«Adam M. Kleinbaum Toby E. Stuart Harvard Business School Harvard Business School Boston, MA 02163 Boston, MA 02163 akleinbaum tstuart ...»
Communication (and Coordination?) in a Modern, Complex Organization
Adam M. Kleinbaum Toby E. Stuart
Harvard Business School Harvard Business School
Boston, MA 02163 Boston, MA 02163
Harvard Business School Boston, MA 02163 email@example.com 617.495.5442 Under review at Administrative Science Quarterly The authors would like to thank Pierre Azoulay, Mikolaj Piskorski, Bill Simpson, and Julie Wulf for helpful suggestions. We also benefited from the comments of seminar participants at Stanford, MIT, Columbia, Yahoo Research, and Harvard. The first author gratefully acknowledges financial support from the Ewing Marion Kauffman Foundation. Please direct correspondence to Adam Kleinbaum (firstname.lastname@example.org).
-1- Communication (and Coordination?) in a Modern, Complex Organization Abstract This is a descriptive study of the structure of communications in a modern organization. We analyze a dataset with millions of electronic mail messages, calendar meetings and teleconferences for many thousands of employees of a single, multidivisional firm during a three-month period in calendar 2006. The basic question we explore asks, what is the role of observable (to us) boundaries between individuals in structuring communications inside the firm? We measure three general types of boundaries: organizational boundaries (strategic business unit and function memberships), spatial boundaries (office locations and inter-office distances), and social categories (gender, tenure within the firm). In dyad-level models of the probability that pairs of individuals communicate, we find very large effects of formal organization structure and spatial collocation on the rate of communication. Homophily effects based on sociodemographic categories are much weaker. In individual-level regressions of engagement in category-spanning communication patterns, we find that women, mid- to high- level executives, and members of the executive management, sales and marketing functions are most likely to participate in cross-group communications. In effect, these individuals bridge the lacunae between distant groups in the company‘s social structure.
-2- Communication (and Coordination?) in aModern, Complex Organization
Introduction Coordination (and the communication it implies) is central to the very existence of organizations. Theories of the firm are variously rooted in the coordination benefits of hierarchical control (Thompson, 1967) – relative to the market, the superior capacity of managerial hierarchies to efficiently coordinate transactions involving specific assets (Williamson, 1975) – and the synergistic potential of coordinating multiple activities within a single corporate enterprise (Chandler, 1977). In these and other theories of the firm, the key managerial task is to effect coordination. And consistent with these theories, survey and ethnographic studies of managerial behavior have revealed that leaders spend upward of 80 percent of their time interacting with other people (Mintzberg, 1973;Kanter, 1977). The implication of extant theories is that organizational members communicate to coordinate activities. Moreover, the quintessentially social nature of managerial work is evident in the vast proportion of management time devoted to interacting with others.
Despite the fundamental role of coordination – and the communication that enables it – to the purpose of organizations, we have little understanding of actual interaction patterns in modern, complex, multi-unit firms. To open the proverbial ―black box‖ and begin to reveal the internal wiring of the firm, this paper presents a detailed, descriptive analysis of the network of communications among members of a large, structurally, functionally, geographically, and strategically diverse firm (hereafter, ―BigCo‖). The full dataset comprises more than 100 million electronic mail messages and over 60 million electronic calendar entries for a sample of more
also possess basic organizational, demographic, and social information, including gender, salary band, tenure, business unit, job function, and office location.
Because of the importance of boundaries in general theories of social structure and specific theories of organizational behavior, we focus on the role of observable (to us) boundaries between individuals in structuring communications inside the firm. Our data enable us to discern three types of boundaries: organizational unit (e.g., strategic business unit, function membership), spatial boundaries (e.g., office locations), and social categories (e.g., gender, tenure within the firm). We conduct many of the analyses in the paper at the level of the pair of individuals: we examine how organizational, spatial, and social boundaries affect the frequency of dyadic interactions.
After generating estimates of the effects of boundaries on the frequency of communication, we then flip this analysis on its head: we calculate person-specific measures of the degree to which each individual in the sample engages in communications that deviate from the modal pattern of intra- and inter-group interaction in the data. Aggregating across all of an individual‘s dyadic exchanges, people who score highly on this variable are category spanners – their interactions connect rarely traversed organizational and social groups. Because communication is dense within most categories and sparse between them, category spanners are far more likely to create the ―weak‖ (Granovetter, 1973) and ―bridging‖ (Burt, 1992) ties in the organization. Without these people, BigCo would devolve to a structure of clan-like silos separated by relational voids.
interactions, the paper‘s primary aim is descriptive and we refrain from formulating any specific hypotheses. Our theoretical development centers on the potential implications of different theories of organization for likely communication structures within complex organizations, but other than opinions formed from our own anecdotal observations of large organizations, we have no compelling a priori reason to give precedence to one theory over another. For this reason, we do not offer predictions. Moreover, at the level at which we can measure the communications network within BigCo, there is not a conclusive, one-to-one mapping between the evidence we marshal and the different theories of organization that populate the literature. Thus, despite our belief that the preponderance of the evidence is most consistent with one point of view – classical organization theory‘s emphasis on formal structure in shaping interaction – we will make no strong claim of proof.
