the ties that lead a social network approach tto leadership, Naukowe, Sieci w organizacji
[ Pobierz całość w formacie PDF ]
The Leadership Quarterly 17 (2006) 419
–
439
The ties that lead: A social network approach to leadership
⁎
, Martin Kilduff
b,1
Prasad Balkundi
a,
a
Department of Organizational Behavior and Human Resources, 274 Jacobs Management Center, University of New York at Buffalo,
Buffalo, NY 14260, United States
b
Department of Management and Organizations, Smeal College of Business, The Pennsylvania State University, University Park,
PA 16802, United States
Abstract
This article investigates, for leadership research, the implications of new directions in social network theory that emphasize
networks as both cognitive structures in the minds of organizational members and opportunity structures that facilitate and
constrain action. We introduce the four core ideas at the heart of the network research program: the importance of relations, actors'
embeddedness, the social utility of connections, and the structural patterning of social life. Then we present a theoretical model of
how network cognitions in the minds of leaders affect three types of networks: the direct ties surrounding leaders, the pattern of
direct and indirect ties within which leaders are embedded in the whole organization and the interorganizational linkages formed by
leaders as representatives of organizations. We suggest that these patterns of ties can contribute to leader effectiveness.
© 2006 Elsevier Inc. All rights reserved.
Keywords: Social networks; Leaders; Centrality
Good administrators sometimes fail to understand social structure, and fail to anticipate its consequences for
organizational survival. This can leave organizations vulnerable to manipulation by skilled political entrepreneurs. In
one example, the entire top management team of a manufacturing company learned from a network analysis that the
bomb threats, shootings, and vandalism threatening the future of the company were instigated by partisans of a lower-
ranking manager, who had had systematically recruited family, friends and neighbors into the company over a 30-year
period. In a district desperate for jobs, these partisans felt loyalty to the informal leader who had provided them
information that allowed them to be first in line for vacancies on Monday morning. The CEO, confronted with an
analysis of the deep cleavages existing in the social structure of the organization resulting from the informal patterns of
recruiti ng over decades , had this to say about those who had been hired :
“…
they just seem ed like waves of turt les
coming over the hill; hired as they made it to our door
(
Burt, 1992
: 1).
This story illustrates the gap at the heart of our understanding of leadership. It illustrates how important it is for
would-be leaders (who may or may not hold formal supervisory positions
”
see
Bedeian & Hunt, in press
) to accurately
perceive the network relations that connect people, and to actively manage these network relations. This story also
—
DOI of original article: 10.1016/j.leaqua.2005.09.004.
⁎
Corresponding author. Tel.: +1 716 645 3250.
E-mail addresses:
(P. Balkundi),
(M. Kilduff).
1
Tel.: +1 814 865 1789.
1048-9843/$ - see front matter © 2006 Elsevier Inc. All rights reserved.
420
P. Balkundi, M. Kilduff / The Leadership Quarterly 17 (2006) 419
–
439
illustrates how informal leaders who may lack formal authority can emerge to frustrate organizational functioning
through the manipulation of network structures and the exercise of social influence. The perception of and the
management of social networks are intrinsic to the leadership role as we define it in this article, but reviews of the social
network literature frequently point out that
”
(
Brass, Galaskiewicz, Greve, & Tsai, 2004
, p. 800). Our goal is to investigate, for leadership research, the implications
of new directions in network theory that emphasize networks as both cognitive structures in the minds of organizational
members and opportunity structures that facilitate and constrain action.
