Abstract
Studies adopting a first-order social-capital perspective have found that firms' innovation is influenced by network embeddedness in scientific collaboration. The present study adopts a second-order social-capital perspective to examine the effect of partners' centrality on exploratory innovation. We further investigate the boundaries of this effect via patent stock and structural holes (SHs), individually and jointly. Our theoretical framework is validated using negative binomial regression models and a sample of patents and joint publications by 194 Chinese pharmaceutical firms between 2007 and 2021. The results indicate that a focal firm's exploratory innovation is boosted by partners' centrality in scientific collaboration, but this effect is weakened by patent stock and SHs. Moreover, patent stock and SHs jointly influence this relationship-that is, partners' centrality has a strong positive effect on exploratory innovation when the firm holds low patent stock and occupies fewer SHs. This study offers insights into how businesses might profit from the innovation advantages of influential partners by exploring the contingencies of patent stock and SHs. Extending the framework to second-order aspects, we further illuminate the scope of the scientific collaboration-innovation issue.