---In STMFC@..., <gn2059@...> wrote :
I'll take a contrary position: If the number of cars in service wasn't important to know, why was there a column for it in the ORER? It would seem to me the GN tried to be accurate, as the number of cars in service changes for each number series in the various issues of the ORER I have examined.
I am also late to the party here. The arguments already presented that the actual number of cars makes no difference, since if the car didn't exist, no one would be looking it up, make some sense, as far as they go.
However, that ignores the times when a shipper complains, "why can't you send me more cars like the last one, instead of this junk." It becomes an easy reference to find the answer, "Because we only own six of them, and they are scattered all over the country." I'm sure there were other instances of agents being asked to find cars with a specific dimension or other characteristic, and having some idea how large the potential target population was would be useful.
Anyway, It is my understanding that updates to the ICC valuation reports were still required at least until 1960, so someone was already counting the cars for this purpose. It would be trivial to mine this data to provide mostly accurate numbers for remaining cars in each series for inclusion in the ORER.