Innovation quality and global collaborations: Insights from Japan

An employee of online fashion retailer Zozo Inc, poses with Zozoglass and its smartphone application, used to measure skin tone for ordering cosmetics online, during a demonstration at the company's office in Tokyo, Japan, January 28, 2021. Picture taken January 28, 2021. REUTERS/Kim Kyung-Hoon


Even if the post-COVID-19 recovery seems already underway—despite that millions are still hurting due to the loss of loved ones, record-high unemployment levels, and despair and uncertainty that might last until after a vaccine is effectively distributed—some of the secular trends that preceded COVID-19 will likely still accompany us, namely, the global productivity slowdown.

The productivity dynamics of Japan, in particular, are quite intriguing. Most developed nations have seen sharp slowdowns in their productivity growth since 2004, but the Japanese case is puzzling because productivity growth decelerated just as sharply there, even though productivity levels continue to lag well behind countries such as the U.S. and Germany (see Bahar and Strauss, 2020; Baily, Bosworth, and Doshi, 2020)

In light of this, Bahar and Strauss (2020) suggest a possible explanation for this phenomenon, backed by the evidence at hand: an important dichotomy between the quantity and quality of Japanese innovation. In particular, they find that while Japan allocates more resources on research and development (R&D) and files more patent applications than the U.S. and Germany, the quality of Japanese innovation severely lags compared to these two nations.

This paper digs deeper into this finding to identify possible policy actions that could tackle this reality. Specifically, we identify an aspect of Japanese innovation—compared to that of countries like the U.S. and Germany—that could play a role in explaining the lower-quality inventions hindering the creative-destruction process. This aspect is the extent to which Japanese inventors engage in Global Collaborative Patents (GCPs). GCPs are patents where the team of inventors includes individuals residing in two or more countries. As Kerr and Kerr (2018) show, GCPs tend to be higher-quality innovations than entirely domestic patents.

In this context, our study uses patent-level data and document two important stylized facts. First, Japanese inventors tend to engage much less in GCPs as compared to other comparable countries; and second, even if fewer, GCPs by Japanese inventors tend to be of higher measurable quality, consistent with the evidence. Having established these facts, we then find that Japanese firms dominate the production of patents in Japan, with very little innovative activity by foreign firms, where most of the GCPs typically would occur. Thus, based on this, we offer some thoughts about possible directions using policy to facilitate the integration of Japanese inventors into the global stage by (1) pursuing the establishment of foreign R&D centers in Japan and (2) opening the country’s doors to more immigration.

The paper is divided as follows. The first section describes the data sources used in our analysis. The following section characterizes Japanese innovation using patent-level data, such as the quantity and quality of innovation, its geographic distribution within Japan, and the most active firms (also referred to as assignees) patenting in the country. After that, we document the relatively low intensity with which Japanese inventors participate in GCPs and establish the fact that Japanese GCPs tend to be of much higher quality than patents filed by all-local inventors. Finally, we present some broad policy guidelines to promote the participation of Japanese inventors in global collaborations before offering concluding remarks.

Download the full working paper