Professional Documents
Culture Documents
This historical overview of leadership theory and research with an eye for commonali-
ties provides an opportunity for integration. Early unproductive research focused on
personality traits and behaviors. A recognition of the more complex nature of the
phenomenon resulted in the development of contingency theories that examined leader
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
characteristics and behavior in the context of situational parameters. The 1970s brought
This document is copyrighted by the American Psychological Association or one of its allied publishers.
For much of its history, leadership theory and integrative framework and a suggested direction
the empirical supporting research have been for future research.
regarded as a fractured and confusing set of In this analysis leadership is defined as "a
contradictory findings and assertions without process of social influence in which one person
coherence or interpretability. In this article T is able to enlist the aid and support of others in
argue that a considerable commonality of the accomplishment of a common task." This
well-accepted findings points the way toward a definition places the subject distinctly within the
successful and useful integration of current purview of social psychology, and the analyses
knowledge. That integration, based on the key that follow reveal how fully the leadership
functions performed by effective leaders, begins literature is indebted to the field of social
to answer the question of how good leaders psychology for its dominant paradigms and
behave and raises intriguing questions about the central variables.
personal characteristics of leaders that facilitate
those behaviors.
This article takes an historical perspective, Before Contingency Theory:
and the analysis is divided into four periods: (a) Lost in the Wilderness
the period prior to the presentation of Fiedler's
(1964) contingency model; (b) the period from Social philosophers have had a long-standing
1965 to 1975, focusing on the development and interest in both organizational and political
elaboration of contingency theories; (c) the leadership. Western European philosophers,
period from 1975 to 1985, when cognitive embedded in a strongly individualistic cultural
theories and concerns about gender differences milieu, looked primarily to the characteristics of
arose; and (d) the period since 1985, which has leaders for explanatory premises, For example,
most extensively focused on transformational Carlyle (1841/1907) proposed the great man
theories and cultural influences. The historical theory of leadership, which argued that success-
analysis is followed by a presentation of an ful leaders possessed traits of personality and
character that set them apart from ordinary
followers. The interest in individual characteris-
tics of leaders was spurred by the emergence of
Correspondence concerning this article should be ad- intelligence tests in the early 20th century.
dressed to Martin M. Chemers, Division of Social Sciences, Empirical psychology turned toward the study
117 Social Sciences I, University of California, 1156 High
Street, Santa Cruz, California 95064. Electronic mail may be of traits, and the nascent leadership field
sent to mchemers@zzyx.ucsc.edu. followed suit.
27
28 CHEMERS
the selected trait measures. Stogdill (1948) seemed to reflect leader intentions to support
This document is copyrighted by the American Psychological Association or one of its allied publishers.
provided an extensive review of 30 years of the positive group morale and follower satisfaction.
trait studies. He reported that a few traits (most A strong second factor, labeled Initiation of
notably intelligence) were sometimes associated Structure, included items measuring the leader's
with reliable differences between leaders and use of standard operating procedures, criticism
followers (i.e., about 35% of the time), but there of poor work, and emphasis on high levels of
was no single variable or even cluster of performance. These behaviors appeared to be
variables that was related to leadership across a related to a leader's focus on building a structure
variety of situations. Stogdill concluded that for task accomplishment.
although individual differences were certainly Although the LBDQ factors were found
important in identifying emergent or effective reliably in ratings of leader behavior across a
leaders, the great diversity of situations in which wide range of settings, they were less than
leaders functioned made it unlikely that any one completely successful at predicting the impor-
trait would be a universal predictor. Although it tant outcomes associated with leadership effec-
was not immediately recognized, StogduTs tiveness, that is, follower satisfaction and group
analysis set the stage for theories of leadership performance (Fleishmann & Harris, 1962; Kor-
that were predicated on an interaction between man, 1966). Considerate leadership was often
leader traits and situational contingencies. related to follower satisfaction or morale, and
Consideration and Initiation of Structure were
sometimes but not always predictive of group
Behaviors and Styles
performance. The failure of this carefully
Daunted by the failure of traits to predict constructed and comprehensively researched
leadership, but unwilling to abandon individual- behavioral measure to predict leadership led
istic explanations, researchers turned to the many researchers to throw up their hands in
study of leader behavior. Observations of the frustration and seemed to be yet another
effects of leadership style (i.e., autocratic vs. instance of leadership research leading to no
democratic) on the atmosphere of small groups coherent conclusions.
(Lewin, Lippitt, & White, 1939), process
analyses of interactions in laboratory discussion Legitimacy
tasks (Bales & Slater, 1955), and reports of
industrial workers on the behavioral styles of One of the brightest spots in the early
their supervisors (Kahn, 1951) sought to iden- empirical work on leadership was the series of
tify patterns of leader behavior associated with studies conducted by Hollander, which illumi-
high productivity or good morale. nated some of the facets of leadership status
The most extensive of these research pro- accrual and legitimacy. In both laboratory and
grams and the one with the most enduring field research venues, Hollander (1964; Hol-
impact on the field of leadership was the set of lander & Julian, 1970) found that individuals in
studies surrounding the development of the groups gain status through the demonstration of
Leader Behavior Description Questionnaire task-related competence and loyalty to group
(LBDQ) at Ohio State University (Hemphill, values. Status acquisition is associated with the
1950). This 150-item behavioral inventory was accrual of so-called "idiosyncrasy credits,"
used to collect ratings of military and industrial which can be thought of as units of group
SPECIAL ISSUE: LEADERSHIP RESEARCH AND THEORY 29
acceptance that can be "spent" to influence negative terms (hypothesized to reflect a greater
others and provide leeway from group norms to emphasis on interpersonal relationships).
