Social Networks and Organizations

Martin Kilduff, Wenpin Tsai

`The authors should be congratulated for not only offering an excellent tour de force of cutting-edge work in social network analysis, but also charting some new possible territories for future organizational research? -

Environment and Planning

Social Networks and Organizations provides a compact introduction to major concepts in the area of organizational social networks.

The book covers the rudiments of methods, explores major debates, and directs attention to theoretical directions, including a vigorous critique of some taken-for-granted assumptions. The book is aimed at all of those who seek a lucid and lively treatment of social network approaches to organizational research, with a particular emphasis on the neglected area of interpersonal networks in organizations.

In this book, Martin Kilduff and Wenpin Tsai offer new insights to those already familiar with network analysis, and motivate those interested in pursuing network research to embark on journeys of discovery.

`This book is extremely timely. It provides a wonderful synthesis of the recently burgeoning literature in the area of organizations and social networks. It should be relevant at once for both the experienced network scholar as well as those entering this growing area? - Ranjay Gulati, Kellogg School of Management, Northwestern University

`Martin Kilduff and Wenpin Tsai have done a marvellous job of not only reviewing and integrating the diverse streams of literatures on social networks, but also of showing the enormous potential of this research approach that still lies untapped. Overall, this book will prove to be an invaluable resource for interested graduate students as well as for established scholars in the field? - Sumantra Ghoshal, Professor of Strategic and International Management, London Business School

`Research on social networks is already one of the most vibrant areas of organizational inquiry. How can it possibly become any more so? This book by Kilduff and Tsai opens up many new avenues for network research and theory-building. Whether you?re newly-interested in social networks or a veteran of the topic, you will benefit from Kilduff and Tsai?s marvellous contribution? - Donald C Hambrick, Smeal College of Business Administration, The Pennsylvania State University

1	Introduction

CHAPTER CONTENTS

• Exploring the range of organizational social network research

• Mapping the chapters in the book

• Summary

• Recommended further reading

On the night of 17 April 1775, two men rode different routes from outside Boston to Lexington warning communities along the way of the imminent threat from the British army. The message delivered by Paul Revere and William Dawes on their midnight rides was dramatic: the next day would see the British army marching on Lexington to arrest colonial leaders and then on to Concord to seize colonial guns and ammunition. Both Revere and Dawes carried the identical message through just as many towns over just as many miles. Paul Revere’s message spread like wildfire in communities such as Charlestown and Medford, but Dawes’s message failed to catch fire, with the result that in towns such as Waltham even the local militia leaders weren’t aware of the British moves. Why was there a difference in the reception of this identical message? Evidence suggests that Paul Revere was connected to an extensive network of strategic relationships whereas William Dawes’s connections were less useful. Paul Revere ‘knew everybody. … When he came upon a town, he would have known exactly whose door to knock on, who the local militia leader was, who the key players in town were’ (Gladwell, 2000: 23). Not only did Revere alert whole towns to the looming threat, the leaders in these towns themselves sent riders to alert the surrounding areas. Dawes’s message failed to spread through the network whereas Revere’s message rapidly diffused.

The moral of this tale is that the network of relationships within which we are embedded may have important consequences for the success or failure of our projects. Evidence suggests that the types of network we form around ourselves affect everything from our health, to our career success, to our very identities. One study of a randomly selected sample of 6,928 residents of Alameda County, California, over a nine-year period showed that people who ‘lacked social and community ties were more likely to die … than those with more extensive contacts’ (Berkman and Syme, 1979: 186). The study controlled for a host of other possible causes of mortality, such as smoking, alcohol consumption, obesity, physical activity and utilization of preventive health services. A follow-up study looked at the same sample over a 17-year period and confirmed these results, but also found that extensive contact with friends and relatives (compared to contact with spouse) was particularly important in reducing mortality for those over the age of 60 (Seeman et al., 1987). Another study showed that maintaining a diverse network reduced susceptibility to the common cold (Cohen et al., 1997). People who had frequent contact (in person or on the telephone) to others across a wide range of relationship types (relatives, neighbours, friends, workmates, members of social groups, etc.) tended to resist infection better than those whose contacts were with a narrower range of relationship types. Maintaining network ties to different groups of people in organizations has been associated with higher performance ratings (Mehra et al., 2001), and faster promotions (Burt, 1992). Having the right contacts can help you get a job (Granovetter, 1974) and can help you negotiate a higher salary (Seidel et al., 2000).

The extent to which people are engaged in social activities in the community may be important not only for the individuals concerned but also for the larger collectivity, according to one version of social capital theory. In this perspective, for any community, the higher the level of citizen engagement in civic life and in voluntary organizations such as sports clubs, the better the overall economic health of the community (e.g., Putnam, 1993). The jury is still out as to whether social capital measured at the individual level does indeed have effects at the community level (Portes, 2000). A similar argument at the organizational level also awaits testing. This argument suggests that individuals’ good citizenship behaviours in organizations helps create organizational social capital that in turn positively affects firm performance (Bolino et al., 2002).

