Taking Stock

This chapter has reviewed a number of conceptual and modeling issues related to the economics of research and development. I introduced the distinction between process and product innovations, macro and micro innovations, and also discussed the concept of the innovation possibilities set and the importance of the nonrivalry of ideas.

We have also seen why ex post monopoly power is important to create incentives for research spending, how incentives to undertake innovations differ between competitive firms and monopolies and how these compare to the social value of innovation. In this context, I emphasized the importance of the appropriability effect, which implies that the private value of innovation often falls short of the social value of innovation, because even with ex post monopoly power an innovating firm will not be able to appropriate the entire consumer surplus created by a better product or a cheaper process. Arrow’s replacement effect, in this context, implies that unless they have a cost advantage, incumbent monopolists will have weaker incentives for innovation than the entrants. Despite the appropriability and the replacement effects, the amount of innovation in equilibrium can be excessive, because of another, countervailing force, the business stealing effect, which encourages firms to undertake innovations in order to become the new monopolist and take over (“steal”) the monopoly rents. Therefore, whether there is too little or too much innovation in equilibrium depends on the market structure and the parameters of the model.

This chapter has also introduced the Dixit-Stiglitz-Spence (or for short the Dixit-Stiglitz) model, which will play an important role in the analysis of the next few chapters. This model offers a very tractable formalization of Chamberlin’s approach to monopolistic competition, where each firm has some monopoly power, but free entry ensures that all firms (or the mar­ginal entrants) make zero profits. The Dixit-Stiglitz model is particularly tractable because the markup charged by monopolists is independent of the number of competing firms. This makes it an ideal model to study endogenous growth, because it will enable innovation to remain profitable even when the number of products (or the number of machines) increases continuously.

12.7.