allow for innovation in group processes and Apparently promising early work found that
views. task-oriented leaders were more effective, but
This pioneering work on how individuals subsequent studies yielded results showing that
decide to follow those in leadership positions relationship-oriented leaders had more effective
has retained its currency to the present day, in teams. Confused but not deterred by these
part because it embodies both cognitive and anomalous findings, Fiedler reanalyzed a large
behavioral elements in its approach. Contempo- number of studies—this time classifying the
rary information-processing theories of leader- group settings in terms of the degree of support
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ship posit leadership "prototypes" that are and cooperation offered by followers, the clarity
This document is copyrighted by the American Psychological Association or one of its allied publishers.
characterized by the elements of competency and structure of the group's task, and the
and trustworthiness that are the bases for the leader's formal authority to direct and reward
accrual of idiosyncrasy credits. Recent work by followers. These three variables were combined
Hogg and his associates (Hogg, Hains, & into a dimension of "situational favorableness"
Mason, 1998) applying social identity theory to (Fiedler, 1967) or "situational control" (Fiedler,
leadership perception indicates that, despite a Chemers, & Maher, 1976), thought to reflect the
tendency for followers to value leaders who degree to which the overall situation gave the
embody group values (the basis for perceptions leader a feeling of certainty, predictability, and
of trustworthiness in Hollander's model), they control over group processes.
also heavily weight task relevant competence in When the leader's orientation (i.e., LPC
leadership evaluation. Those basic determinants score) was correlated with group performance
of leadership status turn up in more recent across the dimension of situational favorable-
approaches to understanding leadership judg- ness, a reliable relationship was found. Specifi-
ments. A fuller discussion of the role of cally, groups led by task-oriented leaders
perception in leadership process appears later. performed best in situations of high control and
predictability or very low control and predictabil-
ity, and groups led by relationship-oriented
The Mid-1960s to the Mid-1970s: leaders performed best in the situations of
The Contingency Era moderate control or predictability. The explana-
tory rationale for these findings was that the
The Contingency Model relatively more directive, task-focused leader-
ship style is most appropriate when an orderly
The study of leadership took a dramatic situation provides the leader with the clarity to
change of direction with the publication of give directions and the follower support to be
Fiedler's first articles (1964) and subsequent sure of his or her performance, and the highly
book (1967), which presented a new approach to volatile and unpredictable environment of the
understanding leadership effectiveness. The very low control situation also requires the
contingency model of leadership effectiveness steadying influence of clear directions and
emerged as an answer to StogduTs (1948) call structuring leader behavior. However, the more
for an approach based on the interaction of interpersonally oriented, participative style of
leader traits with situational parameters, but it leadership was thought to function most effec-
did not start out that way. Early work (Cleven & tively when the complexities of a moderate-
Fiedler, 1956; Fiedler, 1955, 1958) tested the control situation required greater delicacy to
predictive validity of a leadership trait measure navigate a poorly understood task or to avoid the
on the basis of the leader's views of coworkers. dangers associated with uncertain follower
The measure, which eventually came to be support.
known as the Least Preferred Coworker (LPC) The inductive method by which the contin-
scale, differentiated leaders who viewed poorly gency model was constructed and the highly
performing coworkers in very negative terms complex nature of its predictions led to many
(thought to reveal a very strong concern with criticisms of the model during the 1970s
effective task performance) from those who (Ashour, 1973; Graen, Alvarez, Orris, & Mar-
viewed poorly performing coworkers in less tella, 1970). However, subsequent research and
30 CHEMERS
extensive meta-analyses (Peters, Hartke, & generally supportive of its basic premises (Field
Pohlmann, 1983; Strube & Garcia, 1981) & House, 1990).
provided strong support for the basic principles The contingency model and normative deci-
of the model. (For a more complete discussion sion theory have many features in common.
of the development of the contingency model They are both focused on the leader as the
and the controversy surrounding its validity, see central actor in the group's efforts to interface
Chemers, 1997.) Another criticism of the with the task environment. Both theories regard
contingency model was its apparent assumption the leader's task as to gain the group's support in
that a leader could not choose to be both task solving problems and implementing solutions
and relationship oriented when the situation effectively. Also, they both hypothesize that
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
demanded it. Some other contingency ap- more directive approaches will be most effective
This document is copyrighted by the American Psychological Association or one of its allied publishers.
proaches did not make that same assumption. when a clear task and a supportive group give
the leader the certainty to take charge but that
more participative strategies will work better
Normative Decision Theory when a less clear and orderly environment
argues against bold action and autocratic direc-
Energized by the potential of a contingency tion. The two theories part company in the
approach to explain leadership performance, but situation of very low control with the contin-
working from a more deductive theoretical base, gency model more focused on immediate group
Vroom and Yetton (1973) offered a model of performance through leader direct action but
decision-making effectiveness that integrated normative decision theory suggesting more
leaders' decision strategy with situational fac- participative strategies to build a more support-
tors. Leaders were conceived as having a range ive environment over the long run.
of decision-making strategies available to them
that varied in degree of follower involvement in
the process—ranging from autocratic styles Path-Goal Theory
(leader makes the decision with minimal fol- Contingency theories held the promise of
lower input) to consultative styles (leader makes correcting the weaknesses of earlier approaches
the decision after getting follower opinion and to leadership effectiveness prediction. For ex-
advice) to group or participative styles (leader ample, one promising but disappointing ap-
and group make decision together, with equal proach had been the attempt to relate leader
weight). behavior (e.g., the LBDQ score) to organiza-
The situational parameters included in the tional outcomes. R. J. House and his associates
model were represented as a series of questions (R. J. House, 1971; R. J. House & Dessler, 1974;
arranged to yield a decision tree. Leaders R. J. House & Mitchell, 1974) picked up that
seeking the most effective decision strategy gauntlet and attempted to merge traditional
were asked to analyze situational factors that behavioral approaches with emerging develop-
included the clarity and structure of the task and ments in the study of worker motivation to
surrounding information, the degree of support understand the impact of the leader on the
for the leader and the organization among the motivation and performance of followers.
