Identify leader traits and attributes
10.1177/0149206304271765ARTICLEJournal of Management / April 2005Groves / Charismatic Leadership
Linking Leader Skills, Follower Attitudes, and Contextual Variables via an
Integrated Model of Charismatic Leadership†
Kevin S. Groves* College of Business and Economics, Department of Management, California State University, Los Angeles,
5151 State University Drive, Los Angeles, CA 90032-8126
A charismatic leadership model consisting of leader social and emotional skills, follower open- ness to organizational change, and organizational-change magnitude was tested using data from 108 leaders and 325 direct followers in 64 organizations. Leader social control and emotional expressivity skills predicted charismatic leadership whereas follower openness to change medi- ated the relationship between charismatic leadership and leadership effectiveness. Surprisingly, organizational-change magnitude did not moderate the relationship between charismatic leader- ship and leadership effectiveness. Implications for leadership theory, practice, and future research are discussed.
Keywords: charismatic leadership; leader social/emotional skills; organizational change; follower attitudes
Dozens of empirical studies and several meta-analytic reviews have demonstrated the pow- erful effects of charismatic leadership behavior in organizations. Recent research has shown that charismatic leadership strongly affects follower performance and attitudes (Lowe, Kroeck, & Sivasubramaniam, 1996), perceptions of leadership effectiveness (Shamir, Zakay,
† This research project was funded by the Rudolph Haynes Social Science Dissertation Grant, and this article repre- sents work from the author’s dissertation. I wish to thank my dissertation chair, Susan Murphy, for her tutelage throughout the research project, and committee members Ronald Riggio and Stewart Donaldson for their many insightful suggestions.
* Corresponding author. Tel: 323 343-5429; fax: 323 343-6461.
E-mail address: kgroves@calstatela.edu
Journal of Management, Vol. 31 No. 2, April 2005 255-277 DOI: 10.1177/0149206304271765 © 2005 Southern Management Association. All rights reserved.
255
Breinin, & Popper, 1998), stock performance (Agle & Sonnenfeld, 1994), and organizational profitability (Waldman, Ramirez, House, & Puranam, 2001). Although charismatic leader- ship theory and the hypothesized relationships between charismatic behaviors and various outcomes have received considerable empirical support, several conceptual and methodologi- cal weaknesses limit this area of leadership research. First, research on charismatic leadership has generated very limited empirical support for the mediating influence processes through which these leaders create their powerful effects in organizations (Yukl, 1999). Second, very little is known about the role that followers play in charismatic leadership and how charismatic leaders influence follower characteristics and attitudes in organizations (Conger, 1999). Third, few empirical studies (e.g., Shamir & Howell, 1999; Waldman et al., 2001) have inves- tigated the contextual variables that facilitate or inhibit charismatic leadership. Finally, recent meta-analyses (Dumdum, Lowe, & Avolio, 2002; Lowe et al., 1996) indicate that mono- method bias limits the majority of charismatic leadership studies due to a predominance of single-source follower survey ratings for both leadership behavior and leadership effectiveness.
The purpose of the present study is to test an integrated model of charismatic leadership consisting of leader characteristics and behaviors, follower attitudes, and contextual variables and address the limitations of charismatic leadership theory mentioned above. Specifically, I argue that leader social and emotional skills, follower openness to organizational change, and organizational-change magnitude are critical components that collectively explain the effi- cacy of charismatic leadership in organizations. A key goal of the present investigation is to expose the manner in which charismatic leaders influence followers’attitude toward organiza- tional change and subsequent outcomes, thereby establishing support for an important medi- ating process of charismatic leadership. Both the lack of attention devoted to followers’ role in the leadership process and the very limited empirical support for the mediating variables that explain the efficacy of charismatic leadership continue to limit our understanding of this pow- erful phenomenon. Another objective of this study is to advance the developing stream of leadership research on the role of social and emotional intelligence by demonstrating empiri- cal support for specific leader social and emotional skills that contribute to charismatic leader- ship. Furthermore, this study will assess the magnitude of organizational change as a key con- textual variable that facilitates charismatic leadership and also address an important methodological limitation of charismatic leadership research by testing the integrated model with a split sample of followers.
