Neroden has done some pretty good analysis of the incremental cost and revenue. However, that does not account for the common costs that are necessary to run any trains at all.
The two sides of that coin are:
1. If you kill a train those costs do not go away. They just get redistributed to the other trains thus making all of them look less attractive.
2. If you completely remove that common cost then you are unable to run any train effectively, so the entire edifice collapses.
The trick, and a pretty hard one at that, is to figure out how to minimize the common cost without losing the business on the one hand, and trying to get the incremental revenue minus the incremental cost of each (or at least the most number of) trains positive. On step to take is to try to account for actual costs for each train at as fine a per cost item granularity as possible, based on real numbers traceable unequivocally to the operation of the train using a capable inventory and work resource tracking system.
Taking a rolled up number and then trying to distribute it based on some arbitrary formula is at the root of a lot of the evil. In some cases that is all you can do, but at present many believe that due to the weakness of inventory and resource accounting systems used, Amtrak does too much of it, much more than for example airlines do - something that Anderson may actually be able to fix. And trust me, past experience suggests that if that is done effectively many traditional railroaders will be terribly upset about it, just like many traditional airline folks were upset about it when they were hit by such. A lot of sins and favorite oxen hide in those caves.
At least one major complications arises in this exercise when one has to take into consideration second and third order effects and cross elasticities to take into account the network effects. A number of trains together produce more than just the sum of trains, and how much more depends a lot on the quality of design of the network.
So even if someone can gussy up a number to satisfy some Senator’s demands, I am not sure that at the end of the day it will provide the right metric for doing anything. It really is more complicated than that, and that is the reason each time someone comes up with one of these simple minded numbers and tries to justify some action that does not quite pass the "taste test" I cringe. The greater danger in my mind is to be told "I know and this is it" while skating on thin ice, and basing decisions on such.
Edited by jis, 16 June 2018 - 07:37 AM.