followers, the degree of conflict among subordi- Path-goal theory argues that the leader's
nates, and the time urgency for a decision to be main purpose is to motivate subordinates by
made. The model specifies that when the task is helping mem to see how their task-related
clear and the followers supportive, the leader performance could help them to achieve their
should use the more time-efficient autocratic personal goals. Research within the path-goal
styles. If the task or information is unclear, using framework attempted to understand how a
the consultative strategies increases the informa- leader's directiveness (i.e., Initiation of Struc-
tion yield and likelihood of a higher quality ture) or supportiveness (i.e., Consideration)
decision. When the leader lacks follower behaviors might affect subordinate motivation
support, the participative strategy helps to and performance. Proceeding logically, the
ensure follower commitment to the decision and theory predicted that a leader's structuring
its implementation. Empirical research on the behavior would be motivating to a subordinate
normative decision model is not extensive but is when the subordinate's task environment lacked
SPECIAL ISSUE: LEADERSHIP RESEARCH AND THEORY 31
structure because of insufficient training or higher motivation and that both characteristics
experience or a highly complex task. However, of the task and of the subordinate will contribute
when a subordinate had sufficient structure, to that receptiveness.
leader directiveness would be regarded as overly In an interesting extension of path-goal
close monitoring or "pushing" and would have theory, Kerr and Jermier (1978) argued that if
negative effects. Consideration behavior was the leader's purpose is to supply missing
seen to have its most positive effects when the elements in the subordinate's job environment
subordinate needed psychological or emotional (e.g., structure or support), then other sources of
support to deal with an aversive work environ- those missing elements might make the leader's
ment (made so by a boring or unpleasant task). behavior redundant and unnecessary. Their
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
situations that were engaging and intrinsically for example, that if a job provided plenty of
interesting to the subordinate. task-relevant feedback, leader structuring behav-
The typical research paradigm for path-goal ior would be unnecessary, or if a compatible and
theory studies was to divide a group of cohesive work group provided emotional sup-
subordinates into situations of low clarity port, leader consideration would be redundant.
(presumably interesting, but potentially frustrat- Under such conditions, leader behaviors were
ing because of lack of structure) and of very hypothesized to show minimal or even negative
high clarity-predictability (presumably boring relationships with subordinate motivation, satis-
and uninvolving). Leader structuring behavior faction, or performance. However, a review of a
was predicted to have positive effects on number of studies of substitutes-for-leadership
subordinate motivation and performance in the hypotheses indicated very little support for the
former situation but not the latter, whereas the theory's predictions (Podsakoff, Niehoff, Mac-
reverse was true for leader considerate behavior. Kenzie, & Williams, 1993) and revealed that a
Path-goal theory generated a considerable body leader's behavior remains very important to
of empirical research support for the basic subordinates regardless of varying situational
propositions. Considerate behavior, for ex- conditions.
ample, was usually related to positive subordi- The research literature on the contingency
nate attitudes under boring or aversive task theories suggests that actions by a group's leader
situations but often had similarly positive effects can have strong effects on the motivational and
across all situations. Results regarding structur- emotional states of followers and on the
ing behavior were even less consistent. successful accomplishment of the group's task.
A study by Griffin (1981) that included The relationship of the specific leader actions to
measures of subordinates * ' * growth need those outcomes depends on the interaction of
strength" (Hackman & Oldham, 1976) indicated those actions with relevant features of the
that the needs and expectations of subordinates interpersonal and task environment.
played a role in determining when subordinates
needed or wanted different types of leader The Mid-1970s to the Mid-1980s:
behavior. Growth-oriented, challenge-seeking Cognitive Models and Gender Concerns
subordinates were quite comfortable with un-
structured and challenging tasks and were The growing influence of cognitive theories
therefore less receptive to directive leader in social psychology led to a similar interest
behavior under unstructured or structured condi- among leadership researchers. Two broad classes
tions. However, these growth-oriented subordi- of investigation were concerned with percep-
nates were very responsive to supportive tions of leaders by others (i.e., followers,
behavior when the task was boring. The more superiors, and observers) and leaders' percep-
change-averse, low-growth-need subordinates tions and evaluations of subordinates.
were comfortable with leader structure across all
situations but needed less support when a task Leadership Perceptions
was ostensibly boring. A reasonable conclusion
to be drawn from this literature is that leader In the mid-1970s, studies involving ratings of
behavior that is seen as supportive by subordi- leader behavior began to reveal certain anoma-
nates is likely to lead to positive reactions and lous findings. Eden and Leviatan (1975) re-
32 CHEMERS
ported that when research participants were processes. Recognition processes determined
asked to make ratings of leader behavior by when an individual's behavior would result in
simply imagining a leader, the resultant data the perception of that person as a leader.
showed factor structures similar to those derived Observers were found to hold highly articulated
from ratings of actual leaders. Staw (1975) prototypes (Cantor & Mischel, 1979; Rosch,
showed two sets of observers the same video- 1978) of leadership. When an actor's behavior
tape of a group interaction but told the observers showed sufficient overlap with the prototypi-
that the group had either been very successful or cally driven expectations of observers, a leader-
very unsuccessful on task performance. Ob- ship attribution was made. Once an individual
server ratings of the "successful" leader were was seen as a leader, selective attention and
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
supportive leadership than were the ratings of Leadership judgments were also found to be
the "unsuccessful" leader. influenced by inferential processes. Because
That ratings of leaders might be strongly implicit theories of leadership associate team
biased created a problem on both theoretical and success with effective leadership, observers are
methodological grounds. Leader legitimacy, a likely to infer the presence of good leadership
central construct in understanding the bases of from evidence of group success (Phillips &
leader influence, was based on follower percep- Lord, 1981). Thus, once a person is seen as a
tions. Furthermore, almost every research para- leader, observer inferences are likely to rein-
digm in the leadership field depended on ratings force and enhance that perception. Of course, if
of leader behavior (Rush, Thomas, & Lord, a person's characteristics, such as gender or
1977). race, are inconsistent with observers' prototypic
Attribution theory (Jones & Davis, 1965; expectations, then such a person is less likely to
Kelley, 1967) provided a theoretical framework be perceived as a legitimate or effective leader
for the investigation of leadership biases. Taking despite any objective achievements.