We present too many descriptions of the BigCo communication network to summarize in their entirety here, but a few findings are noteworthy and at least somewhat unexpected. First, relative to men, women participate in a greater volume of electronic and face-to-face interactions and do so with a larger and more diverse set of communication partners. This finding cannot be explained by gender sorting into different work roles, such as secretarial positions (although there is a gendered division of labor within the firm). Second, organizational boundaries – business unit, job function, and office location – have an enormous influence on who interacts with whom inside the firm. As a summary statistic, we find that relative to two people that share none of these categories in common and who are geographically separated by the sample‘s mean dyadic distance, a pair of individuals that shares the same business unit, job function, and office location communicates at an estimated rate that is approximately 1,000 times higher. Third,
delimited by the pathways of formal organizational structure. By contrast, in interactions amongst themselves, executives show a somewhat stronger tendency toward sociodemographically (gender and tenure) homophilous communication partners than do other employees. Finally, the category spanners in the firm are women concentrated in the uppermiddle management ranks and in a few functions, most notably sales, marketing, and general executive management.
Theories of Organization & Communication The literature offers starkly different possibilities for how communications might weave together a social fabric inside the firm. Weber (1924) provided us with the enduring image of the rationalized, formalized, hierarchical bureaucracy. This image carries through much of the classic work in organization theory and has implications for the structure of communications we might expect to observe within complex organizations. For instance, Chandler (1962) famously characterized many of the large organizations since the turn of the last century as adopting Mforms, in which operational decisions occur within business units and strategic decisions are managed at the headquarters level. Adopting a Chandlerian view, which is also echoed in many contemporary texts on value creation in the diversified firm, we should expect to observe a high density of communications nested within business units and, insofar as the organization is structurally integrated, the headquarters unit should play a central role in bridging communications among autonomous business units. While subsequent work, such as Lawrence and Lorsch‘s (1967) contingency-theoretic discussion of integration mechanisms and Galbraith‘s (1973) focus on lateral coordinating mechanisms, does complicate the story, classic organization theories nevertheless imply that interaction patterns inside the firm will, for the most part, be
expect hierarchical communication patterns that unfold within formal organizational units.
Albeit not extensive, there is some evidence to suggest that formal structure does strongly influence communication patterns within firms. A few of the survey-based studies of intra-firm networks have suggested that an organization‘s formal structure forms the backbone of the actual relational structure of the firm. For instance, in an analysis of four different types of relations, Han (1996) found that the network of interactions was tightly bound to the formal reporting structure. Although not the primary purpose of the paper, a similarly central role of formal structure in shaping interaction patterns is evident in Burt‘s (2004) analysis of social capital effects in the supply chain function of a large electronics company.
There are, however, competing views. One of organization theory‘s most taken-forgranted assumptions is that informal structures of power, influence, and information exchange emerge within organizations. These informal structures are thought to significantly influence interaction patterns and, indeed, the ―informal‖ organizational chart is often held to be more consequential than the formal one (Mayo, 1949;Krackhardt, and Hanson, 1993). Moreover, some of the most prominent students of organizations have viewed structure and action as being only loosely coupled (Weick, 1976). In Cohen, March and Olsen‘s (1972) garbage can model, for example, people, problems, and solutions admix by chance, and organizational action can appear almost random relative to formally prescribed decision hierarchies.
If communication patterns map to informal structure, the relevant question for us is, what would the observable manifestation of this be in the intra-organizational interaction network?
Although many different social dimensions might serve as the foundations of informal structure,
homophilous interaction patterns (Lazarsfeld, and Merton, 1954;Blau, and Schwartz, 1984), we might expect to find that communication is much more prevalent within sociodemographic categories. For instance, within formal organizations and in social networks more generally, there is considerable evidence of gender homophily: many individuals are immersed in networks comprising primarily same-sex ties (e.g., McPherson, and Smith-Lovin, 1986;Marsden, 1988;Ibarra, 1992;Ridgeway, and Smith-Lovin, 1999). Similarly, there is evidence that social ties cluster within age-based and tenure-based strata (e.g., Zenger, and Lawrence, 1989). Thus, if the communication pathways within organizations are informally structured, we might anticipate that similarities along sociodemographic dimensions have a primary influence on who communicates with whom.
If the formal and informal structure perspectives suggest, respectively, that organizational or sociodemographic boundaries will exert a first-order influence on interaction patterns within an organization, a third point of view – represented in the literature under rubrics such as the ―boundaryless‖ firm, the ―networked‖ organization (Powell, 1990;Nohria, 1996), and the knowledge-based firm (Kogut, and Zander, 1992) – reflect a contemporary image of organizations characterized by free-flowing, lateral, and collaborative interaction patterns. In this perspective, neither rigid hierarchies nor social categories necessarily play a dominant role in orchestrating intra-organizational interaction. One might trace the origins of these perspectives to Burns and Stalker‘s (1961) distinction between mechanistic and organic organizations. The notion of more organic organization structures gained momentum with Ouchi‘s (1980;1981) provocative characterization of ―clan structures‖, in which a commonly held and broadly internalized set of goals were thought to replace hierarchy as the mechanism of governance.
leadership is disassociating from formal hierarchies and is migrating to lower levels of the organization.