Networks in Organizational Populations: the Impact of Extra- Network Organizations on Firm Performance

Almost every business network exists in the context of a larger organizational population, but research has not produced a clear understanding of how similar firms outside a focal network influence the performance of network members. The existing literature offers two conflicting arguments and sets of results. One approach views the direct and indirect ties that exist between organizations as paramount and finds little empirical impact of extra-network organizations. The second approach, based on the concept of structural autonomy, takes into account the opportunities that exist outside of the network of direct and indirect ties, and finds a constant moderating effect of extra-network organizations on organizational action and performance. Using data on buyer-supplier relationships in the automobile industry, we help resolve this conceptual and empirical conflict by arguing that the impact of extra-network firms varies inversely with the relationship-specificity of network ties. Either ignoring extra-network firms or incorrectly attributing a constant moderating effect to them can lead to a dramatically incorrect understanding of how network relationships affect firm performance. 1 Almost every network of business relationships exists in the context of a larger organizational population. For example, a supplier normally sells to some, but not all, of the buyers to which it might potentially sell goods and services. The buyers, in turn, purchase goods from some, but not all, of their potential suppliers. Two primary strands of research in the existing literature offer conflicting arguments and results about the significance of such extra-network firms. One approach sees the ties that exist between organizations as paramount, so that extra-network organizations are generally beyond theoretical consideration (e.g. Galaskiewicz and Burt 1991; Gulati 1996). On the occasions that extra-network organizations enter a study's frame of reference, empirical testing generally finds them irrelevant (e.g. Davis 1991; Darr, Argote, and Epple 1995). The structural autonomy literature, on the other hand, emphasizes the need to understand the opportunities that exist outside a network of ties in order to understand the impact of an organization's network (Burt 1992). Both approaches suffer from a common limitation, however, because they assume that any extra-network influence will produce a constant effect on focal firm performance. Neither approach addresses differences in the nature of network relationships. This paper argues that either ignoring extra-network firms or assuming an unvaried relevance produces misleading conclusions about firm performance. We take a market-defined boundary of a population of organizations (Scott 1991:56-58), which in our case includes …