community that never met a train they didn't like and is of the opinion that there is enough intrinsic goodness in trains of all types that whatever level subsidy it takes is what we should support.  The other camp is that for whatever level of subsidy money is forthcoming politically, there are choices to make -- LD vs corridors, HSR vs 110 MPH vs 79 MPH, expanded network vs increased frequency on existing network.

One litmus test of membership in the camps is when anyone poses a choice, a recent thread by Don Oltmann of "How would you spend the Lautenberg-Lott funding if we ever get it", there were numerous responses, roughly to the effect, that if we have to make choices, rail is underfunded, and shame to those (politicians, lobbyists, taxpayers) whom we can blame for not-enough-money because we ought to fund everything -- HSR, LD trains, LD HSR was proposed and debated on a recent thread, and so on.

Another litmus test is the Sunset Limited.  One camp is that any train discontinuance is a matter of losing ground in the effort to maintain and build up Amtrak; another camp is, maybe if we give up the Sunset Limited there would be money to improve other services. 

I can see some of the reasoning behind "hang on the Sunset Limited" -- do you give up the route so it becomes impossible to establish passenger service, even on corridors within that route, in the future?  Do you really save what you think by axing that train, or would other trains connecting to it see reduced ridership?  If you gave up the Sunset Limited, would you really see that money to spend on other train services, or would it somehow get taken away or otherwise not be available for that purpose?  Is the LD network part of the social contract where people in Kansas are willing to have tax money to to the NEC, and without the LD trains, the NEC is in peril?

On the other hand, there is a faction in the community which "defends ground" and does not even want to hear about giving up the Sunset or any other train.

Do the numbers that Samantha1 lists indicate any latent demand for expansions of existing service ? if so, where? Also, does it not seem suspicious that the routes with the highest cost and expense-to-passenger-mile rations are the ones that are presented as least expensive ?

As to the thin routes being high cost, I don't get what is suspicious about that?  It stands to reason that the sparsest routes are least efficient and hence highest cost per passenger mile.  No, they don't lose that much in absolute terms because they don't serve up that many passenger miles, but they can be still high cost.  It is simple math of taking the ratio of cost over passenger miles to get cost per passenger mile. 

We can quibble about load factors, but when you start getting arline-style high load factors you get airline-style inconvenience of crowded terminals and seats and difficulty of getting bookings.

I am of the opinion that even with Amtrak's timekeeping problems and anecdotal accounts of bad service, ridership is not the problem.  If you are willing to subsidize Amtrak at the required rates to maintain auto and air-competitive fares, I think you could probably fill as many seats as you provide.

The problem is one of cost -- passenger trains are a high-cost means of providing service.  The low-cost means of providing common-carrier service is by air, and the reasons for this had been written up in Trains starting in the early 1960s.  Autos are also high-cost, but many of the costs are up-front or one-time costs of owning the car, and some of them are not accounted for by motorists, if people did that, we would perhaps see even greater usage of airlines, but people value the flexibility of autos an across cultures on this planet, people are willing to pay a lot of money for their convenience when they become wealthy enough to do it.

So if trains are expensive, what do you do?  One way is to see if trains could be operated more economically, and some of this goes under the banner of "Amtrak reform", which has had a checkered history.  Another way is to identify social benefits to trains to justify the subsidy cost, and to interest people among the broader voting public beyond the advocacy community in having trains.

The reason I make a big deal about energy efficiency and stress energy efficiency as a needed Amtrak reform is that if saving on oil is a social benefit that merits subsidy of train service, those trains better save on oil big time, otherwise there are other projects that could use the money.  Energy saving would favor corridor trains, which can be made lighter with more seats per train car.

Fundamentally, if trains require subsidy to operate, whether you have trains or not becomes a political decision, and whether they are politically viable depends on whether people out there even care about trains or have a sense that Amtrak service is providing something people value for their tax dollar.  That is why I bristle at the "shame" argument about Amtrak funding or the lack of HSR -- one is essentially scolding the very people (taxpayers in general) we need to get on our side, and I have never known scolding to be an effective tool of persuasion and making people feel good about something.

