Nobeoka, Dyer, and Madhok (2002) statistically demonstrated the relationship between suppliers' performance and suppliers' customer scope; the latter was measured by (1) the number of automakers to whom components were supplied and (2) the concentration ratio of automakers (Herfindahl index) for seven Japanese automotive assemblers (automakers), using the data for 1995 on Japanese automotive parts industry. They concluded that the greater the number of customers (automakers) and the lower the concentration ratio, the higher the supplier's performance (profit-sales ratio). Min and Song 166 their performance after dividing the suppliers into two groups: (A) suppliers with three or more automaker customers and (B) suppliers with less than three automaker customers. The results show that (A) the suppliers with more customers have significantly higher performances in 1985 and 1995; however, in 2005, (B) the suppliers with less customers have a higher profit-sales ratio, although it is not significant. It is highly likely that the correlation between supplier performance and the number or the concentration ratio of automaker clients might result from a spurious correlation caused by other antecedent variables.

Von Hippel (1988) coined the term "user innovation" by arguing that semiconductor device makers (equipment users) played a critical role as innovators in the development of semiconductor process equipments. However, in the same period of this study, the Japanese semiconductor process equipment manufacturer, ULVAC, autonomously conducted fundamental research on the material composition of cobalt-nickel-chrome (CoNiCr), a study that is generally conducted by device makers. Based on the material research outcome, ULVAC developed new manufacturing equipment for HDD and supplied it to users (device makers). Here we define this phenomenon as "adverse user innovation," which works in an opposite manner to the "user innovation" of von Hippel (1988). In fact, there are likely to be numerous other examples similar to this case.

Does keiretsu in the Japanese automotive industry vary by company in terms of firm performance and strategic behavior? This paper classifies parts-supply keiretsu into (1) Toyota suppliers and non-Toyota suppliers and (2) Nissan suppliers and non-Nissan suppliers and then conducts a comparative analysis of the suppliers to test differences in (a) firm performance, (b) customer scope, and (c) product diversity. The following results emerge from the analysis: (a) in regard to "firm performance (return on sales)," Toyota suppliers outperformed non-Toyota suppliers, but there was no difference between Nissan suppliers and non-Nissan suppliers; (b) with respect to "customer scope," both Toyota suppliers and Nissan suppliers outperformed non-Toyota suppliers and non-Nissan suppliers, respectively; (c) as for "product diversity," Toyota suppliers have less product diversity than non-Toyota suppliers, but there was no difference between Nissan suppliers and non-Nissan suppliers. These results indicate that the strategic behavior of Toyota suppliers differs from that of Nissan suppliers in that Toyota suppliers achieve high profitability by broadening customer scope and simultaneously narrowing product