Knowledge that Transforms

NON-discrimination and multilateral payments represent the foundations of United States foreign trade and exchange policy. The United States has sought to implement this policy through multilateral agreements, such as the International Monetary Fund Agreement and the Charter for an International Trade Organization, and through bilateral agreements, such as the Anglo-American Financial Agreement.' Most American economists and statesmen consider the universal desirability of nondiscrimination to be a self-evident truth readily deducible from the first principles of economics. But in recent years a number of reputable economists have come forth with strong arguments in favor of trade and exchange discrimination, and these arguments have had an important influence on the formation of international economic policy. It is not enough for economists of the free-trade school to refute these arguments by an application of the principles of classical economics, since the proponents of trade discrimination usually begin by denying the classical postulates.2 Rather the problem must be examined within the institutional framework of the postwar world and its solution worked out in a manner consistent with current economic and political realities. Although the economist would be derelict if he assumed the world as it currently exists as inevitable and incapable of change in response to his counsel, mere deduction from first principles is not likely to solve its problems. The purpose of this paper therefore is to analyze the case for discrimination as objectively and sympathetically as possible, with a view to determining the extent to which the problems that the proponents of discrimination have posed can be dealt with a minimum of violence to the wellknown advantages of non-discriminatory and multilateral trade.

IN THE current literature, comparison continues to be made between the Robertsonian and the Swedish conceptions 2 of sequence analysis and of saving and investment (see, for example, the excellent volume, Survey of Contemporary Economics, edited by Howard S. Ellis). These conceptions appear to be in some manner similar, yet different; but just what is the difference seems to be a somewhat baffling question. To guide the reader, I propose to state certain propositions at the outset: I. The Robertsonian formulation may consist merely of a set of definitions, as presented in his I933 Economic Journal article on Saving and Hoarding (and it is this formulation which Keynes comments upon in the General Theory); second, his period analysis may be presented as a definite formulation of the multiplier sequence over time 3 as is done in his November, I936 Quarterly Journal of Economics review article on Keynes' General Theory. The second formulation makes use of the earlier terminology, but it does not stop with mere definitions; it suggests a definite empirical hypothesis.4 Throughout the following I shall be considering the second formulation, not simply the first. 2. The second Robertsonian formulation is not, as is commonly supposed, a mere tautology. The view that it is such overlooks the third equation referred to below. 3. The Swedish analysis based on the divergence of planned saving and planned investment, as originally interpreted, does not appear to offer a satisfactory theory of income expansion or contraction. 4. Useful analyses can, however, be made in terms of the discrepancy between certain magnitudes, such as the divergence between desired consumption and actual consumption (expenditure-lag) and the divergence between intended investment and actual investment (output-lag). These formulations are often stated in terms which bear some resemblance to, though in fact they are different from, the original Swedish formulation. The expenditure-lag is an integral part of the Robertsonian formulation. And the output-lag 5 (divergence of output from sales) has been employed by Lundberg and Metzler. Both formulations play a significant role in the explanation of the cumulative process. In each case, expansion or contraction will follow as a result of the unsatisfied conditions.