On vertical integration into R&D. (research and development)

This article examines incentives of a firm to integrate vertically into R&D. Vertical integration into R&D has the unique characteristic that the research outcome is lumpy and uncertain (that is, there are disproportionately higher rewards to finishing first in the innovation race). Integration into research enables the firm in our model to appropriate gains from innovation that might not otherwise be available. The setting of this model can be visualized as follows. There is some probability of some innovation occurring in the future. The innovation outcome modeled is a one-shot, winner-take-all, nonzero sum game. In other words, multiple innovations and research spillovers are not considered. The problem facing a firm is whether or not to enter the innovation race by setting up its own research facilities. If the firm innovates before anyone else does, it receives all the rewards from innovation. However, failure to innovate first results in the loss of all research investment. Firms are able to benefit from the current best technology until someone innovates. Examples of such integration into one-shot research include pharmaceutical firms deciding whether to pursue development of a new drug. Research into every new drug involves investment in specialized technology and personnel in order to be the first to capture the reward of innovation. A case in point may be the different firms currently pursuing a cure for the AIDS virus.

Frequently cited incentives to integrate vertically include better substitution among inputs, price discrimination by an integrated firm against rival firms, and guarantee of a steady flow of inputs to the integrated firm in case of vulnerable input markets (see, for example, |1, 7, 8~ for details). However, vertical integration into research has not yet been formally studied in the economics literature (see |4~ for a review). It is important to understand private incentives for …