a very strong position, Calder (1977) argued that The strength of common beliefs in the
the very concept of leadership is rooted in importance of leadership for group outcomes
popular language and poorly articulated as a led Meindl (1990) to develop the "romance
scientific construct. He argued that with no way of leadership" concept. In an ingenious series
of measuring leadership apart from social of experiments and naturalistic observations,
perceptions, leadership exists primarily as an Meindl showed that any remarkable group or
attribution rather than a testable construct and organizational outcome, whether highly positive
should, therefore, be abandoned as a subject of or highly negative, is likely to be attributed to
scientific inquiry. Few researchers were willing leadership effects, while other reasonable causes
"to throw the baby out with the bathwater" and are largely ignored.
began instead to make a systematic study of Although the strong susceptibility to percep-
leadership perceptions and the processes that tual biases in the observations of leadership
gave rise to them. might have constituted a problem for research
A useful model was provided by the research methodology, it opened a fascinating area for
on implicit personality theories, which Hastorf, theoretical development. If leadership is a
Schneider, and Polefka (1970) defined as a process of social influence, then factors that
structure of association about what traits or affect the legitimacy, credibility, and influence
characteristics are related that guides and of leaders become a central aspect of leadership
organizes perceptions, thoughts, and memories function.
about a phenomenon. Implicit theories of Another important component of the leader-
leadership, then, would define the assumptions ship process involves the perception of follow-
that people held about what behaviors leaders ers by leaders. Almost every theory of leader-
displayed and how those behaviors were associ- ship posits that a central function of leadership
ated with group and organizational outcomes. involves the direction of subordinates. The
An extensive research program by Lord and follower-oriented contingency theories, such as
his associates (Lord, 1985; Lord, Binning, Rush, path-goal theory, maintain that it is the leader's
& Thomas, 1978; Lord & Maher, 1991) revealed responsibility to provide the subordinate with
that leadership attributions were based on two task-directed guidance or emotional support to
SPECIAL ISSUE: LEADERSHIP RESEARCH AND THEORY 33
help them to be effective and satisfied. Implicit the work of the subordinate. Subordinate failure
in these premises is the expectation that leaders might be evidence of leadership failure. These
are able to judge what kinds of behaviors on factors strengthen the tendency of leaders to
their part are likely to have positive effects on make ego-defensive attributions, blaming subor-
subordinates. In other words, leaders must dinates for poor performance and possibly
observe the actions and reactions of subordi- taking personal credit for group success. Inaccu-
nates to judge what is needed. This clearly rate judgments that arise from these blind spots
places attributionai processes at the center of the can easily erode the working relationship and
relationship between leader and follower. bases of influence between the leader and
Mitchell and his associates (Green & Mitchell, follower.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
former to stereotypic perceptions about the negative expectations that bias women's oppor-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
characteristics of a manager. As late as 1989 tunities for achieving leadership roles and being
Heilman, Block, Martell, and Simon replicated fairly evaluated in those roles. Research on the
Schein's (1973) study and found little change in structural approach supports this view.
these stereotypes. Clearly, then, the common J. House (1981) argued that actors in a social
view was and may still be that women and men structure are often strongly influenced by their
are very different in their leadership style and place in that structure. In a series of studies on
performance. How good is the evidence? women and power in organizations, Ragins
In a classic treatise on the subject of gender (1989,1991; Ragins & Sundstrom, 1989) found
differences, Deaux (1984) effectively dismissed that women face a number of barriers along the
any biological bases for gender differences in path to acquisition of status and power in
social behavior. The evidence simply does not organizations. However, when male and female
support such differences. But what about the managers are matched for level (i.e., power and
possibility that differences in socialization to authority) within the organization almost no
gender roles carry over to behavior in the differences are found in leader behavior, perfor-
workplace, so-called "gender role spillover" mance, or acceptance by subordinates.
(Nieva&Gutek, 1981)? The conclusion that may be drawn from this
In the second edition of the Handbook of literature is that although few real differences in
Leadership, Bass (1981) reported that the leadership behavior or style exist between men
empirical evidence available at that time showed and women, false but persistent stereotypes
no consistent pattern of differences between impede equal access and fair evaluation for
men and women in supervisory style. However, women in organizational leadership.
definitive analysis on this topic waited until
Eagly and her associates conducted a series of The Mid-1980s to the Mid-1990s:
meta-analyses on male-female differences in Transformational Leadership and Cultural
leadership style (Eagly & Johnson, 1990), Awareness Transformational Theories
leadership emergence (Eagly & Karau, 1991),
and leadership evaluation (Eagly, Makhijani, & A major shift of interest in leadership research
Klonsky, 1992). A careful reading of these was sparked by the work of a political historian.
analyses suggests that to the extent that the Burns's (1978) book on great leaders differenti-
observations of leadership were taken in organi- ated transactional leaders, whose relationship to
zational settings; using standard behavioral followers was based on mutually beneficial
measures; by observers, superiors, or subordi- transactions, from transformational leaders,
nates, the differences found between men and who influence followers to transcend personal
women are so small as to be of little practical interests and transform themselves into agents
significance. Women tend to emerge as leaders of collective achievement. This was an exciting
about as often as men, and they tend to be perspective for a field locked in molecular
evaluated similarly to men when all other analyses of trait-situation interactions and
variables are equal. perceptual biases.
How about when other variables are not Anticipating this development by a year, R. J.
equal? Women tend to emerge less frequently House (1977) published a theoretical analysis of
and are evaluated less positively in situations charismatic leadership in which he analyzed the
where followers are hostile to women in characteristics of historical leaders who elicited
SPECIAL ISSUE; LEADERSHIP RESEARCH AND THEORY 35
extraordinary levels of devotion and commit- nates are associated with high-performing teams
ment from followers—for example, Gandhi, and organizations.
