Managing a Partnership Efficiently∗

The focus of this paper is on how control rights should be allocated to ensure a successful partnership in a dynamic environment. Control rights can be viewed as an equity share for joint-ventures and R&D consortia, as patent breadth for cumulative innovation, or as the probability that a firm wins a patent race. Several partners control a multistage project. Each partner has private cost of completing each stage. I describe the set of dynamic contracts that ensures efficient completion of the project in the absence of commitment. I find that in most cases, control rights must be dynamic for the duration of the project. For any stage, the optimal control right allocation depends on the abilities of the partners in the current stage and the sensitivity of future performance to current cost. In an optimal contract, higher expected efficiency of a partner increases his control rights, while higher sensi-tivity of future expected performance to his current cost diminishes his control right. In the absence of a technology spillover, partners invest at the socially optimal level. ∗I am grateful to Tracy Lewis, James Anton and Curtis Taylor for their guidance and advice. I have benefited greatly from the comments and suggestions of Dan Graham, Huseyin Yildirim and Sergei Severinov