In making the link from social networks to leadership, we borrow freely from the entire corpus of social network
theory (see
Kilduff & Tsai, 2003
, for a recent review) and from two perspectives in recent leadership theory: the
cognitive revolution in leadership research (
Lord & Emrich, 2001
) and the theoretical and empirical work that seeks to
extend the leader
“
little empirical work has been done on leadership and social networks
member exchange (LMX) perspective (e.g.,
Graen & Uhl-Bien, 1995; Liden, Sparrowe, & Wayne,
1997; Sparrowe & Liden, 1997, in press; Uhl-Bien, Graen, & Scandura, 2000
). The cognitive approach to leadership
draws attention to the fundamental importance of cognitive structures, such as schemas, in shaping leadership attitudes
and behaviors, whereas the LMX approach draws attention to the importance of relations, particularly the dyadic link
between a formally designated managerial leader and either a subordinate or, more rarely, a higher-level leader (see
Graen & Scandura, 1987
, for a review). We build a model that emphasizes, from a network perspective, how the
cognitions in the mind of the individual influence the network relationships negotiated by the individual, and how this
individual network affects leadership effectiveness both directly and through informal networks, both within
organizations and across organizations. Thus, we link together social cognitions and social structure to forge a
distinctive network approach to leadership that builds upon, but extends, previous work in both the network and the
leadership realms.
–
1. Organizational network research core ideas
The organizational network perspective is a broad-based research program that continually draws inspiration from a
set of distinctive ideas to investigate new empirical phenomena. The
ideas at the heart of network research
define its special character and distinguish it from rival research programs (cf.
Lakatos, 1970
). What are these ideas
familiar to all organizational network researchers? At least four interrelated principles generate network theories and
hypotheses: the importance of relations between organizational actors, actors' embeddedness in social fields, the social
utility of network connections, and the structural patterning of social life (
Kilduff, Tsai, & Hanke, 2005
).
An emphasis on relations between actors is the most important distinguishing feature of the network research
program. As a recent historical treatment of social network research (
Freeman, 2004
, p. 16) pointed out, a core belief
underlying modern social network analysis is the importance of understanding the interactions between actors (rather
than a focus exclusively on the attributes of actors). An early treatment of network research on organizations stated that
“
“
hard-core
”
the social network approach views organizations in society as a system of objects (e.g., people, groups, organizations)
joined by a variety of relationships
(
Tichy, Tushman, & Fombrum, 1979
: 507), whereas a more recent survey
represented organizational network research as a movement
”
away from individualist, essentialist and atomistic
explanations toward more relational, contextual and systemic understandings
“
(
Borgatti & Foster, 2003
: 991). The
importance of understanding relationships as constitutive of human nature was stated as follows in a recent book:
“
”
Human beings are by their very nature gregarious creatures, for whom relationships are defining elements of their
identities and creativeness. The study of such relationships is therefore the study of human nature itself
(
Kilduff &
Tsai, 2003
: 131). Our network approach locates leadership not in the attributes of individuals but in the relationships
connecting individuals.
The second principle that gives organizational network research its distinctiveness as a research program is the
emphasis on embeddedness. For organizational network researchers, human behavior is seen as embedded in networks
of interpersonal relationships (
Granovetter, 1985; Uzzi, 1996
). People in organizations and as representatives of
organizations tend to enter exchange relationships, not with complete strangers, but with family, friends, or
acquaintances. Embeddedness at the system level can refer to a preference for interacting with those within the
community rather than those outside the community. We emphasize that people's perceptions of others as leaders are
reflected through the sets of embedded ties within which people are located.
The third driving principle of social network research is the belief that network connections constitute social capital
that provides value
”
—
including economic returns (
Burt, 2000
). As a previous review of network research on leadership
P. Balkundi, M. Kilduff / The Leadership Quarterly 17 (2006) 419
–
439
421
pointed out,
(
Brass & Krackhardt, 1999
: 180). Depending
upon the arrangement of social connections surrounding an actor, more or less value can be extracted (
Burt, 1992;
Gnyawali &Madhavan, 2001
). At the system level, a generalized civic spirit emerges from and contributes to the many
interactions of trust and interdependence between individual actors within the system (
Coleman, 1990; Portes, 2000
).
Leadership, from the network perspective we develop, involves building and using social capital.