The numbers I would like to see that Amtrak never pubishes are ridership and revenue by OD pair.  That would tell us a lot about how "networky" Amtrak is and whether the small city service arguement holds much water.  But, it would be a rather large report with 500+ stations there would be likely be 10,000+ point pairs. (the theoretical max would be 500 x 500, but many would likely have zero ridership such as Toccoa GA to Florence SC)

oltmannd:

The numbers I would like to see that Amtrak never pubishes are ridership and revenue by OD pair.  That would tell us a lot about how "networky" Amtrak is and whether the small city service arguement holds much water.  But, it would be a rather large report with 500+ stations there would be likely be 10,000+ point pairs. (the theoretical max would be 500 x 500, but many would likely have zero ridership such as Toccoa GA to Florence SC)

I am in the process of asking Amtrak's financial gurus for an explanation of how they allocate interest and depreciation.  I'll throw in a request for segment information.

Sam: It appears to me that the long distance routes suffer from having to allocate a lot to station upkeep, agents, etc. The sunset route comes to mind as allocating these costs to only 6 trains a week really pushes up the costs. Also the New Orleans station also has Greyhound ( what do they pay?) Maybe NARP needs to campaign to either get various citys (like they do airports) to cover these costs or costs be allocated out of a different non operating pie. Samantha: any figures? Does AMTRAK break out these costs?. I think we all need to know the direct operating costs alone for any routes to give an unbiased look at costs. This method probably would show NEC operating profits. Also the additional operating costs due to RR caused delays should also be broken out. (again UP's delays expecially the Sunset.)

I also stated that maintaining a station for very few trains resulted in high costs per train, but Samantha said my thinking was flawed.

I also would like to see a bar graph, broken by route and sorted in stop order, of origin and destination pair passengers.

blue streak 1:

Sam: It appears to me that the long distance routes suffer from having to allocate a lot to station upkeep, agents, etc. The sunset route comes to mind as allocating these costs to only 6 trains a week really pushes up the costs. Also the New Orleans station also has Greyhound ( what do they pay?) Maybe NARP needs to campaign to either get various citys (like they do airports) to cover these costs or costs be allocated out of a different non operating pie. Samantha: any figures? Does AMTRAK break out these costs?. I think we all need to know the direct operating costs alone for any routes to give an unbiased look at costs. This method probably would show NEC operating profits. Also the additional operating costs due to RR caused delays should also be broken out. (again UP's delays expecially the Sunset.)

Only end point to end point costs for each route are made available to the public by Amtrak. Segment information is not published, but I am sure that Amtrak has it because it would be the basis of the fare structure.

The financial and operating information that Amtrak makes public is more than most companies make available to their shareholders.  I don't know whether Amtrak would make the segment information available to the general public.  It is a quasi governmental agency, so invoking the Freedom of Information Act might turn it loose.  However, having it would not change the key financial data or conclusions that can be drawn from it.  

The NEC covers its operating costs.  The other corridor operations and long distance trains do not.     

Even if segment data, e.g. revenue, direct labor, other direct costs, shared costs, and attributed costs was available to the public, it would produce more data than most people are prepared to digest.  For example, if each station from NO to LA is paired as a segment, there are 231 combinations for the Sunset route alone.  For Amtrak it would be thousands.    

As of July 2007 Amtrak owned 46 of the 525 stations that it served.  In addition, it maintained 181 other stations for which it was paid a fee by the owner.  The costs of the wholly owned station are included in other direct costs.  The costs of the shared stations are embedded in shared costs.  Amtrak does not break these costs out, although I am sure that they have them. 

Many of the stations are owned by cities or transit authorities, and Amtrak pays rent for its portion of the space occupied in them.  For example, it shares the Fort Worth Intermodal Transit Center (ITC) with Greyhound and the local transit authority.  How much rent Amtrak pays compared to the other occupants is probably available through the city of Fort Worth, which owns the property, as would be the case in New Orleans.  Interestingly, the ITC was built with tax free municipal financing, just like most of the nation's airports, which some rail advocates bewail.    

The cost of the stations is fixed.  The cost of staffing them could be variable, but if my observations are accurate, Amtrak staffs many of them so that the costs behave like they are fixed.  To its credit Amtrak has eliminated the staffing at many of the smaller stations that it serves.  For example, of the 22 stations on the Sunset route, only six are staffed. 

Clearly, if a station is only served by one train a day, not to mention six a week, all its costs get dumped on that train.  It is inherently inefficient to support a low volume activity (one train a day) with a facility that was intended for higher volumes of activity (two, three, or more trains a day).  It is one of the reasons why long distance trains will never cover their costs. 

Amtrak's trains were delayed a total of 5,610,000 minutes (93,500 hours) during FY 2008.  Although the causes are not given for the year, the numbers for September show that hoist railroads were responsible for roughly 75 per cent of the delays, whilst Amtrak and thirds parties were responsible for 18 and 7 per cent.  These numbers probably reflect those for the year. 

