Spurious Value-Price Correlations: Some Additional Evidence And Arguments

During the last two decades, many Marxian economists have brought forth empirical evidence that supposedly supports a version of the “labor theory of value” that Marx rejected, namely the theory that individual commodities’ prices tend to equal their values. However, recent studies have challenged this conclusion. The present paper offers additional evidence and arguments against it. Firstly, the theory in question implies that prices will be higher, ceteris paribus, in industries in which variable capital is a relatively large component of total cost, but regression analysis of U.S. data compels us to reject this hypothesis. Secondly, although sectoral values and prices are very strongly correlated, simulation results indicate that the observed correlations are no higher than the correlations that can be obtained by aggregation, even if the disaggregated values and prices are uncorrelated and extremely far apart. Finally, many studies have found that average price-value deviations are small, but it is shown here that this finding is meaningless, since aggregation of the data tends systematically to reduce measures of average deviation.