We have focused on the positive effects of network ties and social capital, but we also know that maintaining relationships with people requires resources that some groups of people find it difficult to afford (Riley and Eckenrode, 1986). Problems with relationships can adversely affect people’s well-being (Rook, 1984). People in close relationships such as friendship have the opportunity to betray each other’s trust and to hurt each other in other ways (Granovetter, 1985). Increasing research attention is being directed towards the negative side of social interaction (e.g., Brass and Labianca, 1999; Yager, 2002).

Social relationships affect not only people’s well-being, but also their very identities. Adam Smith declared that ‘the countenance and behaviour of those [we live] with … is the only looking glass by which we can, in some measure, with the eyes of other people, scrutinize the propriety of our conduct’ (quoted in Bryson, 1945: 161). George Herbert Mead (1934: 171) summarized this perspective most succinctly in his remark that the individual only becomes a self ‘in so far as he can take the attitude of another and act toward himself as others act’.

Despite the apparently decisive effects that social contacts can have on the lives and well-being of individuals, much social science research has been silent concerning social influences. In the area of decision-making, for example, both the normative models, such as expected utility theory (e.g., Becker, 1976), and the descriptive models, such as prospect theory (Kahneman and Tversky, 1979), portray individuals making decisions in splendid isolation from the force-field of influences that surround them. As a survey of social network analysis pointed out: ‘In the atomistic perspectives typically assumed by economics and psychology, individual actors are depicted as making choices and acting without regard to the behavior of other actors’ (Knoke and Kuklinski, 1982: 9). In the field of organizational behaviour, the dominance of atomistic approaches such as expectancy theory has contributed to the neglect of social influences.

The neglect of social context has affected not only the more individualistic social sciences of economics and psychology, but also the more structural approaches such as sociology (including organizational sociology). Many sociologists continue to study categories that are assumed to share similar characteristics (Wellman, 1988a: 15). These categories, such as ‘managers’, ‘employed adults’, and ‘churchgoers’, contain aggregated sets of unrelated individuals. Much analysis consists of investigating whether individuals in one set, such as managers, are more likely to belong to another set, such as high-performers. At the organizational level of analysis, enquiries examine whether characteristics such as size and concentration of authority predict important outcomes such as market share. These analyses tell us little about how the structure of actors’ social worlds emerges, and how the structure of interactions affects outcomes.

Studies of organizational social networks have increased in recent years in response to this perceived neglect of social structure and interaction. Useful collections of articles have been published containing original research and thinking concerning social capital (Leenders and Gabbay, 1999) and network ties (Andrews and Knoke, 1999; Grandori, 1999; Nohria and Eccles, 1992). Several monographs have advanced our understanding of specialized topics such as structural holes (Burt, 1992), job-search networks (Granovetter, 1995) and inter-firm alliances (Nooteboom, 1999). Useful reviews of research have focused on networks at the intraorganizational level (Flap et al., 1998; Krackhardt and Brass, 1994; Raider and Krackhardt, 2002), the organizational level (Gulati et al., 2002) and the interorganizational level (Baker and Faulkner, 2002). The field of practitioner-oriented books includes recommendations concerning organizational architecture (e.g., Helgesen, 1995) and managerial relationship-building (e.g., Baker, 1994, 2000). And books of methods have proliferated (e.g., Degenne and Forse, 1999; Knoke and Kuklinski, 1982; Schensul et al., 1999; Scott, 2000; Wasserman and Faust, 1994).

With all of this existing literature, what is the motivation for the current book? Our intention is to provide a compact handbook that introduces major concepts, covers the rudiments of methods, explores major debates, directs attention to new theoretical directions, and presents a vigorous critique of some taken-for-granted assumptions. Our book is aimed at all of those who seek a lucid and lively treatment of social network approaches to organizational research, with a particular emphasis on the neglected area of interpersonal networks in organizations. We aspire to offer new insights to those familiar with network analysis, and to motivate those interested in pursuing network research to embark on journeys of discovery.

The potential application of the social network approach to organizations is, in our view, enormous. The full spectrum of organizational phenomena that network thinking can illuminate extends across levels from micro to macro, and includes topics typically covered in fields such as organizational cognition, organizational behaviour, organizational theory, and strategic management. Network research investigates relational processes and structures at many different levels of analysis. We organize our review of potential applications by unit of observation (the individual, the team, the organization, etc.) and within these units by level of analysis.

EXPLORING THE RANGE OF ORGANIZATIONAL SOCIAL NETWORK RESEARCH

---

Individual Level of Observation

Cognition

Cognitions concerning organizational networks matter. If your colleagues at work think that a prominent person in the organization is your friend, then your colleagues will tend to think of you as a high performer: the perceived friendship link to the prominent person will bathe you in reflected glory. What matters is the perception that you have the friendship tie – irrespective of whether there really is such a tie or not (Kilduff and Krackhardt, 1994). One of the most interesting areas of social network research in organizations concerns such network perceptions.

