United Rail Passenger Alliance

Volume 3, Number 25

1. Amtrak has been naughty, again. The Chicago Sun-Times reported on May 29th the Hoosier State, train no. 318 from Chicago to Indianapolis had a gross passenger service failure.As the Sun-Times reported, the train’s engineer allegedly ran a red signal around 9 P.M., near Dolton, Indiana, a far suburb of Chicago. This would have put the train about 30-40 miles from its originating terminal in downtown Chicago. Running a red signal on a train is one of the worst and most serious safety infractions that can occur on any railroad, and is dealt with harshly by all company and regulatory parties concerned.

   When this occurred, the train went to the stopping point, and the engineer was immediately removed from the train for a mandatory drug test. While all of this was going on, the 70 paying passengers on the train sat from 9 P.M. until 3 A.M. when rescue motorcoaches finally arrived to transport the stranded passengers to their destinations, including Indianapolis, which was still more than 150 miles away. Amtrak police also did arrive at the stranded train, and supplied small snacks to the stranded passengers. The train is a coach-only operation; there is no food service car.

   This happened on a Sunday night, which, in reality, should be irrelevant, especially still in the outer area of a major metropolitan city such as Chicago. Why did it take so long to rescue the passengers? Why was there not a contingency plan available and someone to make immediate decisions that benefited the passengers first, and Amtrak, second? Certainly in the Chicago crew base, one of Amtrak’s largest, there was a rested engineer which could take the train to its endpoint?

   This type of scenario makes it tougher and tougher to recommend Amtrak travel for those who may have any type of chronic medical condition, or need to be at a destination with reasonable certainty. While Amtrak apologists will say this type of thing can’t be anticipated, that’s just not true. Good contingency planned accounts for any scenario. It’s all part of good customer/passenger service. Amtrak is improving, but it still has miles to go. The sooner problems like this are dealt with efficiently, the sooner Amtrak will become a viable company. Until then, Amtrak remains a problem child for almost everyone who has contact with the company.

2. It’s summer, Amtrak’s busiest season. URPA’s Vice President of Law and Policy, Andrew Selden, has composed a snapshot of Amtrak’s fleet availability.
   

   Amtrak Fleet Snapshot

   Amtrak has squandered so much money subsidizing the operations and bloated infrastructure of the Northeast Corridor, and the bloated costs of its own corporate overheads (it still has more executives and administrators on staff than engineers and conductors combined), that in the first week of June, rolling into the peak travel period of the year, the one chance it really has to make some decent revenue, the car fleet was stretched to the breaking point.

   Out of an active fleet of 1,345 cars of all kinds, 1,138 cars were available to cover a minimum daily need of 1,075. So far, so good. But these numbers hide some serious issues.

   First, consists have been truncated, on many long distance trains. The Empire Builder, Coast Starlight, and California Zephyr are all running with one less Superliner sleeping car than in years past; Florida service has been cut almost in half from what it was six years ago. In the NEC, 20 new Acela trainsets have displaced as many as 150 “surplus” Amfleet 1 cars.

   In the NEC, 340 Amfleet cars were available to cover 326 daily assignments. That is: two extra trainsets. The only thing that makes this tolerable in the NEC is load factors are so low and frequencies so high that if a trainset has to be taken out of service, passengers can be bumped to the next train, and if traffic surges on a given day or hour, hundreds of seats are available that Amtrak can’t give away on most days. In the Chicago-hub corridors, 47 Horizon cars covered 49 assignments, two cars short at the outset. In the far west, 69 “California Cars” covered 63 assignments, and 43 “Surfliner” cars covered 43 assignments. There is no margin for error here. In the Pacific Northwest, 61 Talgo cars covered a need for 64 cars.

   Across the system, a total of 207 active cars (15.4%) were out of service for a variety of maintenance related reasons, and hundreds more rot away at Beech Grove and elsewhere, out of service.

   In the long-haul markets, where summer service is critical, only 131 Superliner coaches were available for 123 car lines, and 87 sleepers for 74 assignments. Since a minimum of five to seven cars is required to add one car of a given type to each trip, these margins are razor thin. Adding one sleeper to the Starlight and Zephyr would leave a fleet sleeper reserve (for unexpected mechanical failure or wreck replacement) near zero.

