View article

Research on corporate innovation often focuses on firms with positive US patent activity and reported R&D, thereby excluding over 90% of the firms in Compustat. By exploiting data from 30 global patent offices, we investigate the nature of missing innovation data in the US and around the world. We find systematic and predictable patterns across firms and countries for missing patents and R&D. We then compare the empirical efficacy of excluding firms without US patents or without reported R&D to simple replacement methods, and to various econometric solutions for missing patent and R&D data. We show how excluding or deleting firms without US patents or reported R&D, even in studies of just US firms, provides biased coefficient estimates and standard errors. We also demonstrate how the biases from simply excluding the missing observations lead to specific distortions in tests related to corporate growth and country level innovation capacity. We then discuss best practices or guidelines in handling missing R&D and patent data for studies in the US and elsewhere.