On Wed, 20 Mar 1996, Iwao Kitamura wrote:
> I think your testing the proportionality of price/value is
> valid for your purpose in the paper. I imagine the slight
> upward disparity of the slopes and the nagative constants
> in those four cases come from exclusion of depreciation
> which may be relatively larger in large mechanic industries.
The regressions are logarithmic (following Shaikh's methodology),
and hence the estimates of the intercepts are biased; one needn't
pay too much attention to the negative constants.
Allin Cottrell
Department of Economics
Wake Forest University