INDUSTRIAL RESEARCH AND TECHNOLOGICAL INNOVATION: AN ECONOMETRIC ANALYSIS.

By EdwinMansfield '51. New York: W. W. Norton& Company, Inc., 1968. $7.50.

The study of industrial organization, a research field providing much of the rationale for antitrust policy, has in the past decade been turning away from the so-called industry study in which economic performance of a related group of firms - say cigarette producers - is traced back to the structural and behavioral characteristics of the industry. This traditional approach is being displaced partly by examinations of the "environments" of certain decision-making functions common to a host of firms and crossing over the conventional boundaries of an industry. The examinations increasingly are econometric in approach; that is, they use the tools of statistical inference and a growing body of quantitative data to test hypotheses generated by a priori economic theory. High-up among the important firm decisions receiving the critical attention of economists is that complex of decisions encompassed by research, invention, development, and innovation. Over the years Ed Mansfield, perhaps more than any other economist, has been in the vanguard of this research effort, marking the trail that many others are following.

Research in this particular area of economics can be exasperating. The foremost difficulty is the relative lack of quantitative information that will meet the technical demands of econometric analysis. Of lesser magnitude, but formidable nevertheless, is the problem of devising appropriate theoretical-statistical models by which the numerous vague hypotheses relevant to the researchinnovation process can be tested. The research-innovation process itself must be broken down into realistic subcomponents before meaningful analyses can be performed. Yet, the process is so vital to the continuing viability of an economic system that, these difficulties notwithstanding, the context, determinants, and consequences of research-innovation decisions must be comprehended. Mansfield's monograph moves us a long way toward fuller understanding; it is a brilliant paradigm of hypothesis testing and of making the best of a difficult data situation. In particular, his careful qualification of generalizations should serve as a praiseworthy example to all quantitatively oriented researchers.

A review does not provide adequate opportunity to do justice to the book as a source of methodological insights or as a guide through the pitfalls of econometric research. Suffice it that a trained, intelligent reader will find more than ample reward in these respects. Nor can a review give a full appreciation of the substantive conclusions of the monograph, although the flavor of these can be partially conveyed. The rate of technological change (innovation) is found, for example, to be directly related to the expected profit. Not a surprising result to be sure; but it is comforting, finally, to have empirical confirmation of our suppositions. Moreover, as a percentage of sales industry research and development outlays have been expanding since World War II. But has this paid off in terms of an increased rate of invention? Mansfield's study concludes that it has; however, the further inference that there are scale economies for R and D outlays must be carefully qualified. Moreover, the largest firms appear to be less inventive per dollar of R and D expenditure than firms of lesser size. There is little doubt that innovation - as distinct from invention - is beneficial to a firm's market position, since innovating firms exhibit higher growth rates than other comparable firms.

And so it goes, through rates of return from R and D, timing of innovation, diffusion of innovation, and the like. The number of hypotheses Mansfield subjects to econometric tests is impressive as are the designs of the various tests themselves. By implication, of course, this is not a book for the layman. Mansfield's The Economics ofTechnological Change (Norton: 1968) is much more intelligible to the general reader. For the specialist in R and D activities or in industrial organization generally, however, the monograph is absolutely essential reading.

Professor of Economics at Dartmouth, Mr.Clement teaches courses in Business Cyclesand International Trade, Payments, andCommercial Policy.