Does Geography Matter? A Study of Technological Learning and Innovation in the Solar Photovoltaic Balance-of-Systems Industry

Abstract

Technologies whose future development trajectory depends significantly on local factors (e.g., administrative codes, user preferences, etc.) exhibit interactive feedback between geographical factors and technological learning. However, in such a context it is unclear which factors influence the degree of localized learning and the extent to which and at which geographic level localized learning is important for innovation. To empirically investigate firms’ learning and innovation processes in relation to geography, in this paper we use patent citation analysis on a unique database of PV balance-of-system (BOS) patents in the U.S. issued between 2000 to 2014. The PV BOS industry offers an excellent empirical opportunity to examine the linkages between geography and learning in the context of technologies in the deployment process, wherein features of such technologies are associated with the locations in which they are applied and deployed. We find that localized learning is more important for technologies that are more associated with the local context. We also find that extra-local knowledge significantly contributes to the quality of patents. Overall, our findings suggest that for technological learning and innovation in PV BOS technologies, neither local learning nor extra-local learning is uniformly superior in its importance. Thus policies to support deployment technologies should pay explicit attention to the local vs. extra-local knowledge needs of local firms. For example, local demand-pull policies may be more helpful in earlier phases of market development, while helping firms build bridges to extra-local knowledge networks may be more helpful later on.