Technological Innovation: What Do the Schumpeterian Hypotheses Imply?

Surajit Sinha


A large section of the industrial organization literature is devoted to the interpretation and empirical testing of the hypotheses advanced by Joseph Schumpeter concerning market structure and innovation. One set of hypotheses deals with the monopoly power of firms and its association with innovation, while the other set focusses on the relationship between firm size and innovation. Empirically these two sets of hypotheses may not be independent of each other, although extant empirical studies have concentrated on one in isolation of the other. In this dissertation we build a general framework where these two sets of hypotheses can be considered together in order that we can properly identify either effect -- that is, the effects of monopoly power and of firm size on innovation.

A major objective of our study is to combine two recent developments in the literature. First, it has been empirically demonstrated that proxies for appropriability and technological opportunity, when included in the regression, significantly diminish the explanatory power of concentration (a measure of monopoly power) on innovation and innovation-related activities, such as Research and Development (R&D) expenditures. In this dissertation, we construct several proxies for appropriability and technological opportunity, using new data an innovative activities of firms in Canada collected by the Economic Council of Canada.

Second, it has been recently argued that, except in the short run, both market structure and innovative activities ought to be treated as endogeneous, as opposed to the existing practice of assuming that the market structure is given. Our theoretical framework allows both innovative activity and firm size (a measure of market structure) to be determined by appropriability; technological opportunity and various non-innovative factors such as demand conditions, foreign versus domestic ownership of firms, etc.

It is commonly observed that firms sometimes choose not to hire any R&D personnel, namely, scientists and engineers, and instead entrust the responsibilities of R&D either with a consulting firm or even with its own operating personnel. These decisions result in observations with 'zero' values for R&D personnel in the sample. In this dissertation, we provide an economic underpinning for such decisions by firms, and recognize this 'limited-dependent variable' problem in our estimation through use of the Tobit estimation method.

Finally, as a by-product of our theoretical framework we examine some of the common empirical measures which have been used to test the Schumpeterian hypotheses. One of our major findings in this regard is that many of the existing empirical tests of the hypothesis that there is a positive externality from firm size to innovative activities of firms are misspecified in the sense that they are neither necessary nor sufficient for the existence of the above externality.