Idea Diffusion and Property Rights (Revised November 2025)

We develop a model of industry evolution that links innovation, diffusion, and intellectual property (IP) protection. Motivated by evidence that most innovations are neither patented nor effectively protected by patents, we analyze diffusion under free imitation and derive the socially optimal licensing rate that balances learning and matching externalities. When IP protection is ineffective, innovators rely on lead time to recover investment, but extending this lead time by slowing down diffusion reduces welfare. Calibrating the model to 18 U.S. industries, including automobiles and personal computers, we quantify the theoretical predictions and show how industry dynamics shape optimal policy design.