Theoretical Background and Hypotheses
Integrated model. As illustrated in Figure 1, the central component of the proposed inte- grated model is charismatic leadership. Charismatic leadership is defined as a process of trans- forming followers from an existing present state to an improved future state by demonstrating sensitivity to member needs, environmental sensitivity, vision and articulation, personal risk, and unconventional behavior (Conger & Kanungo, 1994). The integrated model postulates that specific social and emotional skills facilitate a leader’s ability to enact these behavioral dimensions. Furthermore, follower attributions of charismatic qualities to a leader positively
256 Journal of Management / April 2005
influence follower attitudes toward organizational change and perceptions of leadership effec- tiveness. In addition, organizational contexts characterized by highly significant change are modeled to strengthen the relationship between charismatic leadership and leadership effec- tiveness. Thus, follower attribution of charismatic qualities to a leader is collectively deter- mined by the leader’s social and emotional skills, the leader’s demonstration of charismatic behavioral dimensions, and the magnitude of organizational change in the work unit. Below, I offer the integrated model by first examining leader social and emotional skills and their re- lationship to charismatic leadership, followed by a review of charismatic leadership and how it is potentially related to follower openness to change, organizational-change magnitude, and leadership effectiveness.
Groves / Charismatic Leadership 257
Leader Social
Control
Leader Emotional
Expressivity
Leader Emotional
Control
Follower Ratings of
Charismatic Leadership
Follower Ratings of Leadership
Effectiveness
.22*
Leader and Follower Ratings of Organizational
Change Magnitude
.
Control Variables
Leader Gender
Leader Tenure
For-/Non-profit
.05 .19*
.20*
.33*
-.09
H1 (.20*)
H2 (.22*)
H3 (.15+)
H4 (.40**)
H5 (.42**)
H6 (.16 )
Follower Openness to Organizational
Change
+
Figure 1 Proposed Integrated Model of Charismatic Leadershipa
a. Model fit: χ2 = 46.98 (df = 30, p < .05); Goodness-of-Fit Index (GFI) = .96; Adjusted Goodness-of-Fit Index (AGFI) = .90; root mean square error of approximation (RMSEA) = .05; Incremental Fit Index (IFI) = .93; Comparative Fit Index (CFI) = .92. + p < .10 *p < .05 **p < .01
Leader social and emotional skills. The first component of the integrated model addresses several sets of social and emotional skills that enable charismatic leadership behaviors. A recent development in leadership research, particularly transformational and charismatic leadership, is the importance of leader social and emotional intelligence and related interper- sonal communication competencies. For example, Zaccaro’s (2002) notion of social intelli- gence includes various social interaction competencies, such as the ability to identify and interpret social cues, and subsequently adjust behavioral responses. Similarly, Mayer and Salovey’s (1997) ability model of emotional intelligence involves several nonverbal commu- nication competencies, including the ability to perceive and understand emotions, regulate emotions, and generate or express emotions. Clearly, these popular models of social and emo- tional intelligence subsume interpersonal communication competencies, yet there is surpris- ingly little empirical research that links specific social and emotional skills with charismatic leadership.
However, a recent trend in research on emotional intelligence and charismatic leadership considers the role of self-monitoring, which is defined as the ability to monitor and control one’s expressive behaviors (Snyder, 1974). Although theoretical (Garnder & Avolio, 1998) and empirical research (Sosik & Dworakivsky, 1998) has linked self-monitoring ability to charismatic leadership, several scholars have questioned the validity of self-monitoring on grounds that it is a multidimensional construct consisting of several components (acting, extroversion, and other-directedness) that are inconsistent with one another (Briggs & Cheek, 1988). Some research suggests that social control (Riggio, 1989), a social skill conceptually similar to self-monitoring, may be a more valid means of measuring one’s ability to control expressive behavior. Social control is a one-dimensional construct defined as social self- presentation skill that enables an individual to adjust personal behavior to fit with what he or she considers appropriate in a given situation (Riggio, 1989). Persons high in social control are tactful, socially adept, self-confident, and excel in social role-playing abilities. Indeed, Riggio’s (1989) validation studies showed that social control demonstrated strong positive relationships with the acting and extroversion factors of self-monitoring and a strong negative relationship with the other-directedness factor.