The cost of the delays is not shown.  Delays on the hoist railroads could actually result in a positive cash flow for Amtrak.  If a hoist railroad fails to get Amtrak's trains across its property according to the terms of the performance contract, it forfeits the on-time performance incentive payment, which is money in Amtrak's pocket.  But if Amtrak has to put passengers in a motel or on alternate modes of transport because of failed connections, it could easily eat up the saved incentive payment plus more.

My observation regarding the Raleigh Amtrak Station was as much about newspaper stories that don't drill very deep as the adequacy of the station.  The story implied that the station was chockers, but the numbers don't support the conclusion.  Building a new Intermodal facility in Raleigh may be a good option, but the story did not make a compelling argument for it.

Sam:

It's obvious by your posts that you are probably a bean counter by trade, and I have never questioned your numbers.

I do, however still stand by the "chicken or the egg" question.  From Charlotte (USAirways largest hub) to Atlanta (Delta hub) USAirways alone has 8 non stop flights each way a day.  Many more if you don't want to go nonstop.  Delta also has flights between the two cities.  Amtrak has ONE train each way, and it goes through Charlotte at 2:30 AM.  How many more people would take the train if there were morning, midday, and evening trains available?  Do you realize how many people heavily medicate before getting on an airplane?  I know several people who are afraid to fly but don't feel they have a choice.  No, I'm not one of them, I am a retired pilot.  I don't fly because I find the paranoid security at the airport offensive.  I do not, however want to go off on that tangent.  It is not the issue I am debating, other than the fact that I bet I am not the only one.

In other words, are there few trains because there are few passengers, or are there few passengers because there are few trains?

I bet if trains were faster and more frequent, many more people would use them.

To use the Crescent as an example:

There is already an adequate number of trains on the northeast corridor.  Perhaps the Crescent only needs to go to DC.  Three trains a day from ATL or CLT to DC might better serve travelers, even if their origin or destination is NYC.

That is why I said I would like to see a bar graph of origins and destinations.  Not revenue or expenses, just passengers carried.

Sam: So if I understand there is no place in AMTRAK's reports that give above the rail costs

blue streak 1:

Sam: So if I understand there is no place in AMTRAK's reports that give above the rail costs

It depends on what you mean by "above the rail costs". 

Amtrak shows Direct Labor and Other Direct Costs to determine Total Avoidable Costs.  This would be the closest number to above the rail costs.  Total Shared Costs are added to Total Avoidable Costs to get Total Attributed Costs, which subtracted from Total Revenue gives the contribution or loss for each train or train route.  This number is before the application of interest, depreciation, and unallocated system costs. 

Direct Labor includes wages, benefits, and supports costs.  These are clearly operating or above the rail costs.

Other Direct Costs include hoist railroad Maintenance of Way, performance incentives, fuel and power, car and locomotive maintenance, turn around, commissions, reservations, call centers, passenger inconvenience, and route station.  Most cost accountants would classify these as operating or above the rail costs, although you could get an argument whether commissions, reservations, call centers and route stations are operating costs.  However, the classification is approved by the company's external auditors, which carries significant weight.

Share costs include shared stations, supervision and training, yard operations, marketing and distribution, insurance, terminal payments, procurement, police, environment, safety, T&E overheads, NTS Infrastructure, and system.  Are these operating costs?  Yes, although again some accountants might differ with this conclusion.  

It appears that Amtrak has a good cost accounting system.  Equally important, they provide enough information for the public to determine whether its trains are covering their costs.

What is not clear is how the interest and depreciation is allocated.  I assign most of it (80 to 90 per cent) to the NEC, since most of it is owned by Amtrak, and it has invested heavily in it.  Nevertheless, some interest and depreciation must be assigned to other operations since the equipment is still being depreciation and, as far as I can tell, it still carries some of the debt service (interest) associated with the purchase of the equipment.   

It is crystal clear that Amtrak does not cover its costs and requires a significant contribution from federal and state governments.  This is not good from a business person's perspective.  On the other hand, for those who argue that trains provide a valuable social service that should be supported irrespective of the large per passenger mile subsidies required by Amtrak, they can bolster their case by pointing out that Amtrak's deficit as a per cent of the federal deficit is so small as to be unworthy of mention.

In the long run all costs are variable and should be attributed to the activity that drives them.  In Amtrak's case the activity is running passenger trains.  If it was a market driven business, it would be required to recover all of its costs, as well turn a profit for the shareholders if it hoped to stay in business.