Does geography matter for science-based firms? Epistemic communities and the geography of research and patenting in biotechnology

The spatial clustering of innovation has been associated with localized knowledge flows among small, knowledge-intensive firms, but knowledge flows may extend beyond regional boundaries through the participation of firm employees in broader knowledge-based communities. This paper analyzes biotechnology firms jointly engaged in technological innovation and open scientific research and proposes that the geography of their collaborations should reflect the distinctive social logics of these activities. I hypothesize that projects involving local ties are more likely to be patented by a firm than are projects involving distant contacts, both because proximity is conducive to innovation and because a small firm's social capital is likely to be greatest in its home region. However, classic studies in the sociology of science show that scientific communities are socially stratified and geographically dispersed. As a result, I hypothesize that ties to distant partners and prestige in scientific communities are positively associated with scientific impact but negatively associated with firm patenting. The analysis focuses on 5,143 collaborative research papers published by a large sample of small biotechnology firms. The average distance among coauthors on a paper is some 1,500 miles, indicating that the firms are engaged in geographically far-flung research networks; however, the distribution of teams in space is strongly bimodal, revealing an important core of regional ties alongside a set of much more distant ties. Regression analysis show that the spatially clustered teams are more likely to publish papers that are subsequently cited in the authoring firms' patents, whereas teams that are globally dispersed produce papers that are more highly cited in the scientific literature, but less cited in the authoring firms' patents. Status in the scientific community has the expected positive effects on paper citations, but a negative effect on patent citations. The results give evidence of different-and in some respects conflicting-logics governing the creation of new technologies on the one hand and valuable ideas in science on the other, highlighting challenges faced by firms that aim to profit from knowledge created in open scientific communities.