Martin Luther King, Jr., and others. House House and Shamir (1993) returned to the
identified three sets of characteristics that study of transformational leadership, emphasiz-
typified charismatic leaders. Personal character- ing the psychological processes of followers that
istics included a strong belief in the moral mediated the effects of charismatic or transfor-
righteousness of one's beliefs, high levels of selfmational leader actions. Weaving together path-
confidence, and a strong need to influence and goal theory (with its emphasis on expectancy
dominate others. Behaviors included dramatic motivation) with theories of intrinsic motivation
goal articulation, role modeling of desired and self-concept, they argued that transforma-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
attitudes and behaviors, image building, exhibit- tional leaders have several significant psycho-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
ing high expectations of and confidence in logical effects on followers. By placing the
followers, and arousing follower motives that group's mission into moral or spiritual contexts,
were consistent with desired behavior (i.e., such leaders raise the salience of collective
aggressive or altruistic motives). Finally, situ- goals over personal or selfish interests of the
ational influences might include high levels of followers. Second, tying the follower's self-
environmental stress (e.g., economic crises, concept to the group mission makes self-esteem
social upheaval) or an opportunity to express contingent on group success and fosters self-
group goals in moralistic or spiritual terms. motivation and self-regulation by followers.
The most careful, empirical analyses of Both the theoretical explication and perfor-
transformational leadership have been con- mance outcomes associated with transforma-
ducted by Bass and his associates (Bass, 1985; tional leadership make the construct quite
Bass & Avolio, 1990a, 1990b, 1993). Bass compelling but also leave the leadership scholar
started by interviewing managers about transfor- with another conundrum. Transformational theo-
mational leaders they had known. On the basis ries arc stated in terms of "universally"
of the interviews Bass built and validated a effective leadership behavior—that is, for all
questionnaire designed to measure transforma- leaders in all situations. It is difficult to square
tional leadership: the Multi-Factor Leadership that idea with the equally compelling evidence
Questionnaire (MLQ). Factor analysis of the supporting various contingency theories that
MLQ yielded seven factors, including three show that effective leadership is the result of the
"transactional" factors (Contingent Reward, appropriateness or fit between particular behav-
Management By Exception, and Laissez-Faire iors and particular situations. Some recent
Leadership) that were associated with moderate developments applying the concept of self-
to poor leadership effects and four transforma- efficacy (Bandura, 1982) to leadership offers a
tional factors that were associated with high potential resolution of this contradiction.
levels of subordinate motivation and group or
organizational success. The transformational Leadership Efficacy
factors included (a) Idealized Influence (cha-
risma), reflecting extremely high levels of leader A number of contingency model studies
competency, trustworthiness, or both; (b) Inspi- showing that in-match leaders felt more confi-
rational Motivation, involving the articulation of dent and in control (Chemers, Ayman, Sorod, &
the group's goals in emotional, moral, or Akimoto, 1991; Chemers, Hays, Rhodewalt, &
visionary terms; (c) Intellectual Stimulation, Wysocki, 1985) led me and my associates to
entailing the encouragement of followers to conduct a series of studies designed to assess the
think independently and creatively and to move role of leadership confidence or efficacy in
away from past ideas or limitations; and (d) performance. Chemers, Watson, and May (in
Individualized Consideration, relating to the press) reported concurrent, predictive, and
leader's capacity to understand each follower's discriminant validity for a measure of leadership
personal needs and goals. Bass (1998) reported efficacy in a longitudinal study of Reserve
data from many organizations in countries Officer Training Corps cadets. Cadets filled out
around the world that indicate that leaders who a measure of self-perceived leadership ability
are rated highly on transformational leadership and were rated for military leadership potential
characteristics by superiors, peers, or subordi- by their military science instructors. Several
36 CHEMERS
months later, the same cadets were followed up Transformational leadership measures leader-
through a 6-week U.S. Army-sponsored summer ship at the outcome (i.e., dependent variable)
leadership camp in which cadets rotated through level, whereas the contingency theories tend to
everyday leadership roles, participated in train- place more focus on the leader characteristics
ing, and were tested in realistic military (i.e., independent variable) level. Leadership
simulations. Ratings by course instructors, efficacy may be the psychological link between
training camp superiors, peer cadets, and contingent fit and transformational behavior.
simulation exercise grading staff all revealed Later in the present article a functional integra-
dramatic effects indicating superior perfor- tion of contemporary leadership theory will
mance of cadets who expressed greater confi- elaborate these critical functions.
dence in their leadership capability. By way of
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
important. For example, leaders should be seen personal achievement. Here again we see that
as aloof and powerful in high power distance the successful leader is the one who provides
cultures; as expert, confident, and orderly in subordinates with an atmosphere conducive to
high uncertainty avoidance cultures; as paternal- the fulfillment of the followers' personal needs
istic and caring in collectivist cultures, and as and goals but that the nature of those needs and
macho and competitive in masculine cultures. goals is influenced by culturally socialized
Likewise, the needs of followers and the nature values.
of the leader-follower relationship—for ex- The 80 or 90 years of leadership research
ample, as manifested in a desire for structure briefly described in the preceding pages cover a
(uncertainty avoidance) or personal achieve- lot of territory. From contingency theories to
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ment (individualism)—would affect the ways transformational leadership, and cognitive, gen-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
that leaders and followers interact in the der, and cultural factors, a complex pattern of
determination of follower satisfaction, morale, theoretical and empirical material has been
and motivation. generated. The question remaining is whether a
A second, more focused and more empirical coherent integration of these seemingly dispar-
stream of research involved testing various ate findings is possible.
