Network influence and organizational performance: The effects of tie strength and structural equivalence

Summary

We combine theory and empirical evidence to develop and test three related claims about how ties strength and structural equivalence affect interorganizational performance similarity. First, network partners attain similar levels of performance because direct network ties facilitate information and resource sharing, and the diffusion of organizational practices. Organizations occupying similar network positions attain similar levels of performance because they face similar competitive constraints. Second, because mechanisms underlying performance similarity implied by direct network ties and similarity in network positions are different, strength of network ties and network position affect different dimensions of organizational performance. Third, we expect that tie strength will affect aspects of performance that are more likely to be influenced by processes activated by the direct contact between network partners. We expect structural equivalence to affect more strongly those aspects of organizational performance that are more directly influenced by competition. An analysis of interorganizational networks and organizational performance within a regional community of hospitals provides evidence in support of these claims. Models estimated on a sample of 8190 interorganizational dyads suggest that the effects of tie strength and structural equivalence on organizational performance are highly differentiated and contingent on specific aspects of performance. We show that these conclusions are robust with respect to a wide variety of other factors that may be responsible for performance similarity including competitive interdependence, structural homophily, and resource complementarity.

Introduction

How do networks of relations in which organizations are embedded influence their performance? Substantial research on interorganizational networks conducted across a variety of empirical setting and theoretical orientations sustains two broad views (Mizruchi & Marquis, 2005). The first view is based on the concept of strength of contact between organizations and their network partners. Network ties between organizations act as conduits for information, knowledge, organizational practices and material resources. As a consequence the presence of strong network ties is likely to make partners more homogeneous in terms of structures, behavioral orientation and, arguably, performance (Ahuja, 2000, Davis and Greve, 1997, Hedström et al., 2000). As Brass, Galaskiewicz, Greve, and Tsai (2004) note, the claim that information transmission leads to imitation is a basic tenet of both institutional (DiMaggio & Powell, 1983), as well as organizational learning theories (Levitt & March, 1988). The second general view is based on the notion of structural equivalence, or similarity in networks positions that organizations come to occupy by virtue of similarity in their patterns of relation with third parties (Lorraine & White, 1971). Structurally equivalent organizations depend on the same resources (or resources controlled by the same alters) and are, therefore, more likely to compete (Burt, 1987, Mizruchi, 1993). Competition induced by structural equivalence may also lead to similarity in behavior and performance because organizations occupying the same position are more likely to emulate the actions of others perceived to be in a similar position with respect to fundamental resource dependencies (Hallen, 2008, Zaheer and Bell, 2005).

Earlier comparative studies treated tie strength and structural equivalence as alternative mechanisms underlying social influence, and as competing explanations of behavioral similarity (Mizruchi, 1993). For example, in his re-analysis of data from the classic study by Coleman et al., 1996, Burt, 1987 suggested that actors grouped together on the basis of structural equivalence were more likely to behave similarly than actors grouped on the basis of direct, cohesive ties. In contemporary organizational studies this structural view has been somewhat tamed. More than two decades of empirical research have produced the generally accepted view that structural equivalence not only does not preclude direct cohesive ties, but it may actually amplify the tendency to imitate partners connected by direct contacts, and thus strengthen the tendency toward similarity in structure and behavior (Brass et al., 2004). As a consequence, it is now “Generally accepted in organization theory and strategic management that firms are ‘relationally’ and ‘positionally’ embedded in interfirm networks that affect their behavior and performance” (Rowley, Baum, Shipilov, Greve, & Rao, 2004).

In this study we take this view as our point of departure to build a more detailed understanding of the mechanisms underlying tie strength and structural equivalence in interorganizational networks. We do not take issue – and in fact agree – with the claim that strong ties (relational perspective) and structural equivalence (positional perspective) may coexist and have mutually reinforcing effect on organizational similarity. We think, however, that discriminating among rival (unobservable) mechanisms capable of yielding similar empirical outcomes remains a central challenge for empirical research on interorganizational networks.

With this analytical objective in mind, in this paper we develop and test three main claims. The first is that both the strength of network ties and similarity in network position tend to reduce interorganizational performance differentials by diffusing organizational structures, behavioral orientations, and best practices. Following directly from the first, the second claim is based on the recurrent observation that similarity in performance between organizations may be determined both by diffusion of solutions and practices between partners facilitated by direct contact, and by competitive mimicry induced by structural equivalence (Burt, 1987, Burt, 1992, Mizruchi, 1993). If this is the case, tie strength and structural equivalence will be associated to different aspects of organizational performance similarity. Whence our third claim: tie strength will be a better predictor of interorganizational performance similarity when performance depends on activities that may be affected by process of social influence activated by direct contact. Correspondingly, structural equivalence will be a more accurate predictor of interorganizational performance similarity when performance depends on competitive intensity, or rivalry between organizations occupying similar network position.

To examine our claims empirically we rely on data about performance and network relations that we have collected within a regional community of hospital organizations. We model dyadic similarity in organizational performance as conditional on relevant attributes of the partner hospitals, the strength of their network relations, their position in the interorganizational networks, and a variety of institutional features. We selected specific measures of performance in order to facilitate discrimination between aspects of organizational activities that may be influenced mostly by the collaborative diffusion of best practices through contact, and aspects that are influenced by the level of diffuse competition between organizations.

We reconstruct the relevant interorganizational network in terms of acts of dyadic patient transfers. Acts of patient transfer between hospitals are directly observable and are not possible without high levels of coordination and communication between partner hospitals. Patient transfer flows reflect collaboration between hospitals (Iwashyna, Jason, Moody, Kahn, & Asch, 2009). As pointed out by Gittell and Weiss (2004, p. 128), when multiple parties are involved in the care of any given patient: “The clinical information that these parties need to transmit about patients is highly complex, due to the growth and specialization of clinical knowledge, and due to the multitude of conditions patients may suffer in combination”. The complexity of information that partner hospitals need to exchange makes patient transfer impossible without a high level of collaboration and coordination between the hospitals involved. Building on research on interorganizational fields in health care sectors, we consider patient transfer as a reliable proxy for inter-hospital collaboration (Lomi and Pallotti, in press, Iwashyna et al., 2009).