So, why are we seeing program upgrades for Washington Union Station in the $7 billion range which only serves one city, and not seeing any realistic proposals to replace and upgrade the sleeping car fleet of the long distance system, which generates huge amounts of cash? Amtrak continues to have its priorities wrong, focusing on the care and comfort of the Northeast Corridor traveler, and mostly ignoring the unloved, but cash-generating long distance train travelerLet's start with the question of whether Amtrak's sleeper passengers generate "huge amounts of cash." In Fiscal Year 2011, Amtrak's long distance sleepers generated $171,319,021 in ticket revenue, 35% of the total of $481,262,202 in ticket revenue that long distance trains generated. While that's a significant fraction of total long distance revenue, and a higher per passenger fare than any other Amtrak service, it's not really what should be considered a "huge amount of cash."It is, after all, less than the $204.9 million operating surplus that the Acela posted (excluding OPEBs and other costs).
However, that's just revenue and ignores the significant costs incurred with sleeper service and with long distance trains in general. Now, I am a critic of Amtrak's overstaffing: I see absolutely no reason why the Surfliner, for instance, should run with a crew of four to five. But that's half the crew size of a typical long distance train (page 92). In addition to the sleeper attendant, whose costs must be apportioned amongst those in his car, a significant fraction of the food service costs must be attributed to sleeper passengers, who receive food complimentary as part of their fare. That alone should dispel the idea of a sleeper passenger being a cash-generator since one should also include the opportunity costs of those free meals. In addition to this, there is the significant problem of late trains resulting in ordering busses and overnight hotel accommodations for connecting passengers.
What of the idea that Amtrak passes costs from the Northeast Corridor onto the long distance trains? That may very well have happened decades ago, but without any hard evidence, it's difficult to say that that currently is the case. Certainly it seems an odd thing to claim when Amtrak's costs are higher for their NEC trains per train-mile than for the long distance trains. It also doesn't pass the smell test: There simply aren't terribly many costs which could be plausibly passed on to long distance trains and what costs there are are, or should be, fairly insignificant next to the total costs of the long distance trains. Take for instance, the cost of maintaining the NEC, which is the most cited item I've seen for costs passed on to long distance trains inappropriately. At $150-200,000 per double tracked route-mile for high speed rail, a worst case scenario for Amtrak puts the cost of maintaining the 1,555 track-miles of NEC and Harrisburg Line rail at $116-155 million. With long distance trains expensed at over a billion dollars in Fiscal Year 2011, the idea that Amtrak is unfairly burdening the long distance trains seems outlandish.
Lastly, there is the question of opportunity costs. Every item of equipment which is currently operated in long distance service could also be potentially used in corridor service, even the diner equipment, though sleepers would likely need to be remodeled into a coach or business class car. Certain of the peak Surfliners, for instance, are projected for over a thousand passengers on their travel; it stands to reason that a dining car would be of far more use there than on a long distance train which serves only a few hundred passengers across its thousands of miles. But new or enhanced corridor service (such as turning the Cardinal/Hoosier state into an 8x daily service between Chicago and Indianapolis) is another area where equipment could be put to more productive use and this is fairly crucial when taking into account asset depreciation, something Amtrak currently does not display on a per route basis but which will hit long distance trains heavily.