leadership theories across different national
groups. In Japan, Misumi (1984; Misumi & A Functional Integration
Peterson, 1985) conducted an extensive pro-
gram of laboratory and field research, influenced The apparent complexity of research findings
by work in the United States on the LBDQ and theoretical perspectives in the field of
factors Initiation of Structure and Consideration. leadership might be reduced if one examined
Misumi adapated the LBDQ to Japanese culture this literature by focusing on the major functions
by identifying two broad classes of leader that leaders need to fulfill to be successful. I
behavior: (a) behavior associated with work believe that there are three such functions. A
accomplishment through direction and produc- leader must build credibility in the legitimacy of
tivity emphasis (called Performance) and (b) his or her authority by projecting an image that
behavior intended to maintain high group arouses feelings of trust in followers (image
morale (called Maintenance). Misumi and Peter- management). A leader must develop relation-
son (1985) reported that the most productive ships with subordinates that enable those
work groups in Japanese organizations were led subordinates to move toward individual and
by supervisors who were high on both Perfor- collective goal attainment (relationship develop-
mance and Maintenance behavior ment). Finally, leaders must effectively use lie
Ayman and Chemers (1983) reported similar knowledge, skills, and material resources pre-
results for Iranian managers. They factor sent within their group to accomplish the
analyzed a Persian translation of the LBDQ to group's mission {resource deployment).
which they added some probe items related to
the tendency for the worker to identify the Image Management
supervisor in fatherly terms (i.e., "My supervi-
sor is like a kind father to me"). Ayman and The definition of leadership provided earlier
Chemers found that structuring, consideration, stressed that social influence is at the core of the
and the new items collapsed into a single factor, leadership function. Influence depends on cred-
which they labeled Benevolent Paternalism and ibility. For followers to abdicate personal
which was strongly associated with subordinate autonomy and allow themselves to be led, they
satisfaction and performance ratings by superi- must believe that the leader's authority is
ors. Ayman and Chemers concluded that subor- legitimate. The information-processing models
dinates in highly collectivistic and power- of leadership make clear the central role of the
oriented cultures derive satisfaction from a perceptions on which legitimacy is based.
leader who is both directive and nurturant, Individuals who are seen as behaving in ways
whereas subordinates in individualistic, low- that are consistent with observer-held leadership
power cultures such as the United States are prototypes are afforded authority, and subse-
more satisfied with a leader who provides quent perception, attention, and memory are
followers with autonomy and opportunities for more likely to reaffirm the leader's legitimacy.
CHEMERS
We also know that the specific traits or ments about where the subordinate is and where
behaviors associated with leadership credibility the subordinate needs to go next. At the highest
vary somewhat by leadership domain (e.g., levels of leadership performance, these capabili-
political leadership vs. business leadership) and ties are captured by Bass's (1985) concepts of
across cultures (Lord, Foti, & De Vader, 1984; intellectual stimulation (i.e., delicately targeted
Lord, Foti, & Phillips, 1982). Nonetheless, a coaching and guidance that arouse intrinsic
certain commonality exists across leadership motivation) and individualized consideration
prototypes. As Hollander (1958, 1964) demon- (i.e., the subtle but comprehensive awareness of
strated in his early work on leadership status the follower's situation).
accrual, leaders must be seen as competent in
task-relevant abilities and as honest, trustwor-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Resource Deployment
This document is copyrighted by the American Psychological Association or one of its allied publishers.
contingency effects on leadership performance judgments and allow for empathic relationships
at the personal level. with followers.
Both Fiedler's (1967; Fiedler & Chemers, The deployment of personal and team re-
1974, 1984) contingency model and Vroom and sources should be especially affected by confi-
Yetton's (1973) normative decision model are dence. On the personal level, a voluminous
built around the notion that internal group literature in social psychology tells us that
processes, such as decision-making processes, confident and optimistic people are better able to
must match with external task demands to cope with environmental demands (Scheier &
ensure high levels of group performance. For Carver, 1985) and stressful life events (Taylor &
example, overly centralized information-process- Brown, 1988) and are more likely to take risks
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ing strategies and autocratic decision structures (Isen, Nygren, & Ashby, 1988) and solve
This document is copyrighted by the American Psychological Association or one of its allied publishers.
are likely to be more effective in highly routine problems creatively (Isen, Daubman, & Nan-
and predictable leadership situations than they icki, 1987)—all of which are characteristics that
are in more ambiguous, less predictable situa- are related to effective leadership.
tions requiring creative solutions to novel At the level of team deployment that requires
problems. Effective leadership depends on the ability to read the environment accurately
recognizing the nature of the group's environ- and respond flexibly, confidence should also
ment and matching group process to external play a role. Staw and Barsade (1992) measured
demand. management students for positive or negative
dispositional affect. More upbeat individuals
On reflection, it appears that self-confidence performed more effectively at a business deci-
or self-efficacy might play an important role in sion task, integrating more information and
many aspects of leadership effectiveness. If one making better decisions. Experiments by Guzzo
examines the three functions just discussed, one (1986) and by Zaccaro, Peterson, Blair, and
sees a role for self-confidence in each. Effective Gilbert (1990) also show that collective efficacy
image management depends on projecting the (i.e., the shared perception of group members of
appearance of competence. Confidence in one's the capability and effectiveness of the group)
abilities provides a good marker for compe- has been positively related to group perfor-
tence. When we observe competent people it is mance in both experimental and organizational
natural to infer that their confidence is based on settings.
some actual competence. R. J. House's (1977) We also know that when leaders feel that they
analysis of charismatic leadership emphasized are in a congenial, accepting environment they
that outstanding leaders do indeed exhibit high are more likely to act in a directive, "take-
levels of self-confidence. Staw and Barsade charge" fashion (Eagly & Johnson, 1990) and
(1992) reported that MBA students with more that directive leaders are more likely to make
positive emotional dispositions were seen as effective use of their cognitive abilities (Fiedler
more appropriate for leadership roles by ob- & Garcia, 1987).
servers in an assessment center management
simulation.
