The Social Structure of Organization : Coordination in a Large, Multi-Business Firm

In my dissertation, I examine intraorganizational social networks and their antecedents and consequences. The first paper, jointly authored with Michael Tushman, is a theoretical discussion of the role of social networks in inter-divisional coordination. Most large organizations fail to develop new businesses that combine resources from disparate parts of the firm. I define and explore a brand of corporate entrepreneurship based on interdependent innovation - the deliberate creation of interdependence between autonomous divisions of multi-business firms to create new products. I argue that interdependent innovation is difficult because the social structures that promote exploration of new possibilities are inconsistent with the social structures needed to successfully execute interdependent innovation; I suggest that senior leadership plays a crucial role in transitioning the organization between different network structures.Empirically, there are at least two methodological hurdles to researching the complex interaction between formal structure and social structure in contributing to organizational outcomes. First, the kind and quality of data that have typically been collected to conduct network analysis are inadequate; and second, there is a paucity of research that accounts for the embeddedness of the informal structure in the formal. In the second dissertation paper, I begin to resolve these two issues. I argue for data collection methods relying on electronic communication archives (e.g., e-mail) for network analysis. I also empirically develop novel measures that use this data to quantify the social structural relationships between formal divisions; in doing so, I explicitly embed informal structure within formal structure in novel ways.The third dissertation paper, jointly authored with Toby Stuart and Michael Tushman, is an empirical study of the pattern of communications - and, by extension, the coordination that it enables - in a modern organization. We analyze a dataset with more than 100 million electronic mail messages, calendar meetings and teleconferences for a sample of more than 30,000 employees of a single, multidivisional firm. In dyad-level models of the probability that pairs of individuals communicate, we find very large effects of spatial proximity and formal organizational structure on the rate of communication; homophily effects based on gender, organizational tenure, and salary levels are much weaker