Unsurprisingly, sleeper service and dining cars on the long distance trains result in a significant increase in costs. The Silver Meteor is 38% more per seat-mile and the Silver Star 54% more per seat-mile compared to the all coach Palmetto running along most of the same route. Using this consist list and this site for car capacity figures the Palmetto has 311 seats plus diner and baggage car while the Meteor and Star possess 236 coach seats and 90/60 sleeper "seats" as well as diner, lounge, and baggage making them a good point of comparison. Similarly, I am unsurprised at the high costs of the Acela. It has a small number of seats, a high level of service, and probably some rather extravagant energy consumption costs.Yes, and those higher-cost services command a price premium.
While they aren't all near the bottom, and a couple of trains do outperform them, on average the long distance trains cost less than other trains per passenger mile. This is strictly a function of occupancy however. Cheap trains with few passengers are more expensive in that metric than are expensive trains that are well filled.The corridor services tend to be less full on average than the long distance trains (all of which function as the sole local train service for many communities). The post suggests some opportunity for Amtrak to raise fares, particularly at peak times.
There's another matter which should address itself to our attention and to the defenders of the long distance trains as they currently stand. The Palmetto, which is all coach as we know, has an average ticket yield of 20.1¢ per passenger mile. This is rather intriguing because it ties it for second alongside the Crescent for highest yielding long distance passenger train after the Auto Train despite the fact that it possesses no sleeping passengers. With an average trip length of 436 miles, these aren't short little jaunts either, but are comparable to the long distance coach trips of other trains. This strongly suggests that the coach fares of the rest of the long distance network could stand to be raised by 50-100%, gradually of course, but raised sharply nonetheless. This would also raise the price of sleepers by several cents per mile, of course, as they pay a basic rail fare, but that's hardly objectionable and would help to improve the financial performance of the train.The post notes some of the oddities of cost allocation that make analysis of a train's performance difficult. One way to reduce the overhead burden on a train is to add trains to its route. The Ohio Association of Rail Passengers have been making a case for additional trains across Ohio's northern tier (that's going to require additional passenger platforms for Norfolk Southern to buy in), and their focus is on extending the lightly-used Empire Service trains that leave Buffalo early in the morning and arrive late at night, as through passengers will fill seats currently riding empty.