R&D: Competition, Risk, and Performance
分析预期创新后合谋对研发投资的影响，发现合谋程度低于古诺双头垄断时绩效更优，且不同市场结构下研发企业数量与最优水平存在偏差。
Anticipated postinnovation collusion encourages RD sometimes they do not. Both usually outperform perfectly collusive industries. The optimal level of collusion is often less collusive than Cournot duopoly. In Bertrand industries, too few firms do R&D. The same goes for long-shot or high-cost projects in all industries. However, in perfectly collusive industries, too many firms invest when a project has medium to high chances of success. Investment by Cournot industries is often close to optimal.