Summary and a Few Conclusions
Confidence may also play a role in relation-
ship development. Effective coaching and guid- This historical overview of leadership re-
ance are dependent on accurate perceptions of search reveals the extent to which this research
subordinates and cogent attributions about the area, like many others, is influenced by periodic
causes of their behavior and performance. The fashions in research theory; for example, an
major impediment to such perceptual accuracy emphasis on traits at one time; on cognition at
are the ego-defensive motivations created by the another time, and so on. When we take a longer
leader's own concern for positive evaluations. view, we are able to find common findings and
When a group performs poorly, the leader may streams of thought across theoretical perspec-
blame subordinates and be less attentive to tives. The functional integration offered in this
problems caused by other factors, such as the article is an attempt to take such a perspective
support structure or his or her own shortcom- driven especially by an emphasis on what
ings. Confidence in one's own abilities might leaders must do to be effective, that is, to
allow leaders to be less concerned about such influence followers toward goal attainment.
40 CHEMERS
My analysis argues that leaders must first Bales, R. E, & Slater, P. E. (1955). Role differentia-
establish the legitimacy of their authority by tion in small decision-making groups. In T. Parsons
appearing competent and trustworthy to their (Ed.), Family, socialization, and interaction pro-
followers. When leaders are extremely effective cesses. Glencoe, IL: Free Press.
in image management they are seen as possess- Bandura, A. (1982). Self-efficacy mechanism in
human agency. American Psychologist, 37, 122-
ing remarkable, charismatic levels of capability
147.
and trust. Next, leaders must coach, guide, and Bass, B. M. (1981). Stogdill's handbook of leadership
support their followers in a way that allows the (2nd ed.). New York: Free Press.
followers to contribute to group goal attainment Bass, B. M. (1985). Leadership and performance
while satisfying their own personal needs and beyond expectations. New York: Free Press.
goals. To do this, leaders must understand the
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
abilities, values, and personalities of their Industry, military, and educational impact. Mahwah,
subordinates, so they can provide the type of NJ:Erlbaum.
coaching and support that will be most effective. Bass, B. M., & Avolio, B. J. (1990a). The implications
Sometimes leaders are so effective at creating a of transactional and transformational leadership for
motivational environment that followers merge individual, team, and organizational development.
their personal goals with collective group goals In R. W. Woodman & W. A. Passmore (Eds.),
and are transformed in the process. Finally, Research in organizational change and develop-
ment. Greenwich, CT: JAJ Press.
effective leaders must use the skills and abilities
Bass, B. M., & Avolio, B. J. (1990b). Manual for the
possessed by themselves and their followers to multifactor leadership questionnaire. Palo Alto,
accomplish the group's mission. The first step in CA: Consulting Psychologists Press.
utilizing these resources is creating a sense of Bass, B. M., & Avolio, B. J. (1993). Transformational
confidence and personal empowerment that leadership: A response to critiques. In M. M.
encourages each group member to release his or Chemers & R. Ayman (Eds.). Leadership theory
her best efforts. The second step is focusing the and research: Perspectives and directions (pp.
resultant resources on the task environment in a 49-80). San Diego, CA: Academic Press.
way that provides the best fit between group Bass, B. M., Krusell, J., & Alexander, R. A. (1971).
process and environmental demand. Sensitive Male managers' attitudes toward working women.
information processing and intelligent decision American Behavioral Scientist, 15, 221-236.
making are the keys to the group environmental Bowman, G. W., Worthy, N. B., & Greyser, S. A.
interface. (1965). Are women executives people? Harvard
Business Review, 43, 14-28.
In the final paragraphs of this article I raise Brown, K. A. (1984). Explaining group poor perfor-
the hypothesis that leadership efficacy and mance: An attributional analysis. Academy of
group collective efficacy may be the most Management Review, 9, 54-63.
important contributors to each of the functional Burns, J. M. (1978). Leadership. New York: Harper &
necessities of leadership performance. Feelings Row.
of efficacy in the leadership role are thought to Calder, B. J. (1977). An attribution theory of
lead to calm decision making, sensitive interper- leadership. In B. M. Staw & G. R. Salancik (Eds.),
sonal relations, ambitious goal setting, bold New directions in organizational behavior. Chi-
cago: St. Clair Press.
action, and long-term perseverance that energize
Cantor, N., & Mischel, W. (1979). Prototypes in
and maintain the leader and the followers to person perception. In L. Berkowitz (Ed.), Advances
effective common effort. in experimental social psychology (Vol. 12). New
York: Academic Press.
References Carlyle, T. (1907). Heroes and hero worship. Boston:
Adams. (Original work published 1841)
Ashour, A. S. (1973). The contingency model of Chemers, M. M. (1997). An integrative theory of
leadership effectiveness: An evaluation. Organiza- leadership. Mahwah, NJ: Erlbaum.
tional Behavior and Human Performance, 9, Chemers, M. M , & Ayman, R. (1985). Leadership
339-355. orientation as a moderator of the relationship
Ayman, R., & Chemers, M. M. (1983). The between performance and satisfaction of Mexican
relationship of supervisory behavior ratings to managers. Personality and Social Psychology
work group effectiveness and subordinate satisfac- Bulletin, //, 359-367."
tion among Iranian managers. Journal of Applied Chemers, M. M., Ayman, R., Sorod, B., & Akimoto,
Psychology, 68, 338-341. S. (1991, July). Self-monitoring as a moderator of
SPECIAL ISSUE; LEADERSHIP RESEARCH AND THEORY 41
leader-follower relationships. Paper presented at nate reports. Journal of Applied Psychology, 75,
the International Congress of Psychology, Brus- 362-366.
sels, Belgium. Fiske, A. P. (1991). Structures of social life: The four
Chemers, M. M , Hays, R., Rhodewalt, F., & elementary forms of human relations. New York:
Wysocki, J. (1985). A person-environment analy- Free Press.
sis of job stress: A contingency model explanation. Fleishmann, E. A., & Harris, E. F. (1962). Patterns of
Journal of Personality and Social Psychology, 49, leadership behavior related to employee greivances
628-635. and turnover. Personnel Psychology, 15, 43-54.
