Japanese Journal of Human Geography
Online ISSN : 1883-4086
Print ISSN : 0018-7216
The Location of the Software Industry
An Extended Shift-share Analysis Based on International Trade
Naoto YABE
Author information

2005 Volume 57 Issue 4 Pages 428-443


In the late 1980s, the software industry in Japan grew rapidly not only in the 'global city' of Tokyo but also in some rural areas. In short, it was dispersed from the global city to other areas in Japan. In the late 1990s, however, the software industry became concentrated mainly in Tokyo, unlike the late 1980s. The purpose of this paper is to investigate the background of that concentration within the context of international trade.
Two hypotheses were examined as driving forces of that concentration. 1) In the late 1980s, software firms in Tokyo outsourced software business to rural areas in Japan where a low cost labor force was available. In the late 1990s, however, software firms in China and India provided lower costs than the rural areas of Japan. Thus, it was expected that software firms in Tokyo outsourced software not to the rural areas in Japan, but to overseas software firms. 2) As the demand for producer services by Trans National Corporations (TNCs) located in the global city grew with the economic globalization of the 1990s, it was anticipated that demand for software would increase in the 'global city' of Tokyo rather than in the rural areas.
First, the performance of an extended shift-share analysis on the software industry revealed that the import substitution effect is rarely seen in the Japanese software industry although labor costs are significantly reduced by outsourcing to China and India. One reason for limited overseas outsourcing is that electronic devices made by Japanese manufacturers need highly skilled workers. Embedded software in electronic devices like mobile phones and digital cameras need highly skilled workers most of whom live in the Tokyo metropolitan area. The demand for skilled workers for high value electronic manufacturing products prevented further outsourcing of the software overseas and facilitated the concentration of the software industry in Tokyo. Second, the extended shift-share analysis of TNCs growing demand for software shows that export-led job growth became evident in the late 1990s. The correlation coefficient between export-oriented job growth and software job growth is significant at the 1% level. The high demand for software in the 'global city' and the decrease in government public works investment in rural areas have greatly contributed to the concentration of the software industry in Tokyo in the late 1990s. That helped shift software demand from the rural areas in Japan to the 'global city', Tokyo.
With a governmental income transfer policy from urban to rural areas, the software industry grew relatively all over Japan. However, the growth gap between the urban and rural areas was mainly aggravated by economic globalization and the minimal government policy in the 1990s.

Information related to the author
© The Human Geographical Society of Japan
Previous article Next article