Gilbert B Norman wrote:Source data from the AAR
From these 2006 'stats':
1772T Ton-Miles
3163 Tons per Train
$50.3B Freight Revenue
Therefore (check my math volks) 1772T/3163=560,227,631 Train Miles
$50.3B Freight Revenue
Therefore (check my math) $50.3B/560,227,631= $8.97 revenue per train mile.
This would certainly suggest that there is a 2.13:1 or better (8.97/4.20) direct subsidy by the railroad industry to Amtrak for each train mile they operate over "freight" railroad property.
But somehow I think the indirect subsidy arising from the opportunity cost, i.e. the net "cookie jar' revenue lost, from any train a railroad could not operate because of their lawful obligation to handle the Amtrak train on a priority basis, is much higher than this arithmetical calculation would suggest.
First, some of the basis of the cost is related to maintenance, and there's no reason to think a passenger train moving at relatively high speeds does the same damage to track that slow heavy freight does.
As to opportunity cost, there are two major problems with this. First, the marginal price that would be paid for freight that isn't currently shipped is likely to be lower, not higher, than the average price. Second, there are areas where the track isn't in significant demand, and there is essentially NO marginal price that could be asked for any additional freight.
Bottom line, there's ample reason to think that even aside from the track improvement projects undertaken (which you've handily left out of your summation), that Amtrak might be pretty close to the marginal costs.
Or to put it another way, you've left so much out of the picture that we simply don't know. Certainly your analysis doesn't bear the weight of your conclusion.
The other thing that we can easily learn from other interactions of business with government is that there is substantial money to be made in weaseling. Look at major league sports, where they can almost always portray their bottom line in a way designed to weasel governments out of a little more. That makes it extremely dangerous to extrapolate from "CN says they don't want a passenger train on that track" to "CN doesn't want a passenger train on that track". Something that seems so straightforward may be utterly false.