Chemers, M. M., Watson, C. B., & May, S. (in press). Graen, G. (1976). Role making processes within
Dispositional affect and leadership performance: A complex organizations. In M. D. Dunnette (Ed.),
comparison of self-esteem, optimism, and efficacy. Handbook of industrial and organizational psychol-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Cleven, W. A., & Fiedler, F. E. (1956). Interpersonal Graen, G., Alvarez, K., Orris, J. B., & MarteUa, S. A.
perceptions of open-hearth foremen and steel (1970). Contingency model of leadership effective-
production. Journal of Applied Psychology, 40, ness: Antecedent and evidential results. Psychologi-
312-314. cal Bulletin, 74,285-296.
Deaux, K. (1984). From individual differences to Gracn, G., & Cashman, J, (1975). A role-making
social categories. American Psychologist, 39, model of leadership in formal organizations: A
105-116. developmental approach. In J. G. Hunt & L. L.
Eagly, A. R , & Johnson, B. T. (1990). Gender and Larson (Eds.), Leadership frontiers. Kent, OH:
leadership style: A meta-analysis. Psychological Kent State University Press.
Bulletin, 108, 233-256. Graen, G., Cashman, J. E, Ginsburgh, S., &
Eagly, A. H., & KarauT S. J. (1991). Gender and the Schiemann, W. (1978). Effects of linking-pin
emergence of leaders: A meta-analysis. Journal of quality on the quality of working life of lower
Personality and Social Psychology. 60, 685-710. participants: A longitudinal investigation of the
Eagly, A. R , Makhijani, M. G., & Klonsky, B. G. managerial understructure. Administrative Science
(1992). Gender and the evaluation of leaders: A Quarterly, 22, 491-504.
meta-analysis. Psychological Bulletin, 111, 3-22. Graen, G., & Scandura, T. A. (1987). Toward a
Eden, D., & Leviatan, U. (1975). Implicit leadership psychology of dyadic organizing. Research in
theory as a determinant of the factor structure Organizational Behavior, 9, 175-208.
underlying supervisory behavior scales. Journal of Green, S. G., & Mitchell, T. R. (1979). Attribution^
Applied Psychology, 60, 736-741. processes of leaders in leader-member interac-
Fiedler, F. E. (1955). The influence of leader-keyman tions. Organizational Behavior and Human Perfor-
relations on combat crew effectiveness. Journal of mance, 23, 429-458.
Abnormal and Social Psychology, 51, 227-235. Griffin, R. N. (1981). Relationships among indi-
Fiedler, F. E. (1958). Leader attitudes and group vidual, task design, and leader behavior variables.
effectiveness. New York: McGraw-Hill. Academy of Management Journal, 23, 665-683.
Fiedler, F. E. (1964). A contingency model of Guzzo, R. A. (1986). Group decision making and
leadership effectiveness. In L. Berkowitz (Ed.), group effectiveness in organizations. In P. Good-
Advances in experimental social psychology (Vol. man & Associates (Eds.), Designing effective work
1, pp. 149-190). New York: Academic Press. groups (pp. 34-71). San Francisco: Jossey-Bass.
Fiedler, F. E. (1967). A theory of leadership Hackman, J. R., & Oldham, G. R. (1976). Motivation
effectiveness. New York: McGraw-Hill. through the design of work: Test of a theory.
Fiedler, F. E., & Chemers, M. M. (1974). Leadership Organizational Behavior and Human Perfor-
and effective management. Glenview, IL: Scott, mance, 16, 250-279.
Foresman. Halpin, A. W., & Winer, B. J. (1957). A factorial study
Fiedler, F. E., & Chemers, M. M. (1984). Improving of the leader behavior descriptions. In R. M.
leadership effectiveness: The Leader Match con- Stogdill & A. E. Coons (Eds.), Leader behavior: Its
cept (2nd ed.). New York: Wiley. description and measurement. Columbus: Ohio
Fiedler, F. E., Chemers, M. M., & Maher, L. (1976). State University Bureau of Business Research.
Improving leadership effectiveness: The Leader Hastorf, A. R , Schneider, D., & Polefka, J. (1970).
Match concept. New York: Wiley. Person perception. Reading, MA: Addison-
Fiedler, F. E., & Garcia, J. E. (1987), New approaches Wesley.
to effective leadership: Cognitive resources and Heilman, M. E., Block, C. J., Martell, R. F., & Simon,
organizational performance. New York: Wiley. M. C. (1989). Has anything changed? Current
Field, R. H. G., & House, R. J. (1990). A test of the characterizations of men, women, and managers.
Vroom-Yetton model using manager and subordi- Journal ofApplied Psychology, 74, 935-942.
42 CHEMERS
Hemphill, J. K. (1950). Leader behavior description. of Personality and Social Psychology, 55, 710—
Columbus: Ohio State University Personnel Re- 717.
search Board. Jones, E. E., & Davis, K. E. (1965). From acts to
Hennig, M., & Jardim, A. (1977). The managerial dispositions: The attribution process in person
woman. New York: Anchor Press. perception. In L. Berkowitz (Ed.), Advances in
Hofstede, G. (1980). Culture's consequences; Interna- experimental social psychology (Vol. 2). New
tional differences in work-related values. Beverly York: Academic Press.
Hills, CA: Sage. Jones, E. E., & Nisbett, R. E. (1971). The actor and
Hofstede, G. (1983). Motivation, leadership, and the observer: Divergent perceptions of the causes
organization: Do American theories apply abroad? of behavior. Morristown, NJ: General Learning
Organizational Dynamics, 9, 42-63. Press.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Hogg, M. A., Hains, S. C , & Mason, I. (1998). Kahn, R. L. (1951). An analysis of supervisory
This document is copyrighted by the American Psychological Association or one of its allied publishers.
Leader behavior and situational moderators in tics among female managers. Journal of Applied
This document is copyrighted by the American Psychological Association or one of its allied publishers.