The fourth leading idea distinctive to the social network research program
“
Social capital is at the heart of social network analysis
”
–
the emphasis on structural patterning
–
often leads social network research to be referred to as the
“
structural approach.
”
Network researchers look for the
patterns of
in social systems (
Wellman, 1988
: 26). Not content with merely describing the
surface pattern of ties, researchers look for the underlying structural factors through which actors generate and re-create
network ties. At the local level surrounding a particular actor, the structure of ties can be described, for example, as
relatively closed (actors tend to be connected to each other) or open (actors tend to be disconnected from each other)
(
Burt, 1992
). At the system level, organizational networks can be assessed for the degree of clustering they exhibit and
the extent to which any two actors can reach each other through a short number of network connections (e.g.,
Kogut &
Walker, 2001
). To understand who is a leader from a network perspective is to investigate the social-structural positions
occupied by particular individuals in the social system.
These four leading ideas
“
connectivity and cleavage
”
the importance of relationships, the principle of embeddedness, the social utility of
network connections, and the emphasis on structural patterning
–
provide the common culture for organizational
network research that allows the diversity of viewpoints from which fresh theoretical initiatives emerge (cf.
Burns &
Stalker, 1961
: 119). Network research is also characterized by vigorous development of methods and analytical
programs to facilitate the examination of phenomena highlighted by theory (see
Wasserman & Faust, 1994
for a review
of methods; and the UCINET suite of programs
–
–
–
for statistical software). Our
review in this article will touch on these methodological developments where appropriate.
The organizational network research program is progressive in the sense that new theory is constantly being
developed from the metaphysical core of ideas that makes up the heart of the research program, highlighting new areas
of application. It is the purpose of this review to highlight the area of leadership from a network perspective. The four
leading ideas that comprise the intellectual source of theory development for organizational network research are best
understood as mutually reinforcing core beliefs that, like the planks of a ship, keep the research program afloat
Borgatti, Everett, & Freeman, 2002
in
terms of new theory development and exploration of new phenomena. At the level of network theory and research, all
four ideas tend to be inextricably involved. We will invoke these ideas as appropriate throughout this review.
In contrast to network research, traditional leadership research has focused on human capital attributes of leaders
and situational attributes of leadership contexts. Human capital attributes of leaders include traits (e.g.,
House, 1977;
Kenny & Zaccaro, 1983
) and behavioral styles (e.g.,
Lewin, Lippitt, &White, 1939; Podsakoff, Todor, & Skov, 1982
);
whereas situational attributes of leadership contexts include task structure (
Fiedler, 1971
), the availability of leadership
substitutes (
Kerr & Jermier, 1978
), the nature of the decision process (
Vroom & Yetton, 1973
) and the quality of
leader
—
member exchange (
Dansereau, Graen, & Haga, 1975; Graen, Novak, & Sommerkamp, 1982
). A social network
perspective does not eclipse the valuable results of conventional leadership research; rather, a network perspective can
complement existing work without repeating it. In particular, in this review we amplify the voices that have called for a
new understanding of leadership effectiveness to include leaders' cognitions about networks and the actual structure of
leaders' ties (e.g.,
Hooijberg, Hunt, & Dodge, 1997
; see also
Bass, 1990
: 19).
As with all theoretical perspectives the network approach has boundary conditions that limit its range of application.
Social network processes are less likely to have the effects we discuss to the extent that organizations are characterized
by perfect competition between equally informed actors all of whom have the same opportunities (see the discussion in
Burt, 1992
). (Even under conditions of perfect information, however, some actors are likely to be more influenced by
social networks than others
–
—
see
Kilduff, 1992
.) A further limiting condition is the extent of work interdependence:
under conditions of low interdependence between actors and little or no social interaction, network processes and their
effects will tend to be minimized.
In network terms, leadership embodies the four principles that we articulated earlier. Leadership can be understood
as social capital that collects around certain individuals
based on the
acuity of their social perceptions and the structure of their social ties (cf.
