Network Effects in the Japanese Word-Processing Software Market (2003.12)

Home >Japan Fair Trade Commission >CPRC Home >Report/Paper >

File （PDF)	Contents，Author
CPDP- 2-E (PDF:136KB)	"Network Effects in the Japanese Word-Processing Software Market" (2003.12) Yoshihito YASAKI(Research Center for Advanced Science and Technology, University of Tokyo and Competition Policy Research Center, Fair Trade Commission of Japan) Reiko MURAKAMI(Competition Policy Research Center, Fair Trade Commission of Japan) This paper examines the importance of network effects in the Japanese market for word-processing software in the period between 1998 and 2001, using hedonic price and nested logit models. This was a period in which Microsoft Word had a large and steadily increasing share of the market. The presence of network effects was verified in the full regressions covering the entire period. Separate regressions for two consecutive years showed that network effects, as measured by the positive effect of the size of the user base on the product price or on the probability of the software group being chosen, were weak or insignificant in the beginning of the sample period. The paper discusses data and methodological issues, and possible reasons for these results, including the effects of switching costs. JEL classification codes: L86; L11; L13. Keywords: Network Effects, Software, Hedonics, Discrete Choice Model, Switching Costs.

File （PDF)

Contents，Author

"Network Effects in the Japanese Word-Processing Software Market" (2003.12)
Yoshihito YASAKI(Research Center for Advanced Science and Technology, University of Tokyo and Competition Policy Research Center, Fair Trade Commission of Japan)
Reiko MURAKAMI(Competition Policy Research Center, Fair Trade Commission of Japan)
This paper examines the importance of network effects in the Japanese market for word-processing software in the period between 1998 and 2001, using hedonic price and nested logit models. This was a period in which Microsoft Word had a large and steadily increasing share of the market. The presence of network effects was verified in the full regressions covering the entire period. Separate regressions for two consecutive years showed that network effects, as measured by the positive effect of the size of the user base on the product price or on the probability of the software group being chosen, were weak or insignificant in the beginning of the sample period. The paper discusses data and methodological issues, and possible reasons for these results, including the effects of switching costs.
JEL classification codes: L86; L11; L13.
Keywords: Network Effects, Software, Hedonics, Discrete Choice Model, Switching Costs.