Multinational companies (MNCs) have increasingly been trying to make use of globally dispersed knowledge. Several research groups, including geographers, have examined the resulting globalization of R&D, but most seem to have ignored the interaction between international and domestic R&D. It is reasonable to consider these two activities separately provided that they are not significantly interrelated. However, Japanese MNCs typically concentrate most of their primary technological resources and R&D personnel in Japan. Hence, to understand how these companies establish links between their core competencies and the knowledge acquired in a foreign environment, it is important to discern the processes of both the evolution and globalization of their domestic R&D functions.

To study this issue, Japanese synthetic chemical companies established by zaibatsu, i.e., large Japanese industrial and financial groups, were examined. Using securities reports, newspaper articles, and information on company history, the author analyzed: 1) the locational hysteresis and organizational structure of main R&D functions in Japan; and 2) changes in the spatial division of labor between domestic and overseas R&D facilities. Then, by applying social network analysis based on the patent applications by the three companies during 2005–2012, the knowledge flow among internal R&D facilities and between internal and external organizations was determined. These analyses were supported by interviews with company staff in 2012.

The results indicate that Sumitomo disperses its R&D efforts, but Mitsui and Mitsubishi have concentrated their R&D activities into a single site each, located in the Tokyo metropolitan area, since the 1990s megamergers conducted by the two zaibatsu. In addition, Sumitomo’s R&D system is “vertical” and focuses on its product divisions, while those of Mitsui and Mitsubishi are “horizontal” and encourage cooperation among R&D divisions. Furthermore, these differences affect the knowledge flow not only among their facilities and organizations in Japan, but also among their domestic and overseas R&D facilities. Because this study focuses on both domestic and overseas R&D activities, it provides a profound understanding of the spatial division of labor in R&D.

Recently, some cities have adopted urban policies based on a compact city design to facilitate accessibility mainly for the mobility-impaired in Japan’s aging society. The city of Toyama developed a “cluster-type” compact city model and suggested two main measures in the city master plan established in 2008: 1) improvement of public transportation connecting the city center and local subcenters; and 2) concentration of the population in the target area along public transportation lines from suburban areas. The aim of this study was to analyze the effects of those measures on accessibility to facilities used in daily life (the central railway station, general hospital, shopping center, and supermarket) and determine their implications for urban policy.

Accessibility, as represented in terms of time-distance to the nearest facilities, was measured based on road network analysis and bus timetables in the Fuchu area, a suburban area of Toyama. The results showed at the level that most (more than 90%) of the population would be able to reach those facilities. Simulations were conducted for 1) greater frequency of bus transportation and 2) migration into the target area. The major findings were as follows.

1) Accessibility could be substantially improved if bus frequency were increased under five-fold or the rate of migration were greater than 50%.

2) When both simulations were manipulated simultaneously, trade-offs occurred between the contribution of each measure to accessibility.

3) Accessibility for people aged 65 years and older improved relatively well because the ratio of senior citizens is high in peripheral areas.

As implications for urban policy on a cluster-type compact city, a short-term aim is to increase transportation frequency with the migration of the mobility-impaired population like senior citizens. A significant long-term goal is to concentrate the population to improve accessibility.