Pastor, Meindl, & Mayo, 2002
). Patterns of
informal leadership can complement or complicate the patterns of formal leadership in organizations. Individuals can
invest in social relations with others, can structure their social networks by adding and subtracting relationships, and
can reap rewards both in terms of their own personal performance and organizational unit performance (
Sparrowe,
–
whether formally designated as leaders or not
–
422
P. Balkundi, M. Kilduff / The Leadership Quarterly 17 (2006) 419
–
439
Liden, Wayne, & Kraimer, 2001
). But embeddedness in social networks always involves the paradox that social
relations, particularly those outside the immediate circle of the individual, may be difficult both to perceive accurately
and to manage (cf.
Uzzi, 1997
). Thus, although the social structure of the organization determines opportunities and
constraints for emergent leaders, the social structure is not within the control of any particular individual.
2. Leadership and the structure of ties
We start our network approach to leadership theory with a discussion of actor cognitions concerning networks,
move out to the inner circle around the actor and then further zoom out to include progressively more of the social
structure of the organization and the interorganizational realm (see
Ibarra, Kilduff, & Tsai, 2005
, for more on the
“
metaphor in relation to network research). The theoretical framework is illustrated in
Fig. 1
, and represents a
tentative model of leadership effectiveness from a network perspective. We provide an overview of the causal
connections of the model before zooming in to discuss in more detail the dynamics within each part of the model.
As
Fig. 1
shows, the first step in our conceptual model indicates that leaders' cognitions about social networks affect
zooming
”
the
79, for a
review) suggests that people in general shape their immediate social ties to others to be congruent with their schematic
expectations concerning how relationships such as friendship and influence should be structured. The schematic
expectations of leaders affect their ability to notice and change the structure of social ties (e.g.,
Janicik & Larrick,
2005
). Thus, cognitions in the mind of the leader are the starting point for our theorizing concerning the formation of
ties connecting the leader to others.
The network cognitions of leaders concerning such crucial organizational phenomena as the flow of social capital
within and across organizational boundaries, and the presence and meaning of social divides, are hypothesized to affect
the extent to which leaders occupy strategically important positions in the organizational network. An accurate
perception of the informal influence network can itself be a base of power in the organization (
Krackhardt, 1990
) and
can facilitate the leader's ability to forge successful coalitions (
Janicik & Larrick, 2005
). We extend these insights to
hypothesize that the acuity of leader cognitions will affect the extent to which a leader plays a strategically important
role in the relevant interorganizational network. We know of no research bearing on this thesis, although recent work
concerning interorganizational relationships increasingly concerns itself with hypothesized perceptual processes such
as organizational reputation and status (e.g.,
Podolny, 1998; Zuckerman, 1999
).
The extent to which a leader plays a role in these three actual networks
“
ego networks
”
that surround each leader. Cognitive network theory (see
Kilduff & Tsai, 2003
, pp. 70
–
–
the ego network, the organizational network
and the interorganizational network
is hypothesized to affect leader effectiveness. This critical hypothesis derives from
our basic understanding of how the four guiding principles of the network approach extend leadership theory. Modern
concepts of leadership identify the relational content of the interaction between people as the key aspect involved in the
structuring of situations and the altering of perceptions and expectations (e.g.,
Bass, 1990
, p. 19). Modern network
theory suggests that individuals who are central in the immediate networks around them and in the larger networks that
connect them to others throughout the organization and beyond the organization are likely to acquire a particular type
–
Fig. 1. Theoretical framework linking leader's network accuracy to leader-relevant outcomes.
P. Balkundi, M. Kilduff / The Leadership Quarterly 17 (2006) 419
–
439
423
of expert power-knowledge of and access to those few powerful others whose words and deeds control resource flows
and business opportunities (e.g.,
Burt, 2005
). Leaders may not be able to move into the center of every important
network, of course. Embeddedness in one social network may come at the price of marginality in another network.
There are trade-offs involved in building social capital, particularly when brokerage across social divides may engender
distrust rather than gains.
