Understanding network formation in strategy research: Exponential random graph models

Research summary: This article uses Exponential Random Graph Models (ERGMs) to advance strategic management research, focusing on an application to board interlock network tie formation. Networks form as the result of actor attributes as well as through the influence of existing ties. Conventional regression models require assumptions of independence between observations, and fail to incorporate endogenous structural effects of the observed network. ERGMs represent a methodological innovation for network formation research given their ability to model actor attributes along with endogenous structural processes. We illustrate these advantages by modeling board interlock formation among Fortune 100 firms. We also demonstrate how ERGMs offer significant opportunities to extend existing strategy research and open new pathways in multiparty alliances, microfoundations of interorganizational network formation, and multiplexity of ties among actors. Managerial summary: Social networks are increasingly important in the business world, not only between individuals but also between organizations. Firms can obtain information, resources, and status through their external network connections, and understanding how these outside ties form is an important goal of strategy research. Our paper helps advance this effort by introducing a new tool for social network analysis, Exponential Random Graph Models (ERGMs) to the management and strategy fields. We provide an example of this method, demonstrating how social network ties form between companies when they hire common directors to their boards. Executives can benefit from this research through a greater understanding of how corporate relationships are built with allies as well as among competitors.