Embedded patterns of international alliance formation

Organization Studies, Fall, 1998 by Richard N. Osborn, John Hagedoorn, Johannes G. Denekamp, Geert Duysters, C. Christopher Baughn

Even though much of the current alliance literature seems to use product/knowledge flows to characterize the nature of alliance's reciprocal interdependence, there appears to be a split opinion on how to deal with different types of product/knowledge flows. On the one hand, there are those who emphasize transaction-cost minimization (e.g. Williamson 1991). Here, the joint-venture form, with its administrative logic and emphasis on control, would be preferred over less hierarchical administrative forms for coping with the uncertainty stemming from reciprocal flows and for monitoring compliance with the initial agreement (e.g. Burgers et al. 1993; Hennart 1988). On the other hand, Hagedoorn and his colleagues (Duysters and Hagedoorn 1993; Hagedoorn 1993b) suggest that the associative logic found in the agreement form might be more appropriate for stimulating exchange and development. Osborn and his colleagues (see Osborn and Baughn 1993; Baughn et al. 1997) have a similar argument. They suggest that mutual control in an agreement form with reciprocal interdependence may be accomplished by the potential threat of one party to withhold the knowledge needed to proceed to the next stage of commercialization.

International Alliance Sub-populations as Specific Patterns of Embeddedness

So far, the analysis has attempted to recognize the duality and complexity of international alliances (see Osborn and Hagedoorn 1997). They are cooperations (frameworks for action) between self-interested corporations, where these sponsors are located in separate nations and cultures. Executives are expected to imitate other apparently viable alliances in the alliance's industry. Yet, we also suggested that to form an alliance the executives need to agree to a logic of association, contract or administration and to establish a pattern of product/knowledge exchanges. Sponsors are expected to socially construct an alliance consistent with their own interests and bargaining power as well as the alliance's function. We explicitly noted some of the inconsistent recommendations from the current literature concerning the choice of administrative forms and product/knowledge flows.

Holding other factors constant (such as nationality, corporate size and differences in bargaining power), we now contend that some specific embedded combinations are more likely than others. That is, consistent with industry-specific conditions and opportunities for economic and/or technical adaptation, some combinations of administrative form and product/knowledge flow might be favoured over others. While there are potentially many combinations, two are suggested by the more traditional alliance analyses. One is based on transaction cost economics. The second relies more on the technology literature.

Dominance

Even in technically stable industries, there may still be considerable uncertainty in forming an alliance across national boundaries. Sponsors should be prepared, for instance, to react to a host of shifting national, firm-specific, and transaction-specific advantages as they struggle in a continuing battle for global competitive dominance (cf. Dunning 1993). The shifting setting may provide ample opportunities for opportunism (Williamson 1991, 1996). Thus, firms may seek the protection of a hierarchical administrative form (e.g. joint venture). The product/knowledge flows are likely to stem from national or firm comparative advantages and favour unidirectional rather than reciprocal exchanges. In Thompson's (1967) terms, sponsors will opt for pooled and sequential interdependence. The pattern of more stable industrial conditions with unidirectional product/knowledge flows through equity alliance forms to minimize opportunism may be called a dominance pattern. The pattern reflects the underlying rationale presented by many transaction cost and internalization theorists (e.g. Burgers et al. 1993; Dunning 1993; Gulati 1995; Williamson 1996). In sum: