Academic spill-ins or spill-outs? Examining knowledge spillovers of university patents

Abstract In this article, we investigate whether academic technology-based knowledge crosses university boundaries or remains trapped inside the ivory tower. To do so, we rely on a matched sample approach to compare the spillovers generated by academic and firm patents using measures that take into account knowledge spilling-in and knowledge spilling-out of academia. Although it is true that knowledge exchanges among universities may inflate the overall spillovers of university patents vis-à-vis firm patents, our results indicate that university patents generate more spillovers than a comparable sample of matched corporate patents, even when knowledge flows among universities are not regarded as spillovers. This suggests that, in our sample, firm technologies more frequently rely on academic patents than on technologies from other corporations. In addition, we find that the gap between university and industry spillovers differs across industries, with industries where patents are important for appropriating returns from R&D (i.e. more economically valuable), such as drugs, presenting a smaller gap than in industries with complex technologies (where firms have strong incentives to patent aggressively), such as computers. Finally, we show that industry patents generate more spillovers locally and that academic knowledge spillovers are less geographically localized than those of corporate research.