Title: Incentives for Collective Innovation

Abstract: Identical agents exert hidden effort to produce randomly-sized improvements in a technology they share. Their payoff flow grows as the technology develops, but so does the opportunity cost of effort, due to a resource trade-off between using and improving the technology. In the unique strongly symmetric equilibrium, small innovations may hurt all agents as they severely reduce effort. Allowing each agent to discard the innovations she produces (after observing their size) increases equilibrium effort and welfare. If agents can instead conceal innovations for a period of time, under linear payoffs, there exists an equilibrium in which improvements are refined in secret until they are sufficiently large, and progress stops after disclosure. Although concealment is inefficient due to forgone benefits and the risk of redundancy, under natural conditions, this equilibrium induces higher welfare than the one with forced disclosure.