Theoretical and empirical research suggests that social control skill may be a key interper- sonal competency related to charismatic leadership. For instance, research has shown that individuals possessing social control report greater comfort in public speaking (Riggio, 1989), accurately assess interpersonal relationships, understand the meaning of verbal and behav- ioral cues in different contexts, and successfully pose basic emotional expressions (Tucker & Riggio, 1988). Charismatic leaders rely heavily on impression management and image build- ing to express conviction, self-confidence, and dedication to their vision (Conger, 1999; Con- ger & Kanungo, 1994) and construct a charismatic identity that is valued by followers (Gardner & Avolio, 1998). Furthermore, charismatic leaders must be acutely sensitive to fol- lowers’ needs and values and use these cues as input for constructing and articulating their vision. Social control skills may also be used by charismatic leaders to estimate how well they are convincing both supporters and detractors that their vision is right for the organization. In short, leaders with social control skills can detect critical cues from adversaries and refine their message accordingly. Given the evidence linking social control and charismatic leadership, the following hypothesis is presented:
258 Journal of Management / April 2005
Hypothesis 1: Leader social control skills will be positively related to charismatic leadership.
In addition to the integral role of social control skills, recent research suggests that charis- matic leaders rely on various nonverbal, emotional skills to influence and motivate followers. Although no study to date has examined the relationship between specific emotional skills and charismatic leadership, several studies have examined the role of vision delivery in eliciting perceptions of leader charisma. Studies by Howell and Frost (1989), Holladay and Coombs (1994), and Awamleh and Gardner (1999) found that a strong delivery style characterized by nonverbal, emotional communication is a key determinant of perceived charisma. In each of these studies, delivery style was manipulated according to such nonverbal communication skills as eye contact, animated facial expressions, body gestures, and posture. Interestingly, Holladay and Coombs (1994) found that the combination of nonvisionary content and strong delivery elicited greater perceptions of charisma than visionary content and weak delivery.
Although these studies provide support for the importance of vision delivery through non- verbal communication, the findings discussed above are somewhat limited in that each study used professional actors to portray strong (charismatic) and weak (noncharismatic) delivery. The present investigation will test the relationship between emotional skills and charismatic behavior using a sample of senior organizational leaders and a validated measure of nonverbal emotional skills, emotional expressivity (Riggio, 1989). Defined as basic nonverbal expres- sion of emotions through facial expressions, tone of voice and other paralinguistic cues, and posture/body movements, emotional expressivity has been described by many as an integral component of charisma (Riggio, 1992). In a review of the research on emotional expressivity, Riggio (1992) concluded that emotionally expressive individuals have the ability to establish subtle connections with others at an emotional level and transmit emotional messages that can affect the moods and emotional states of others. Similarly, Gardner and Avolio (1998) argued that charismatic leaders use body posture and gestures, speaking rate, smiles, eye contact, and touch to project a powerful and confident presence and that such nonverbal displays may be used to elicit desired responses from followers. Based on the preceding discussion of leader emotional expressivity and charismatic leadership, the following hypothesis is presented:
Hypothesis 2: Leader emotional expressivity skills will be positively related to charismatic leadership.
Although research supports the notion that nonverbal expressiveness and the ability to communicate emotional messages to followers are important skills for charismatic leaders, no empirical study to date has addressed leaders’ability to regulate or control their emotional dis- plays. Riggio (1989) defined emotional control as the ability to control and regulate nonverbal and emotional displays and asserted that individuals high on emotional control are likely to be good nonverbal actors capable of posing emotions on cue and adjusting their nonverbal behav- ior to “fit in” to various social situations. If emotional displays and nonverbal expressiveness are critical elements of charismatic leaders’ ability to motivate followers, it stands to reason that these leaders must also sufficiently regulate their emotional communications. Certainly, ill-timed emotional pleas or excessive emotional displays may cause followers to perceive their leader as frivolous and disingenuous. Indeed, several authors have asserted that percep-
Groves / Charismatic Leadership 259
tions of charisma may reflect a leader’s balanced possession of communication skills rather than exclusively emotional expressiveness (Gardner & Avolio, 1998; Riggio, 1989; Sosik & Dworakivsky, 1998). Thus, the following hypothesis is proposed:
Hypothesis 3: Leader emotional control skills will be positively related to charismatic leadership.
Charismatic leadership. The concept of charisma and how it relates to organizations can be traced back to Weber’s (1947) analysis of authority and imperative control in organizations. Intrigued by the forces of authority in society, Weber defined charisma as legitimacy derived from a leader’s exceptional powers or qualities. Upon recognizing the exceptional qualities of a charismatic leader, followers are likely to exhibit complete personal devotion to the leader and his or her articulated vision. Weber was also intrigued with the creation, maintenance, and transformation of institutional arrangements and categorized charisma as bringing about change and innovation in society. Given this conceptualization, many leadership researchers of the 1980s and 1990s examined charismatic leaders as the focal point for understanding organizational change.
