Duncan, in reply to Ajit:
> No. When you say "We can definitely observe the technology with which these
> goods and services and their raw materials are produced", I'm pointing out
> that the only method available to us right now to do this is input/output
> matrices, which are based on market price value transactions, not on direct
> measures of physical inputs.
The tables are in price terms, but when you calculate the
technical coefficients the prices "cancel out". That is,
you'll get the same answer for the a_ij's regardless of the
price vector. Paul C and I showed this in our paper
presented at the IWGVT session earlier this year --
available at http://www.wfu.edu/~cottrell.
Allin Cottrell.