Also there is the pay one price then order the most expensive bias.
Most sleeper passengers will zero in on the higher priced steaks, seafood, and chicken. Very few will settle for the cheaper salads, hamburgers, etc. This bias towards ordering the more desireable items raises variable costs while the fixed payment in the form of a sleeping car ticket markup might not be working in paying the piper.
On what basis do you claim that most sleeper passengers go for the higher-priced stuff? I don't have specific data, but my anecdotal observations are that people will read/listen to what the options are, and choose based on what they want to eat. The sleeper passengers that get their meals included probably don't even look at the price column.
For what it's worth, the chicken is usually among the lowest-priced items on the dinner menu, and I see plenty of folks ordering that.
For my unscientific sample size of one, I've shelled out for the steak with cash at least a dozen times, and quite possibly close to twice that. But I can also confess to being an outlier in virtually every way on this front. What can I say, I like steak and Amtrak usually does it well (though preferably not well done!). Indeed the diner prices are quite reasonable, being exempt from sales/prepared foods taxes. I think I once tallied up the cost of french toast, bacon, juice/milk, and coffee on Amtrak and at IHOP, and Amtrak came out cheaper...I forget precisely what I put in on each side, but I remember Amtrak beating IHOP on a comparable meal.
I've run into some lousy meals, yes, but I've run into those at very nice restaurants as well when the chef was having a lousy evening or a disagreement with the stove. If anything, Amtrak tends to average much better than your average "causal dining" restaurant in terms of quality (I'd put it on the lower range of "fine dining" in terms of quality as a rule, and in some ways the food is preferable to "fancy" food in my mind). The food is solid, and it is a good value for your money as a rule.
The big questions in the above are, to my mind:
1) What's the breakdown on the OBS losses? i.e. How much of that is coming off of the Acelas vs. the Regionals vs. the LD trains?
2) How are FC meals (Acela), sleeper meals (LD), and BC drinks (Regionals/corridor trains) being accounted for in the mix? If Amtrak sells a million BC seats and books off a $.25 charge for a soda for each one, that's $250,000 in "lost revenue" that could either be deducted from the BC accommodation revenue or just left on the food service budget as a $0 sale. $5/meal in costs for FC on the Acela (and I think I'm using a
very low price point) could easily rack up several million in uncompensated losses to food service. Ditto the sleepers (where you're probably averaging 2-3 meals per sleeper passenger; even at a low price point, that's close to 700,000 passengers times whatever your average price point per meal is times the average meals per passenger; assuming 3 meals at $5/meal times those passengers is going to give you close to $10m/year in costs).
3) What costs
are being allocated to food service? No, really...how much overhead is being offloaded here by Mica et al to make things look bad? Yes, I know that some of that overhead is justified...but in this particular instance, anything outside of direct costs for running the food service is just being "fudged" onto it.
4) Finally, how much revenue is there, what
are the expenses, and what does that leave as losses? My guess is that the losses, while not trivial, are still well under 50%...my best guess, based on the soda example, is that you're covering around 60% of food service costs on food service revenue...and this is potentially excluding several tens of millions in "discarded" revenue due to FC/BC/sleeper food expenses not getting paid for in cash...which would be sloppy, stupid accounting on someone's part, but I don't think we can exclude the possibility of this. This would be roughly in line with the $110m or so in revenue from food and beverage noted last year.
If the sleeper/FC/BC situation is being handled poorly, I can probably shake about $25 million/year loose in "losses" depending on meal cost assumptions; I can probably even run that higher if you bill the "losses" of the meal against the sleeper/first class/business class accommodation charges.