Across the many prominent theories of charismatic leadership, including House’s 1976 theory (1977); Shamir, House, and Arthur’s (1993) self-concept-based theory; and Gardner and Avolio’s (1998) model, Conger and Kanungo’s (1994) behavioral model (C-K) provides several advantages for the present investigation. The C-K model defines charismatic leader- ship as a process of transforming followers from an existing present state to an improved future state through the following behavioral dimensions: sensitivity to member needs, envi- ronmental sensitivity, vision and articulation, personal risk, and unconventional behavior. Fol- lower attribution of charismatic qualities to a leader is determined by the leader’s demonstra- tion of the behavioral dimensions, several of which are well suited for testing the integrated model and examining the research questions of the present study. For example, a leader’s abil- ity to demonstrate sensitivity to followers’ needs and feelings is a critical facet of charismatic leadership that requires social and emotional skills. Similarly, effective articulation of an inspiring vision and goals to followers often requires emotional, nonverbal communication skills such as emotional expressivity. Furthermore, the C-K model makes explicit the role of leadership behaviors in the context of organizational change; the present study’s analysis of the change context and follower attitudes toward change necessitates a leadership theory that models the necessary leadership behaviors for effectively navigating followers through the change process. Finally, the other prominent charismatic leadership theories (e.g., Bass, 1985; House, 1977; Shamir et al., 1993) have been criticized for including both leader behaviors and follower effects as part of their conceptualizations of the leadership process (Conger, 1999), whereas the C-K model is unique in its focus on the specific distinguishing behavioral dimensions of charismatic leaders. For a more detailed discussion of the C-K model compared to other charismatic leadership theories, see Conger and Kanungo (1998).
Follower attitudes toward organizational change. A central goal of the present study is to examine follower attitudes toward organizational change as an important factor that helps explain the efficacy of charismatic leadership in organizations, particularly those organiza-
260 Journal of Management / April 2005
tions experiencing large-scale changes. Of all the streams of research on transformational and charismatic leadership, issues surrounding the role of followers and how they affect the charis- matic relationship and organizational change have been neglected most. Although there are theoretical contributions regarding followers’ role in charismatic leadership (Gardner & Avolio, 1998), no empirical research to date supports the notion that follower attitudes toward change influence the charismatic relationship between leaders and followers or follower per- ceptions of leadership effectiveness. Because charismatic leadership behavior is predicated on the development and implementation of a vision that contrasts with the status quo (Bass, 1985; Conger & Kanungo, 1994; Shamir et al., 1993), significant organizational changes and ensu- ing employee resistance to such changes are likely outcomes of the charismatic leadership process. Although there is a great deal of theoretical research on employee resistance to change (e.g., Piderit, 2000), no empirical study to date has addressed the relationship between charismatic leadership and follower attitudes toward organizational change.
However, recent work on psychological ownership in organizations (Dirks, Cummings, & Pierce, 1996) suggests that key follower beliefs may help explain the relationships between charismatic leadership, organizational change, and leadership effectiveness. On the basis of the concept of psychological ownership, defined as the feeling of being psychologically tied to an organization, Dirks et al. assert that three human needs are related to psychological owner- ship: self-enhancement, self-continuity, and control/efficacy. The model predicts that individ- uals will promote change efforts under conditions that meet their needs for self-enhancement, self-continuity, and/or control and efficacy. Accordingly, when charismatic leaders provide followers with a sense of ownership of the organization’s collective vision (Conger & Kanungo, 1994; Shamir et al., 1993), followers’ needs for self-enhancement, self-continuity, and control are maintained. Indeed, research on Shamir et al.’s self-concept-based theory shows that follower self-esteem, self-worth, and self-consistency are enhanced when follow- ers perceive a sense of ownership of the vision (Shamir et al., 1998). Thus, charismatic leader- ship may suppress followers’ resistance to organizational change by actively addressing their needs throughout the change process.
