I can answer that.
Around or slightly in excess of $2,000 per car is not a bad number over a longer period, without refurbishment. That is running repairs.
Maintenance cost, as an economic activity, depends a lot on the usage of the cars, the volume, and the distance that's being travelled.
The $4,300-per-car number that has been used by FRCC is based on an estimate of what the car cost would have been in 1992. Mr. Atamanenko, 1992 was a 35-million-tonne crop year. That is the highest that we can find on record. If we adjusted the revenue cap in that year to make it equal, we would have had, instead of the $700-million revenue cap for railroads in 2004, a $1.1-billion revenue cap calculated by the government.
The reality is that you have adjust the maintenance cost to the actual volume and the distance being travelled. The CTA, when they did that estimate and debated it publicly, came down to $3,600 as an estimate that would be volume-adjusted. We don't actually agree with this number either, but the reality is we're not talking about $1,600 versus $4,300; we're talking about something much smaller than that, perhaps a lot less than half of that--but we're in a highly technical debate trying to reconstruct costs that 15 years ago would have been 5% of the overall cost structure of a railroad.
The reality is we've moved away from that. Today, we are getting a revenue. That revenue is 35% less than what U.S. railroads receive for the same basic activity just a hundred miles south of the border. The revenue is fair to the farmers, and it's just fair to the railroad.
It is highly misleading to single out the maintenance cost, which is 6% or 7% of the overall cost, and to construct 15-year mathematics to try to get to a number, and it is not very relevant to the long-term issue of what we do to make sure there is enough revenue for the railroads to invest in the cars for the long term and avoid taxpayers' having to spend $1.5 billion to replace the cars.