R&D Data Sharing in New Product Development.

The article investigates when firms in data-driven industries voluntarily share proprietary data and how such voluntary sharing compares with government-mandated (centralized) data sharing aimed at promoting innovation. Using a game-theoretic innovation contest model, it identifies two key factors influencing data-sharing decisions: (i) the relationship among firms' data sets—whether they are complementary or redundant—and (ii) the degree of uncertainty in the innovation's market and technology. Firms tend to voluntarily share data when their data sets are complementary or when uncertainty is high, but refrain from sharing when data sets are redundant and uncertainty is moderate. Compared to the centralized setting, firms undershare data when data sets are complementary and uncertainty is moderate, while they overshare when data sets are redundant and uncertainty is high, highlighting potential misalignments between private incentives and social welfare. The findings have implications for industries such as autonomous vehicles and cybersecurity, and suggest that government interventions like mandatory data sharing and subsidies should be carefully designed to balance innovation incentives and competition.