Global knowledge flow and Japanese multinational firms’ offshore R&D allocation and innovation

Abstract

This paper examines whether allocating more research and development (R&D) activities to a country-industry pair with a higher intensity of knowledge flows improves the innovation performance of multinational enterprises (MNEs). We use firm-patent-matched data for Japanese manufacturing MNEs, including data on MNEs’ offshore R&D expenditure and information on patents filed by both parent firms and overseas affiliates. Moreover, as a proxy for the intensity of knowledge flows, we use the eigenvector centrality of each country-industry pair in the global knowledge flow network, utilizing patent citation information.

We find that the quality-adjusted number of patent applications tends to be higher for MNEs that allocate more R&D activities to country-industry pairs that are more central in the network of global knowledge flows. However, we did not find any significant relationship between the country and industry distribution of offshore R&D and the number of patent applications.