The Econometric Approach: A Critical Analysis

Abstract

TfHE rapid strides made by econometricians during the war years have so advanced their science that it is enjoying an increasingly important role in the formulation and testing of national economic policies. Econometrics is largely an attempt to carry out in a quantitative sense the macro-theories set forth by J. M. Keynes in his General Theory of Employment, Interest, and Money. Yet economic mathematicians appear to be experiencing some difficulty in keeping sight of their objective through the maze of mathematical and technical problems. There is a growing tendency under way, probably unconscious, to warp the economic theory to suit the exigencies of the mathematical theory. Mathematical economists should be reminded that, after all, mathematics is a tool and must be shaped to fit the task for which it is designed. economic considerations are paramount. While, for the most part, the present discussion has universal application, reference is made to a particular empirical model for the purpose of clarifying and exemplifying the points of the argument. specific model chosen is Lawrence R. Klein's Model III, as published in his article, The Use of Econometric Models as a Guide to Economic Policy.' Briefly, the Klein model covers the twenty-one-year period, I92I-4I, inclusive. Five structural equations are used. They include equations (3.i), demand for private producers' plant and equipment; (3.2), demand for inventories; (3.5), demand for new owner-occupied nonfarm residences; (3.6), demand for new rented nonfarm residences; and (3.9), demand for consumer goods. As Dr. Klein points out, the inclusion of equation (3.3), an output adjustment equation, would permit the determination of the general level of prices. He omits it because of the large estimate of the variance of the disturbance. abnormal volume of postwar gold imports, which has affected the volume of money, Klein apparently does not take into account when estimating the level of prices for I947 as an exogenous factor.2 Also, his lumping of gross exports and gross imports into a net export figure has led to distortion.3 This lumping process has concealed a large volume of trade, including investment goods, and, without doubt, is partly responsible for his faulty forecast. Exports-possibly divided into consumer and investment goods, commodity and service imports, and gold imports-should have been considered as separate exogenous items. External trade normally forms a small part of total American trade, but the current strong demand for exports is a factor which cannot be ignored. Exports, of course, are exogenous. Since many countries which are traditional suppliers of imports have been affected by the war, imports (and particularly gold imports) might well be considered exogenous also. In Klein's consumption function two weaknesses appear. First, the durable-goods consumption cycle is not taken into account. Second, currently large cash balances in the hands of wage-earning consumers are ignored. Since neither recur with regularity at short intervals in the time series, they are unlikely to show up in the size of the estimate of the variance of the disturbance. Yet they are of importance in forecasting. This