   Power is critically short. Only 69 locomotives covered 68 daily assignments in the NEC, and trains have been annulled due to shortages of engines; 163 units covered 150 assignments in intercity, but reliability issues pervade the fleet.

   Bottom line is this: the reason Amtrak can’t show any meaningful growth (hint: airlines are swamped with traffic this summer, their strongest performance in five years) is that it has squandered billions of free federal capital dollars propping up the great myth of the NEC (itself losing more money than it ever has), rather than invest a small fraction of that freely available annual capital into its most revenue-productive capital assets, the Superliner fleet.

3. Andrew Selden also tackles the issue of corporate numbers.
   

   Numbers

   by Andrew C. Selden

   If you want to understand Amtrak’s financial results, the October 2005 monthly report is instructive. In that one month, the company lost (in very round numbers) about $100 million. Of that, about $65 million was real cash (excluding depreciation of $43 million) funded by federal operating grant money. The cash losses were allocated, by Amtrak, as follows (per “Federal Railroad Administration Defined Train Cost”):

   NEC:	$17,000
   State-supported:	$486,000
   Other Short Distance:	$961,000
   Long Distance:	$18,500,000

   What rational person, seeing those numbers, wouldn’t conclude there was a real serious issue with the long distance trains as a group, and maybe get out a pocket calculator and start dividing that number by “ridership” to get those bogus “loss-per passenger” numbers that pols, the National Association of Railroad Passengers (NARP), and the media like to flagrantly toss around?

   Here’s where the big lie comes in: the astute number-crunchers may have noted 17+486+961+18,500 does not sum to $65 million, the reported loss. It sums to just shy of $20 million. Where is the remainder? The missing $45 million – a figure, in CASH, that Uncle Sam had to cover that one month that is more than TWICE the aggregate cost of ALL of Amtrak’s train operations?

   Well, the monthly report has that money, in its own line, right below the segmented FRA train costs: $45 million, described as (and I am not making this up): “General Operating Funding.”

   Now, since the $18.5 million assigned to the long distance trains times 12 months comes out to $222 million, we can safely conclude the $18.5 assigned to the long distance group above is pretty fully-allocated, such that none, or very, very little, of the missing $45 million, can be allocated to them as “overhead” (and if we use Amtrak’s grossed-up but unsupported number for long distance losses of $300 million/year, the difference is about $6.5 million a month, which would still leave $38.5 million of the “General Operating Fund” in October to be charged against something other than long distance trains).

   The pretty clear message from Amtrak’s own report therefore is something they do on an ongoing basis (called “general operating”) costs twice as much as running trains; the mystery activity does NOT involve the long distance trains as a group; and, the mystery activity is a steadily recurring item. Since the long distance trains are all paid for outside of the mystery activity, what is left?

   We can reasonably surmise the “State-supported” trains costs are at least fully-allocated (knowing what we do about how Amtrak overcharges California and Illinois to run their trains), and the “Other Short Distance” category is peanuts. What is left? Headquarters costs? Which of Amtrak’s activity centers requires a very large headquarters and other-than-train-operations staff? Could it – just maybe – be that “successful” Northeast Corridor, with all of its infrastructure, monster stations, and those 11,000 employees? Is it even remotely possible that to get a true picture of the financial results of Amtrak’s operations of its “successful” NEC segment we should add to the reported $17,000 loss on NEC train operations all of the net $38.5 million in “general operations” costs, and discover that the long distance trains as a group “lost $18.5 million” but the NEC as a group lost nearly $40 million, more than twice as much as the long distance group?

   Is it possible if the NEC infrastructure were owned by a different entity, and Amtrak shed those costs but had to pay rent to use the NEC its costs for “FRA Defined Train Operations” in the NEC might leapupwards by a large fraction of the $38.5 million (or more)? Unless it sharply curtailed the number and velocity of its trains that use that property? And even if it cut those costs in half they would still exceed all of the long distance trains as a group?

   And, we still haven’t accounted for the “noncash” depreciation (and since 90+% of the capital is in the NEC, that’s where 90+% of the depreciation charges are incurred), which would add another $43 million a month (90% of which would be another $38 million) to the losses of the NEC, for a total monthly result, using real-world business accounting (the kind you go to jail for not using if you are the CFO of a publicly-traded company like Enron), of a monthly NEC loss of around $75 million. Just in October. Using the most favorable and conservative assumptions.

