ESTIMATES OF HICKSIAN IS AND LM CURVES FOR THE UNITED STATES*

Abstract

THERE ARE MORE FACTS AROUND than there are theories capable of explaining. But, while a theory should be capable of explaining a large number of facts in order to be useful, it is generally supposed that theories should not try to explain all of the facts relevant to their domain. This is because the theory, if it permits of no exceptions whatsoever, is necessarily as complex as the set of facts it is designed to explain. Current developments in the field of macroeconomic model building provide a case in point. On the one hand, there is the momental work of Professor Friedman on the influence of changes in the stock of money on the aggregate level of economic activity. It is testimony to his perseverance and the merit of his findings that many economists now protest that they never meant to imply that money didn't matter. The direct evidence that he has provided has undeniably made its mark, and the efforts of many economists are now directed toward uncovering evidence of variables other than the money supply that matter significantly. On the other hand, a magnificent chain of developments following in the heritage of Tinbergen has led to the construction of large and elaborate macroeconomic models. It is a living tribute to the ingenuity and creativity of men like Klein, Goldberger, Suits, and others, that the Brookings-SSRC model of the economy of the United States has been constructed.' But, this model has not been constructed without some misgivings, misgivings that reflect serious concern over the problems of complexity that are bound to arise whenever a model contains hundreds of equations.2 The model to be presented here, stemming from Hicks,3 could perhaps be