Software Choice and Productivity in Japan - Custom Software vs. Packaged Software (Japanese)

The emphasis on custom software is often cited as a reason why Japanese software makers lack international competitiveness, however, the dominance of packaged software in the Japanese market is also a major contributing factor. Custom software is said to be inefficient compared with packaged software in terms of cost and quality because it can't derive benefit from economies of scale and global market competition. And although U.S. firms changed their IT investment from custom software to packaged software in the 1980s, Japanese firms still continue to use custom software. So, why do Japanese firms continue to use custom software, when it is believed to be antiquated? This paper aims to determine why Japanese firms stick to custom software by analyzing industry-based data from software-user firms in Japan. Unexpectedly, the results show that custom software is an efficient choice for Japanese firms. Although there are some negative reasons for choosing custom software such as low IT literacy of workers and network externality, the main reason for choosing custom software is to reflect the firm-specific knowhow on the information system. Custom software is suitable for reflecting firm specific knowhow because it is designed exclusively for a specific firm. This conclusion is supported by the striking fact that Japanese firms using custom software show higher productivity than firms using packaged software. Since it is unlikely that software choice causes an increase in productivity, we should say that Japanese firms with high productivity tend to choose custom software. The competitiveness of Japanese firms comes from the accumulated firm specific knowhow of long-term employees, making custom software an efficient choice to utilize the knowhow of these workers. Conversely, the competitiveness of U.S. firms comes from quick resource reallocation of workers and capital, making packaged software an efficient choice for U.S. firms because of its short lead time and standardization. We find that the difference in software choice between Japanese and U.S. firms is based on each firm's attempt to maximize their own competitive advantage.