Industry Concentration and Allocation of Resources to Basic Research

While the mainstream of Western economic thought believes in competitive markets, the Schumpeterian school considers concentrated industry to be the ideal vehicle for the advancement of industrial technology. Empirical evidence on the Schumpeterian hypothesis that a concentrated industry allocates relatively more resources to its research and development (R&D) activities has been inconclusive. We believe the reason is that the Schumpeterian hypothesis is too general and vague. The fact is that there are several dimensions of industrial concentration, for example, total number of firms in an industry, interfirm differences on a variety of characteristics such as sales, assets, resources, and so forth. Similarly, R&D activities are of several types such as long-term or short-term research, and basic research or applied research and development. Unfortunately, modeling efforts focusing on resource allocation by specific types of concentrated industries to specific types of R&D activities are almost nonexistent, with the exception of Joglekar and Hamburg (Joglekar, P., M. Hamburg. 1986. A homogeneous industry model of resource allocation to basic research and its policy implications. Management Sci. (February) 225--236; Joglekar, P., M. Hamburg. 1987. Industry resource allocation to basic research under normally distributed benefits. Decision Sci. (Winter) 1--24.) who showed that industries consisting of smaller numbers of firms allocate relatively greater amounts of resources to their basic research than industries consisting of larger numbers of firms. Extending Joglekar and Hamburg's models here we find that when two industries consist of the same number of firms, in most cases, a heterogeneous industry falls as short of its Pareto optimal investment in basic research as its comparable homogeneous counterpart. Numerical examples suggest that in some cases, concentrated (heterogeneous) industries may do worse. Policy implications of the results as well as directions for further research are discussed.