Undoubtedly, the difficulties that beset the literature on the Schum peterian arguments are in large part a reflection of the complexity and difficulty of the subject matter. But part of the problem has been that economists have not tried to confront those complexities with a model designed for that task.
We have presented an abstract model of Schumpeterian competi tion that is designed to help economists think through some of the processes involved, sharpen their intuitions regarding the relation ships between technical progress and market structure, and see more clearly the public policy issues that are latent in those relations. Some of the implications of our model are consonant with less formal asser tions about the characteristics of Schumpeterian competition. There are advantages to a firm in being large. In particular, the larger a firm, the greater the ability to appropriate returns from its own successful R&D efforts. Also, the larger R&D spending of a larger firm tends to provide a smoother, more reliable advance of productivity- and thus a lower vulnerability to declines brought on by temporary innovative dearth.
Our model helps sharpen certain propositions that when pre sented informally tend to be rather fuzzy. In particular, its workings are such that market structure per se matters. If the rest of the in dustry is small in total size or consists largely of very small firms, this enhances the extent to which a firm of a given size can appropriate the returns from its own innovation. Also, market structure and behavior can shelter or lead to the decline of firms that spend on in-novative R&D in circumstances in which a strategy of imitating is more profitable.
Some policy conundrums that may well be presented by economic reality are illustrated by the model. It leads us to contemplate the possibility that not only may a relatively concentrated industry pro vide a better shelter for R&D than a more fragm ented industry struc ture, but that production and technical advance may also be more efficient in such a setting. Thus, some tradeoffs are revealed, but not necessarily the same ones that have been identified previously. In particular, in an industry in which technology is science-based, greater concentration “trades off” higher markups for a smaller gap between average and best practice and for more efficient R&D (in the sense that a given productivity level is tracked more cheaply), but does not buy a faster rate of growth of productivity. Further, the tradeoff may be different in some industries than in others; in our experiments with an industry whose technology is cumulative, a more sheltered competitive environment, with its associated higher markups, does lead to more rapid productivity growth.
The results that show a tendency for firms that do innovative R&D to lose o ut in a competitive struggle with skillful and aggressive imi tators are particularly provocative, and illustrate a possibility not much discussed in the economic literature. Nor has there been much discussion differentiating the kinds of regimes for technical progress under which the social costs are slight (science-based industries) or heavy (cumulative technology industries) when firms that invest in innovative R&D are driven to the wall or out of business.
There are some interesting predicted empirical relationships that derive from the model, but, as with the policy conundrums, they are somewhat more subtle than those that many economists seem to think reside in Schumpeterian competition. Most of the reported em pirical “tests” have rested on the supposition that the Schumpeterian arguments imply that large firms tend to spend relatively more on R&D than do small firms . Under the stylized conditions of our exper iments, s uch a correlation can arise only as a result of selection- that is, of differential growth of innovators relative to imitators. In exper imental settings in which innovative R&D is profitable, the firms that spend on innovative R&D (and hence that have a higher ratio of total R&D to capital) do tend to grow in relation to the imitators, but in s uch a setting the small firms tend to be eliminated. Where innova· tive R&D is not profitable but where market structure permits it to s urvive, the R&D-intensive firms tend to be small.
On the basis of o ur model, one can predict that industries with rapid technical progress ought to be marked by high average R&D intensity and, as the industry matures, by a more concentrated in-dustry structure than industries in which technical progress is slower. And, interestingly enough, various studies attempting to ex plain cross-industry differences in productivity growth rates have identified roughly the above relationship. The model also suggests that concentration is likely to increase over time in a technologically progressive industry- another relationship that seems to hold em pirically.
Some other interesting relationships are suggested that present intriguing (though difficult) targets for empirical research. For ex ample, the relationship between R&D expenditure and rate of pro ductivity growth in an industry may depend on the character of tech nical advance in the industry-in particular, on whether there is an exogenous flow of new innovative opportunities and on whether technical change in the industry is cumulative in the sense that today’s advances build on yesterday’S. In actuality, these possibili ties are not likely to be mutually exclusive, but there may be some situations that approach one or the other of the pure cases. It would be interesting to try to classify industries by regimes of technological change and to test whether there are differences between the regimes in the relations connecting technical progress with internal R&D expenditure. Similarly, it would be interesting to attempt to measure “technological opportunity” and thus to explore directly the question of whether industries in which opportunities are rapidly expanding tend to generate both high innovative R&D expenditure and a con centrated industry structure (as contrasted with inferring this indirectly from regressions relating industry R&D intensity and concen tration to rates of measured technical change) . Does the survivability of innovative R&D depend on the difficulty of imitation? Again, an answer to the question depends on being able to measure the ease of imitation.
It should be clear that Schumpeter’s appraisal of progressive capi talism continues to present an imposing challenge to theorists, econ ometricians, policy analysts, and scholars of technological change.4 We hope that this analysis will prompt researchers to view the Schumpeteri an arguments in a new light.
Source: Nelson Richard R., Winter Sidney G. (1985), An Evolutionary Theory of Economic Change, Belknap Press: An Imprint of Harvard University Press.