Knowledge that Transforms

In writing this paper I benefited greatly from information provided by visiting scholars (at Berkeley) from the Soviet Union, Hungary, Czechoslovakia, Poland, and Yugoslavia. To them, and to the following friends and colleagues, I am deeply indebted not only for stimulating discussions and research cooperation but also for observations based on personal experience-particularly on issues of fundamental disagreement: Carlo M. Cipolla, Gerard Debreu, Howard S. Ellis, Oldrich Kyn, Abba P. Lerner, Mark Perlman, Richard Roehl, and Benjamin Ward.

This working paper was prepared by the author while serving in the Research Department, Office of the Assistant Secretary for International Affairs, U.S. Treasury, Washington, D.C., on leave from Simon Fraser University. The views expressed in this paper are those of the author. The U.S. Treasury Department does not necessarily agree or disagree with these views. The author acknowledges gratefully the comments of P. Clark, J. Makin, P. Kenen, M. Keran, and W. Schmidt made on an earlier draft of this paper.

IN THIS PAPER the reader will find a review of econometric studies of investment in fixed capital. A review of these studies through 1953 was given in 1957 by J. Meyer and E. Kuh [86], and a detailed review through 1960 was presented by R. Eisner and R. H. Strotz in 1963 [36]. In this review we concentrate on recent research on time series of investment expenditures for individual firms and industries. Our point of departure is the flexible accelerator model of investment originated by H. B. Chenery [13, 1952] and L. M. Koyck [74, 1954]. In this model attention is focused on the time structure of the investment process. The desired level of capital is determined by longrun considerations. Changes in desired capital are transformed into actual investment expenditures by a geometric distributed lag fuinction-the specification of desired capital has been the subject of a wide variety of alternative theories; the alternative theories do agree on the validity of the flexible accelerator mechanism. Denoting the actual level of capital by K and the desired level by K+, capital is adjusted toward its desired level by a constant proportion of the difference between desired and actual capital,

The philosophy adopted and the literature coverage demonstrated owe mnuch to the assistance of: Thomas W. Calmus, Graduate School of Management and Business, University of Oregon; Karl Haiiser, Seminar fur Volkswirtschaftslehre, Universitdt Frankfurt/Al; J. C. L. Hulqkamp, International Bureau of Fiscal Documentation; and T. A. Kennedy, National Economic Development Office of the U.K. Errors of fact and interpretation are, of course, entirely of my doing.