Where do these perceptions come from? Learning happens as a result of personal interactions between the individual and others: people learn who their friends are. But people also learn by observing others’ interactions, by noticing, for example, who is friendly with whom among not only their own circle of friends but also among those outside of this circle. Research on network learning shows that individuals expect network relationships to follow certain patterns. For example, people expect to see friendship relations between two individuals as reciprocated rather than as an unrequited flow of friendship from one person to the other (De Soto, 1960). People tend to bias their perceptions of network relations in organizations in the direction of such expectations (Krackhardt and Kilduff, 1999).

Thus each individual develops a more or less accurate map of the relationships between all the people in an organizational department or other social arena in which the individual is routinely involved. Individual cognitive maps of a social network can differ widely from one person to another for a variety of reasons that are still not well understood but that might include factors such as susceptibility to biased perceptions, differential opportunities to learn the social network, individual position in the organization, and so on. People who are more accurate in their perceptions may gain advantages in organizations. For example, one study showed that having an accurate perception of who goes to whom for advice in an organization significantly predicted how powerful the individual was perceived to be by others (Krackhardt, 1990). In our view, there is great potential for further work from a social network perspective at the cognitive level of analysis. We spend time in Chapter 4 outlining an emerging cognitive network theory and its implications for organizational analysis.

Relations between individuals

Networks exist not only as sets of cognitions inside the heads of individuals in organizations, but also as structures of constraint and opportunity negotiated and reinforced between interacting individuals. People tend to rely on others in their networks for help in making major decisions (Kilduff, 1990). Further, employees not only tend to interact with group members who are similar on distinctive attributes such as ethnicity and gender (Ibarra, 1992), but the lower the relative proportion of such group members in the organization, the higher the likelihood of within-group identification and friendship (Mehra et al., 1998).

Given the general preference people have for social interaction with others similar to themselves, there arise opportunities for those who bridge across social divides. People whose network connections allow them to act as go-betweens in organizations, connecting otherwise disconnected individuals and groups, tend to garner many benefits, including faster promotions (Podolny and Baron, 1997). One of the newest areas of research concerns the ways in which people of different personality types tend to build distinctively different types of network connection (with respect to spanning across social divides, for example). We outline an emerging personality approach to social structure and discuss its relevance for organizational network research in Chapter 4. Potential applications of interpersonal network research include such standard organizational topics as power, job design, motivation and leadership (Krackhardt and Brass, 1994). Much of this work remains to be done.

Dyads, triads and cliques

There is unrealized potential for looking at two- and three-person units within network structures. Some recent work has taken network data collected at the individual level and used it to examine the ways in which a certain type of two-person unit (or ‘dyad’) experiences organizational life. The basic idea is that pairs of friends who have friends in common (compared to pairs of friends who have no friends in common) are likely to find themselves constrained in their attitudes and behaviours. For example, if Stacy and Kay have a disagreement and become angry with each other concerning some organizational decision, they are likely to repair their wounded feelings if they have mutual friends who can intercede, and whose relationships would also be disrupted by any breach between Stacy and Kay. This emerging stream of work suggests looking carefully at the network contexts in which pairs of individuals are located. (See Krackhardt, 1998, 1999; Krackhardt and Kilduff, 2002, for more on this.)

A three-person group is quite different from a two-person group in that coalitions, mediation, and a host of other sociological processes become possible (see Fernandez and Gould, 1994, for a recent treatment of this topic). Three-person groups (known as ‘triads’) have long been considered the building blocks of informal networks (Holland and Leinhardt, 1977) but have been relatively neglected in organizational network research (but see Krackhardt and Kilduff, 1999, for recent work).

Similarly neglected in organizational research (despite a rich tradition of research in sociology) have been cliques. A clique consists of people who all interact with each other but have no common links to anyone else. Cliques may form on the basis of shared demographic characteristics that are relatively rare in a particular organizational setting (Mehra et al., 1998). The effects of cliques on individuals in the clique (in terms of individual-level issues such as motivation and work performance) and on organizational functioning (in terms of organizational-level issues such as knowledge gathering and dissemination) are still relatively little studied, however.

Business Unit Level of Observation

As an organization expands into a heterogeneous environment (such as a different country), it is likely to establish a business unit focused on the complexities of that environment. Such business units, despite remaining part of the overall multidivisional enterprise, may achieve semi-autonomous status. Formal mechanisms exist to coordinate the activities of such business units (see Thompson, 1967), but social networks are just as likely to play a role here as they are at the interpersonal level. The study of the social networks that spring up between busine...

Table of contents

• Cover Page
• Title
• Copyright
• Contents
• Acknowledgements
• 1 Introduction
• 2 Understanding Social Network Research
• 3 Is There Social Network Theory? A Critical Examination of Theoretical Foundations
• 4 In Pursuit of Lost Questions: Bridging the Gap between Structuralist and Individualist Approaches to Social Networks
• 5 Network Trajectories: Goal-directed and Serendipitous Processes
• 6 Towards a Poststructuralist Network Approach to Organizations
• 7 Conclusion
• Glossary of Technical Terms
• Appendix 1 Getting Started on Data Analysis and Interpretation
• Appendix 2 Analysing Cognitive Network Data Using MRQAP – A Methodological Note
• References
• Index