Inventor Rewards, Specialization, and Innovation Performance

Inventor rewards—monetary compensation for employees who generate patents—are widely used to promote innovation. Yet the same incentives can steer innovation toward productivity (quantity) or toward inventiveness (quality). Based on social identity theorizing, we argue that inventors’ social identity order—lower versus higher—explains this heterogeneity. Lower-order identities, exemplified by technological specialists, are tightly constructed around exclusive, proximal membership with actively policed expectations—such as strict inventiveness norms—whereas higher-order identities, exemplified by generalists, are more distal and loosely policed, providing limited social constraints on behavior. Rewards activate identity, and the desire for incentive-identity alignment then guides behavior, with stronger pressures to conform to ingroup standards under lower-order than higher-order identities. Hence, we predict that, in response to inventor rewards, specialists produce fewer but more-inventive patents whereas generalists increase productivity at the expense of inventiveness, defaulting to prototypical reward-maximizing behavior. Further, in mixed-identity teams, specialists’ norms are projected onto generalists, raising their inventiveness in response to inventor rewards. We test these predictions using a difference-in-differences estimation and matched inventor-patent data from before and after a regulatory change in China mandating inventor rewards in state-owned enterprises. Consistent with our theory, after the mandate, generalists filed more but less-inventive patents, specialists produced fewer but more-inventive patents, and generalists in mixed teams became more inventive. By linking social identity order to reward responses, we clarify how to align incentives and the R&D workforce with desired innovation outcomes.