One blow-by-blow account of an organizational power struggle contrasted the networking strategies of two
combatants for sole control of the CEO position they currently shared. Whereas co-CEO Louis Glucksman was central
within the Lehman Brothers organization as a whole and occupied a particularly strategic position among the traders,
his rival and co-CEO Pete Petersen neglected internal networking in pursuit of connections with the leaders of other
organizations (
Auletta, 1986
). Both men were effective leaders
Glucksman contributing to internal effectiveness and
Petersen building and maintaining the external relationships that brought contracts to the partnership. But both had
built quite different social network bases of power.
The role of external affective ties with the representatives of other organizations in providing vital help to companies
in financial trouble has been emphasized by research on the survival prospects of small firms in the New York garment
industry (
Uzzi, 1996
). More generally, the organizational theory and strategy literatures have examined the extent to
which ties between organizations constitute a knowledge base important for outcomes such as firm growth (e.g.,
Powell, Koput & Smith-Doerr, 1996
), new ties (e.g.,
Gulati & Gargiulo, 1999; Larson, 1992
), and innovation
(
Hargadon & Sutton, 1997
). Thus, the extent to which leaders are effective in terms of accessing important resources is
likely to depend on the social-structural positions they occupy in the key networks within and between organizations.
What are the outcomes associated with leader effectiveness from a social network perspective? We have mentioned
above such aspects of leader effectiveness as organizational growth, survival, and innovation. These are the
responsibility of formal leaders and are outcomes at the organizational level of analysis. As
Fig. 1
summarizes, leader
effectiveness from the network perspective we articulate would also include such components of internal
organizational functioning as coalition building, mentoring, and brokering. These are intrinsically networking
outcomes of both formal and informal leadership that can enhance coordination across functions within the
organization. We return to these internal measures of leader effectiveness later in the paper.
The model outlined in
Fig. 1
necessarily simplifies the relationships between cognition, social networks, and
leadership effectiveness. We neglect, for example, the ways in which occupancy of social structural positions in
networks affects individuals' cognitions and expectations about networks (see
Ibarra et al., 2005
, for a review). The
organization and the environment within which it operates can be jointly considered a set of cyclical processes captured
in networks of cognitions (cf.
Bougon, Binkhorst, &Weick, 1977
). We focus in this article on leadership, and therefore
emphasize the proactive enactment of outcomes leading to leader effectiveness.
—
3. Network cognition and leadership
A key discovery of modern social network research is that cognitions matter (e.g.,
Kilduff & Krackhardt, 1994
), and
thus we start the in-depth discussion of the theoretical framework with an emphasis on network cognition, a topic
relatively neglected within conventional leadership research (but see early LMX work on whether peers within units
accurately perceive the quality of dyadic leader
Graen & Cashman, 1975
). Depending upon
how the boundary is drawn around a particular individual in an organization, that individual may or may not appear to
be influential in the eyes of others. That implicit leadership theories may be triggered by the structural position of
certain individuals in the eyes of others is a possibility hinted at in recent leadership theory (
Lord & Emrich, 2001
), but
yet to be systematically examined. From the perspective of perceivers located in small groups, certain actors may
appear influential, but perceivers surveying the larger context of the whole organization may dismiss these same actors
as relatively inconsequential (see the discussion in
Brass, 1992
). Conversely, people who seem relatively powerless
within one local group may be revealed to have close connections with powerful others outside the group. Thus, we
organize our discussion by progressively zooming out from individuals' network cognitions to include expanding
social circles within and beyond the organization.
From a network perspective that emphasizes the importance of relationships, embeddedness, social capital, and
social structure, the ability of formal or would-be informal leaders to implement any leadership strategy depends on the
accurate perception of how these principles operate in the social context of the organization. To be an effective leader of
a social unit is to be aware of: (a) the relations between actors in that unit; (b) the extent to which such relationships
–
subordinate relations
—
[ Pobierz całość w formacie PDF ]