Science, social networks and spillovers

Previous empirical research has established that science appears to stimulate the widespread diffusion of knowledge. The exact mechanism through which science catalyzes knowledge flow, however, remains somewhat ambiguous. This paper investigates whether the observed knowledge diffusion associated with science-based innovation genuinely stems from the norm of openness and incentives for publication, or whether it arises as an artifact of scientists having more dispersed social networks that facilitate the dissemination of tacit knowledge. Our findings support the former possibility: We use patent citation patterns to track knowledge flows, and find that science-based innovations diffuse more widely even after controlling for the underlying social networks of researchers as measured using data on prior collaborations.