Pawar and Eastman’s (1997) analysis of organizational factors that affect the receptivity of charismatic leadership provides additional evidence of the relationship between charismatic leadership and follower attitudes toward organization change. They suggest that the leader’s role in organizations undergoing adaptation periods, which are characterized by large-scale organizational change, is to overcome follower resistance to change and align the organization to its new environment through charismatic leadership behaviors. According to Pawar and Eastman, follower resistance to change will occur unless leaders cause followers to experience a “felt need” for the changes taking place. Indeed, several core charismatic behaviors affect followers’ receptivity of organizational change, including the ability to powerfully articulate a deficient status quo and inspiring vision, communicate to followers a sense of ownership of the vision, and demonstrate confidence in followers’ abilities to realize the vision. Given the pre- ceding discussion of charismatic leadership and follower attitudes toward change, the following hypotheses are presented:
Hypothesis 4: Charismatic leadership will be positively related to follower openness to organiza- tional change.
Groves / Charismatic Leadership 261
Hypothesis 5: Follower openness to organizational change will partially mediate the relationship between charismatic leadership and follower ratings of leadership effectiveness.
Leadership effectiveness and organizational change. The final component of the integrated model concerns the effectiveness of charismatic leaders in the context of organizational change. Clearly, empirical research has provided great support for the theorized direct effects of charismatic leadership in organizations. Meta-analytic reviews (e.g., Dumdum et al., 2002; Lowe et al., 1996) summarizing years of empirical research have concluded that charismatic leadership demonstrates strong positive relationships with a range of important outcomes, particularly follower perceptions of leadership effectiveness. However, as noted by several leadership scholars (e.g., Conger, 1999; Shamir & Howell, 1999; Yukl, 1999), the literature on charismatic leadership has failed to adequately address the organizational context in which such leadership and follower effects are embedded. Although several theorists (Shamir et al., 1993; Shamir & Howell, 1999) have proposed contextual variables that facilitate charismatic leadership, including environmental uncertainty, crisis, and organic organizational structure, only a few empirical studies have demonstrated situational effects (Waldman et al., 2001).
An important contextual variable of charismatic leadership that has gained recent momen- tum from researchers is the organizational environment in which such leadership takes place. In a review of the organizational and contextual influences on charismatic leadership, Shamir and Howell (1999) proposed that the emergence and effectiveness of charismatic leadership are more likely in organizational environments characterized by a high degree of change than in stable environments that provide few opportunities for change. Because dynamic environ- ments and requisite organizational changes require novel responses from leaders and extraor- dinary effort and commitment from followers, Shamir and Howell assert that “the charismatic leader’s vision often brings to potential followers’ attention the existence of opportunities for change, infuses them with hope and faith regarding that change, and mobilizes their energies to single mindedly devote themselves to the vision” (p. 264). Similarly, Nadler and Tushman (1990) argued that although organizations routinely experience change, the magnitude of organizational changes varies from minor changes to large-scale changes. Furthermore, they assert that the role of leadership differs according to organizational-change magnitude such that existing management structures and processes are capable of implementing minor changes, whereas charismatic leadership is necessary for engineering large-scale changes. Thus, charismatic leadership behaviors, including sensitivity to follower needs, environmen- tal sensitivity, and visioning, are critical for initiating major movements in organizations by inducing necessary changes in followers’ attitudes toward change, level of effort, and trust in the leader’s vision. On the basis of the preceding discussion of organizational change and charismatic leadership, I present the following hypothesis:
Hypothesis 6: The strength of the relationship between charismatic leadership and leadership effec- tiveness will vary depending on the magnitude of organizational change. The charismatic leader- ship–leadership effectiveness relationship will be stronger in conditions of high-change magni- tude compared with low-change magnitude.
262 Journal of Management / April 2005
Method
Participants, Sample Organizations, and Procedures
A total of 433 respondents, both senior organizational leaders (n = 108) and their direct fol- lowers (n = 325) from 64 organizations participated in this study. The organizations represent the following industries: higher education (n = 21, 33%), community development (n = 14, 22%), government (n = 10, 16%), health care (n = 7, 11%), and other (n = 12, 19%), whereas 78% (n = 50) were not-for-profit institutions. The senior leaders in the sample included 67 men (62%) and 41 women (38%). The ethnic breakdown of the leaders was as follows: White (n = 70, 65%), Asian American (n = 10, 9%), African American (n = 10, 9%), Hispanic (n = 14, 13%), and Other (n = 4, 4%). The leaders were employed at their current management position for an average of 5.59 years (SD = 5.89). There was a total of 325 followers in this study, with 3.01 followers per leader. Across the sample of 108 leaders, 34 leaders had two fol- lower participants, 44 leaders had three followers, 25 leaders had four followers, and 5 leaders had five followers.
