Herbert asked me a few questions about the program; I'll answer them here.
I'm pretty sure that it runs on DOS (IBM/clones).
There's nothing wrong with calculating FCCs numerically, I do it all the time.
I use a brute force [flux(param + d) - flux(param - d)]*param/(flux*d), and
make sure that the result is not changed by an order of magnitude decrease in
d. When I check this against matrix methods (with the elasticities calculated
in the same way :-) ) the results are identical. I mentioned this in case
there was a preferance for matrix methods.
OK, I'll contact Dr. Jackson in the near future to ask about the program & if
we can obtain some copies. Also documentation/refs.
Paul
