Do TBED Policies Help or Hinder Knowledge Sharing?

A central tenet in the understanding of regional economic clusters is the idea that the closer two actors are to one another, the more likely they are to collaborate. This belief is based on decades of research done to examine knowledge spillovers and the effect of spatial proximity on tacit knowledge sharing. In a recent article, however, Jasjit Singh of INSEAD and Matt Marx of MIT differentiate the varying effects of crude distance on knowledge sharing compared to the effects of geopolitical borders. Despite the explosion of the Internet and advancement in communication, borders still matter.

Singh and Marx examined 30 years of patent citations to analyze the effects of geopolitical boundaries on knowledge sharing. Although patent citations do not capture the true impact of knowledge sharing, they can provide a meaningful understanding of the path and scope in which knowledge flows, and the potential benefits it may produce. The authors compiled over four million patent citations of more than 630,000 patents between the years 1975-2005 from the U.S. Patent and Trademark Office (USPTO) into their data set. Patents where inventor teams were geographically dispersed were omitted from the database since their geographical knowledge source seems ambiguous. Citations from "nonfirm" sources, such as universities or government bodies, or those unassigned to a specific organization also were excluded from the data set to focus on interfirm diffusion of knowledge.

The authors then used a regression framework to determine the likelihood of a citation between two random patents. Firms' patent citations are 1.3 times as likely to be from a patent within the patent's origin country, more than two times as likely to spread across the same U.S. state, and almost three times as likely to flow within the same metropolitan area. Same-country localization of knowledge is to be expected because of linguistic, cultural and institutional similarity; however, the authors of this study found that the intensity of this localization has increased over time, despite the move towards globalization, while the likelihood of intra-state localization has decreased.

Of the 943 core based statistical areas (CBSA) that were studied, 63 contained more than one state, such as Cincinnati, OH, where the metropolitan area also contains Kentucky and Indiana. Even in these areas, patents are still more likely to be cited in their state of origin, providing further evidence of a strong state border effect on knowledge flows. The authors suggest further exploration of government support for research, higher ed and other related policies might provide a better understanding of this phenomenon. Much of technology-based economic development (TBED) and general economic development policy is state-centric, and therefore we are left to question how this affects innovation and to consider whether TBED policies contribute to this behavior in beneficial and/or disadvantageous ways. Read the paper...