Business Groups and Knowledge Flows: Analysis Using Acquisition Events

This study conducts empirical analysis of the impact of corporate acquisitions and subsidiarity on knowledge flows within business groups, utilizing the Ministry of Economy, Trade and Industry's Basic Survey of Japanese Business Structure and Activities combined with patent data.

In Japan, subsidiaries are responsible for approximately 30% of research and development investment, making the understanding of knowledge flows within business groups a crucial issue for both corporate management and innovation policy. However, previous studies have not sufficiently analyzed the distinction between selection effects of corporate choices and the actual treatment effects of group formation.

This research constructed panel data by linking the Ministry of Economy, Trade and Industry's Basic Survey of Japanese Business Structure and Activities with the Institute of Intellectual Property (IIP) Patent Database (IIP-DB), measuring knowledge flows through patent citation relationships between companies. The study detailed tracked citation relationships of patents filed between target companies, quantifying how much knowledge flows from which company to which company and at what speed.

According to the analysis results, for subsidiaries, approximately 20% of prior art sources were from other companies within the group, including parent companies, which was at a comparably high level to the frequency of the company's own prior art. Additionally, approximately 20% of the utilization destinations for subsidiary inventions as prior art were other companies within the group, confirming that the share of intra-group companies is relatively high.

Regarding the impact of acquisitions on knowledge flows, a selection effect (0.213) existed for knowledge flows from parent companies even before the parent-subsidiary relationship was established, and a significant additional treatment effect of 0.164 was also recognized after the actual parent-subsidiary relationship formation. Furthermore, a significant treatment effect (0.155) was found for knowledge flows from subsidiaries to parent companies, demonstrating that bidirectional knowledge flows are promoted.

Analysis of knowledge flow speed was also conducted, showing that self-citations have citation lags approximately 21% shorter than other citation relationships. For knowledge flows from parent to subsidiary companies, speed improvements of 7% selection effect and 1% treatment effect were confirmed, while for knowledge flows from subsidiary to parent companies, speed improvements of approximately 4% selection effect and 10% treatment effect were observed.

The article concludes by empirically demonstrating that subsidiarity through acquisition has the effect of both increasing the level of knowledge flows and accelerating their speed, with this effect manifesting similarly in partial ownership as in complete subsidiarity.