Abstract
The literature on price-value relations has frequently used correlation measures in testing the theory of labor value--that is, values are good proxies to prices or, alternatively, they explain, in some sense, prices. However, several authors have detected important problems that affect those measures-- particularly, a problem of spurious correlation that invalidates the statistical results typically obtained in the literature. In this paper, we show, using data from Spain (1986-94), that spurious correlation cannot be empirically treated, due to a more general problem of indeterminacy affecting any correlation measure.