Sample organizations. Leaders’ and their respective followers’ participation in the present study was completely voluntary, as the present study was an independent research project completed by the author. Participants were drawn from four sources: a professional health sci- ences university, a public services and works government agency, a professional association of U.S. medical colleges, and a community-based leadership development program. The pro- fessional health sciences university participants included 24 institutional leaders at several management levels, including senior vice presidents (n = 4, 17%), deans (n = 5, 21%), and functional managers (n = 15, 62%). Participants from the public services and works govern- ment agency included 22 engineering managers responsible for a range of public services such as environmental services and flood control engineering. Participants from the professional association of medical colleges included 12 presidents and chief executive officers (CEOs) of U.S. osteopathic medical schools. Finally, participants from the community-based leadership development program included 50 organizational leaders from a wide range of local organiza- tions, including governmental agencies, for-profit businesses, universities, and nonprofit organizations. The leaders in this program did not receive training in charismatic leadership behaviors nor any of the leader skills or behaviors in this study.
Procedures. For the health sciences university, government agency, and professional asso- ciation of medical colleges, an invitation to participate in the study was sent via e-mail to man- agerial personnel, enclosed with a questionnaire measuring organizational change and social/ emotional skills (“leader questionnaire” hereafter). After leader participants identified the members of their departments or work groups, I sent a second questionnaire to each leader’s direct followers that measured follower openness to organizational change and their respec- tive manager’s charismatic leadership behavior and leadership effectiveness (“follower ques- tionnaire” hereafter). The veracity of the reported members of each manager’s department or
Groves / Charismatic Leadership 263
work unit was established through consultation with a human resource manager at each orga- nization and official organizational charts. For the community-based leadership development program, 162 local business leaders who had participated in the program during the past 3 years were sent an invitation to participate in the study, the leader questionnaire, and five sepa- rate envelopes that each included the follower questionnaire. The leaders were instructed to deliver the separate envelopes to their respective direct reports. Of the 162 leaders who received the packet of materials, 50 agreed to participate for a response rate of 31%. The over- all response rate for leaders at all sample organizations was 45%, whereas the overall response rate for followers at all sample organizations was 62%.
Measures
Leader social and emotional skills. Leader social control, emotional expressivity, and emo- tional control skills were measured using scales from an abbreviated version of Riggio’s (1989) Social Skills Inventory, which assesses an individual’s repertoire of social and emo- tional communication skills. On a Likert-type scale ranging from 1 (not at all like me) to 5 (exactly like me), leaders were asked to provide a self-report of their social and emotional skills. To help limit the potential for inflated self-assessments, this section of the leader survey was titled “Self-Description Inventory,” and leader participants were instructed to rate a series of statements that “indicate an attitude or behavior that may or may not be characteristic or descriptive of you” (Riggio, 1989). Thus, leader participants had no reason to believe they were self-assessing social and emotional skills, which is critical given research suggesting that individuals tend to overrate their abilities and skills on self-assessments (Kruger & Dunning, 1999). The following are sample items for social control, emotional expressivity, and emo- tional control, respectively: “I can easily adjust to being in just about any social situation,” “My facial expression is generally neutral” (reverse coded), and “I am rarely able to hide a strong emotion” (reverse coded). Cronbach’s alpha reliabilities for social control, emotional expressivity, and emotional control were .75 (5 items), .71 (6 items), and .74 (6 items), respectively.
Charismatic leadership. Charismatic leadership behavior was measured by Conger and Kanungo’s (1994) 25-item charismatic leadership scale (C-K). On a Likert-type scale ranging from 1 (very uncharacteristic) to 7 (very characteristic), followers were asked to comment on the extent to which their leader exhibits a series of leadership behaviors. The C-K scale is com- posed of six subscales that measure behavioral dimensions of charismatic leadership, includ- ing the leader’s desire to change the status quo (Does Not Maintain Status Quo or DNMSQ), heightened sensitivity to environmental opportunities and constraints (Environmental Sensi- tivity) and followers’needs (Sensitivity to Member Needs), formulation of a shared and ideal- ized vision and inspirational articulation of the vision to organizational members (Vision and Articulation), engaging in exemplary acts that involve great personal risk and sacrifice (Per- sonal Risk), and using innovative and unconventional means for achieving the vision (Uncon- ventional Behavior). The alpha for the overall modified scale was .82, and each of the subscales also demonstrated acceptable reliability: vision and articulation (6 items, α = .91),
264 Journal of Management / April 2005
environmental sensitivity (7 items, α = .81), sensitivity to member needs (3 items, α = .81), personal risk (4 items, α = .87), unconventional behavior (3 items, α = .76), and DNMSQ (2 items, α = .72).