   NOW do we understand why NARP and the Northeastern pols and media dread having Floyd Hall and Enrique Sosa, two accomplished CEOs of very large public companies, on the Amtrak Board overseeing Amtrak’s planning and investment strategy? And why Amtrak, David Gunn and NARP are so bitterly opposed to breaking out the ownership, and costs, of the NEC infrastructure?

4. Amtrak’s experiment with operating one long distance train the way long distance trains are supposed to be run continues to provide interesting results. FY 2005 route performance numbers for the Empire Builder show solid results and solid improvement in a number of areas.Ridership (the headcount of all passengers who used the train) was up 9% to 476,500, or about 653 passengers per trip, on average. Output (total transportation volume, measured in revenue passenger miles) rose 6.35% to 359.3 million, on 1.892 million train-miles. Revenue rose 7.67% to $42.1 million (by far the highest for any single train Amtrak operates). Passenger miles per train mile (the average number of passengers on board at any one point on the route) rose 5.82% to 189.9 (higher than Wondertrain Acela). September ’05 pm/tm reached 228.1, up 13.8%. Load factor was about 57%, nearly 20% higher than Acela, and almost twice the average load factor in the entire NEC.

   Tourism promotion in Montana played a notable part in Empire Builder results. Boardings at East Glacier, Montana rose 18% to 11,943, and at Whitefish (gateway to the Big Mountain ski resort) rose 10%, to 62,719. Belton (West Glacier) was up 25%, and the Builder’s only flag stop, Essex (the Izaak Walton Inn) rose 5.5% to 3,947. Montana’s total boardings rose 9% to 142,783.

   The Empire Builder is the last long distance train (except Auto Train) to feature a traditional cooked-onboard full meal dining service. Sources inside Amtrak report financial margins on Empire Builder dining service have improved in the last year, with enhanced service and higher fares.

   Operations of the train have been subject to the usual glitches. On time performance in the last 12 months has been at an unacceptable 50-60% due to winter weather and equipment problems.

   Some interesting demographics about the Builder:

   • The Empire Builder is Amtrak’s most geographically-isolated train, and (except at its endpoints) serves the least populated route in the system, but consistently produces the most transportation output and the most ticket revenue of any single train in the U.S.
   • The Builder cost 40% less to operate than NEC trains (per pass/mile) and 27% less federal subsidy (per pass/mile).
   • The Builder’s route by itself (without considering off-line connections) includes 47 stations with more than 1,000 city-pair combinations. One train serves all of these overlapping needs for travel, just like a transcontinental Interstate highway.
   • The top volume city-pairs: Chicago – Minneapolis, Chicago – Seattle, and Chicago – Portland; Chicago – Seattle and Chicago – Portland together only account for 10% of the riders, but about 25% (more than $10 million) of revenue. The intermediate small town stations account for more than two-thirds of passengers, and revenue.
   • The average trip on the Builder is more than 800 miles, a 15-hour journey. One in three Empire Builder passengers rides more than 1,000 miles; these people produce almost $30 million in revenue.
   • None of this data accounts for inter-route connections (possible only at Chicago or Portland, and to a very limited extent at Seattle).
   • Passengers making short trips (under 500 miles) account for half of all riders but just 20% of revenues.
   • Empire Builder transportation output (about 360 million pass/mile) dwarfs such trains as the Coast Starlight (206), all of the Pacific Surfliners (204), or all of the San Joaquins (118).
5. Traffic on Wondertrain Acela was so soft in March Amtrak resorted to a “buy two/get one free” ticket promotion in March. Acela load factors remain very weak, at about 50%. U.S. airline load factors are currently near 90% overall.Last year, Amtrak settled its Acela procurement dispute with Bombardier by releasing Bombardier from its warranty/maintenance contract on the trouble-plagued Acela trains. In March, Amtrak turned to Acela co-builder Alstom for an $80 million, five year contract to deliver overhaul parts kits, and scheduled maintenance parts and supplies. Amtrak didn’t say whether it would amortize these costs and charge them off against Acela revenues.