Sequential Innovation and Optimal Patent Design

We study optimal patent design in a setting with sequential innovation. Firms innovate by undertaking "research" activities to generate new ideas and by undertaking "development" activities to transform these ideas into viable products. Both innovation incentives and the welfare costs of patent monopoly are multidimensional. We characterize optimal patent policy, and in particular, the tradeoff between patent length and patent breadth in this setting. The optimal size of the patent reward is smaller for patents associated with a higher deadweight loss. For a given reward size, a better patent that generates higher social surplus is shorter but broader. The optimal patent length may be finite or infinite.