The indeterminacy of price-value correlations: a comment on papers by Simo Mohun and Anwar Shaikh

This paper is the first published critique of the indeterminacy of price-value correlations and their inadequacy as empirical evidence for the determination of prices by values. It comments on the approach developed by Shaikh, Petrovic, Parys, Ochoa and others, according to which prices, as asserted by Ricardo, are empirically ‘97%’ determined by values. This method calculates measures of distance (according to some or other measure such as Mean Absolute Distance, or correlation) between a vector of empirically-observed average price of a set of industrial sectors, and a vector of aggregate values calculated as the vertically-integrated labour coefficients of the same set of industrial sectors. The present paper suggests, and establishes using a Monte Carlo method, that the observed correlations are most likely to ‘spurious’ since they can be entirely accounted for by variations in the size of the industrial sectors concerned. The paper was published in the same volume as the paper from Anwar Shaikh to which it responds, as well as another by Simon Mohun on which the paper also comments. (Bellofiore, R (ed) Marxian Economics: a Reappraisal, Volume 2, pp139-162. Basingstoke: McMillan) The controversy was subsequently developed in a number of exchanges including, in particular, papers in the Cambridge Journal of Economics between Andrew Kliman, Paul Cockshott and Allin Cottrell. It has further been discussed in papers by Ruben Osuna and Emilio Diaz which are at the time of submission unpublished, and in papers by Tsoulfidis and Maniatis also in the Cambridge Journal of Economics.