The Japanese corporate network: a blockmodel analysis

Administrative Science Quarterly,  March, 1992  by Michael L. Gerlach

• E-mail
• Print
• Link

Japanese intercorporate relationships are considered in terms of three different structures of interaction: corporate groupings, financial centrality, and industrial interdependency. The significance of each of these structures is tested with a sample of the forty largest industrial firms and twenty largest financial firms in Japan across networks of equity ownership, interlocking directorships, and banking borrowing. Macro-network blockmodeling analysis demonstrates that financial hierarchy is a pervasive pattern for all sets of relationships, and the position of industrial firms in the network is largely determined by corporate group membership (i.e., keiretsu). However, in contrast to predictions that financial centrality and corporate groupings will be used to manage industrial interdependencies among closely related firms, the results suggest that industrial firms with similar network positions represent highly diversified market sectors. These findings together provide a relatively finely detailed picture of the sources of network structure in Japan, variations that exist among different intercorporate network variables, the role of financial institutions in the Japanese economy, and the hierarchical nature of Japanese corporate groupings.(*)

In response to Japan's continuing competitive performance and to its ongoing trade frictions with important trading partners, researchers have become increasingly interested in recent years in the distinctive patterns of its overall industrial organization (e.g., Aoki, 1988; Hamilton and Biggart, 1988; Lawrence, 1991; Fruin, 1992; Gerlach, 1992). Located at a level intermediate between the internal managerial practices of the Japanese firm and the national and international forces that define Japan's macroeconomy and industrial policies, the Japanese corporate network is marked by an elaborate structure of institutional arrangements that have organized its companies within complex patterns of cooperation and competition. The result, in the words of two observers, is "a thick and complex skein of relations matched in no other industrial country" (Caves and Uekusa, 1976:59).

Among the widely noted characteristics of intercorporate relationships in Japan is the existence of highly visible clique-like patterns based on intercorporate alliances, or keiretsu. Diversified groupings of firms descended from the prewar zaibatsu, or family-centered holding companies, involve firms that often share the same names and logos (e.g., Mitsubishi and Sumitomo) and organize relationships among major financial institutions, trading companies, and industrial producers (Hadley, 1970; Orru, Hamilton, and Suzuki, 1989; Gerlach, 1992). Large industrial concerns (e.g., Toyota Motors and Hitachi) are in turn positioned at the apex of their own vertically linked groupings of smaller affiliated supplier and distribution firms (Aoki, 1988; Asanuma, 1989; Fruin, 1992).

Despite the apparent significance of densely clustered corporate groupings, however, basic questions remain about the overall structure of Japanese business networks: To what extent do nominal groupings correspond to empirical patterns of network structure using standard network methods? Are these groupings the main basis for network structure in Japan, or are other forms of intercorporate relationship equally important? In particular, do Japanese banks and other financial institutions play the same kind of central role in networks of directorships, stock ownership, and corporate control as they do in the U.S.? If so, are ties to common financial institutions the main source of structural coherence in enterprise groupings, or are direct ties among affiliated industrial firms equally important? Furthermore, how do these patterns interact with the business interests of firms in the network - by, for example, linking competitors or vertically related suppliers and customers through common structures of intercorporate coordination and control?

Most studies of Japan's corporate network have proceeded by first identifying nominal groupings of firms based on some criterion for classification. One commonly used measure is formal participation in one of the executive councils that bring together executives of affiliated companies in regular monthly meetings for the purpose of social and strategic interaction (Kosei torihiki iinkai, 1983; Gerlach, 1992). Other researchers have instead relied on classifications provided by public data sources, which are based on considerations such as the history of relationships and contemporary patterns of interaction (Nakatani, 1984; Orru, Hamilton, and Suzuki, 1989). These nominal groupings are then used as the basis for describing characteristics of or relationships among those firms that are of theoretical interest. Structurally oriented researchers have used them to measure relatively precisely the characteristics of the relationships among firms in the same keiretsu and to compare the coherence of different groupings (Hadley, 1970; Futatsugi, 1976; Orru, Hamilton, and Suzuki, 1989; Gerlach, 1992). Econometrically oriented researchers have used patterns of affiliation as the basis for tracing performance differences among Japanese firms in order to infer the functions groups play for member firms (Caves and Uekusa, 1976; Nakatani, 1984; Cable and Yasuki, 1985).