Concentration and co-localization dynamics of technological innovation: The Japanese case

Spatial patterns of innovation and inventive activities are a significant topic in economic geography, as many studies try to unveil the relationship between innovation and proximity. Understanding patterns of concentration and dispersion in space and the spatial relationship across technological categories becomes essential to reveal the characteristics of knowledge sources and flow. This study aims to characterize the spatial patterns of inventive activity in Japan. Using patent data, we identify concentration patterns using point data and a relative concentration ( K emp ) measure. We also assess spatial co-localization and co-dispersion by using the same indicator. These results cross-classify technologies by their concentration and co-localization characteristics and potential knowledge sources and flows, questioning the relevance of national innovation policy in concentration and co-localization patterns. Therefore, the analysis focused on two periods, 1975–1994 and 1995–2014, related to major changes in the national innovation policy. Our analysis reveals that between the two periods, Japan's inventive geography shifted from a highly integrated system to more specialized, isolated clusters. Specifically, the later period demonstrates an increase in spatial concentration within technologies while showing a decline in co-localization across different fields. The importance of this trend is evident, as regionalization policies appear to have successfully promoted local specialization, strengthening regional/local capabilities, but have also resulted in a reduction of regional technological diversity. This situation raises concerns about an increased risk of technological lock-in and the potential decline in breakthrough innovations that depend on broader knowledge flows. • We examined point scale Japanese patent data to determine spatial concentration and technological category interactions. • In the second period (1995–2014), concentration-dispersion patterns strengthen while category interactions decrease. • Dynamics of concentration and co-localization patterns are related to changes in national innovation policy.