The Theory of International Values

Abstract

p OFESSOR GRAHAM' S recent book' issues a powerful challenge to neoclassical economists to make fundamental changes in their assumptions and methods of thinking on the theory of international values. Such challenges inevitably provoke, at first, a feeling of distrust of-almost of antagonism toward-the new proposals; annoyance at expressed or implied criticisms of current views and methods and of those who hold them; and annoyance at the apparent misuse of terms. If it appears, however, that the old terms are misused consistently and that in their new uses they are uniformly related in ways different from the old, a second level is reached; a translation must be attempted and the new set of concepts compared with the old. Then, at a third level, if the new set of concepts looks promising, it must be tested and its limitations learned. Finally, if the model is found to give useful results, there is, on a fourth level, a rush to use the new methods and integrate them with the old. So it is with Graham's challenge. I have sought to force a path to the third level and to point out some things that should be done there; but the way has been difficult, and I may have gone astray. No challenging book, of course, can be adequately summarized; but, for the purposes of this article, a short initial survey is necessary. It will not be convenient, however, to follow the author's arrangement of material. Tastes, Graham holds, are shifting, subjective, and unpredictable; they provide no basis for a theory of normal values: that is, for a theory which purports to describe the enduring tendencies which pervade and shape the economic order. A theory of normal values must be a cost theory. In the field of domestic trade the theory of normal values is based on costs; and the most appropriate theory of normal domestic values is an opportunity-cost theory.2 For, under constant-cost conditions, which ordinarily prevail, at least approximately,3 normal marginal utilities are determined by relative costs; and costs are therefore the ultimate determinants of normal values. Under variable-cost conditions, relative marginal costs are at least a trustworthy guide to normal values. In the field of international values, however, classical and neoclassical economists have held that values are determined by reciprocal demand; such a theory is not a theory of normal values but a theory of market values-and a very poor one at that. In the theory of international trade, classical and neoclassical writers were led astray, Graham's argument continues, by generalizing from the case of two countries and two commodities. Even in this simple case the probability is over-