FIRMS CHOICE OF RESEARCH-AND-DEVELOPMENT INTENSITY IN THE PRESENCE OF AGGREGATE INCREASING RETURNS TO SCALE

When firms possess unique R&D assets such as innovative ideas or bright researchers, and there are aggregate increasing returns to scale in R&D, then there can be several Nash equilibria involving different levels of investment in R&D. In the absence of common knowledge it is not clear whether players will choose strategies that lead to Nash equilibria let alone Pareto-preferred ones. It is shown, in an experimental representation of this situation, that in fact the Pareto-preferred equilibrium may not be reached. This result implies that R&D policy may play a role not just in correcting the traditional market failures but also in solving a coordination problem among firms that are reluctant to research unless other firms research. © 1990.