Organizational-change magnitude. To assess the magnitude of change at the sample orga- nizations, both leaders and their respective followers were asked to characterize the changes that occurred at their respective organization during the past year on a scale ranging from 5 (extremely significant) to 1 (little or no changes). Interrater reliability analyses (rwg; James, Demaree, & Wolf, 1984) indicated sufficient leader-follower agreement with an alpha of .73. Using the same 5-point scale, leaders and followers were asked to rate the magnitude of orga- nizational change on a five-item scale composed of the following dimensions: structural (e.g., reorganization, changes in reporting relationships), cultural (e.g., changes in philosophy, val- ues, expected behavior), technological (e.g., new equipment, computer systems), procedural (e.g., new work processes, policies, practices), and personnel (e.g., new position, job func- tion[s]). Leaders and their respective followers’ ratings across these dimensions of change were summed and averaged to form a composite measure of organizational-change magnitude that produced an alpha of .78. Thus, the organizational-change magnitude measure represents both leader and followers’ overall perceptions of the organizational changes that occurred in their respective work units during the past year.
Openness to organizational change. Follower openness to organizational change was assessed by a seven-item, modified version of Miller, Johnson, and Grau’s (1994) Openness to Change Scale. Because this scale was originally developed to assess employees’ willingness to support the organizational change of implementing work teams, slight modifications were made to make this scale appropriate for a range of organizational changes. Thus, all followers in the present study were able to address their attitude toward change regardless of the specific changes occurring at their organization. On a Likert-type scale ranging from 1 (strongly dis- agree) to 7 (strongly agree), followers were asked to rate the extent to which they were open to the changes occurring at their workplace. The overall reliability of the scale was acceptable, with an alpha of .87.
Leadership effectiveness. Leadership effectiveness was assessed using Agle and Sonnenfeld’s (1994) six-item Leader Accomplishment Scale that addresses leader perfor- mance in terms of the attainment of management goals and objectives. The selection of this scale over alternative measures of leadership effectiveness was based on several factors. First, brevity was an important criterion given the nature of the sample and the potentially large number of total items on the follower questionnaire. Second, the scale was carefully devel- oped according to the procedure outlined by DeVellis (1991), including a comprehensive review by a sample of executives. Finally, the scale’s emphasis on management accomplish- ments rather than follower outcomes was important because the follower sample was split to control for common method bias. The questionnaires from half of each leader’s followers were used to measure charismatic leadership, and the other half were used to measure follower openness to change and leadership effectiveness. The use of an objective-based measure was intended to reduce the likelihood that followers’ ratings of openness to change would affect
Groves / Charismatic Leadership 265
their ratings of leadership effectiveness. On a Likert-type scale ranging from 1 (strongly dis- agree) to 7 (strongly agree), followers were asked to rate the extent to which their leader accomplished the goals and objectives of their work units. The reliability of the overall leadership effectiveness scale was acceptable, with an alpha of .93.
Control variables. Leader tenure and gender, and nonprofit/for-profit organizational status were used as control variables in the present study. Because a manager’s prior work experi- ence is often strongly related to leadership effectiveness, number of years at the current posi- tion was used as a control variable. Leader gender was also controlled because prior research has shown that women tend to have greater social and emotional skills than men (Riggio, 1986). Given the range of organizations and industries from which the leaders were drawn, leader industry was coded into five primary industries represented by the leaders’ respective organizations. Tukey’s Honestly Significant Difference (HSD) test identified no significant differences in any of the study’s variables across the leaders’ respective industries; therefore, leader industry was not included as a control variable for hypothesis testing. However, given the predominance of nonprofit organizations in the sample, nonprofit/for-profit organiza- tional status was included as a control variable.
Analyses
Level of analysis. The leader was chosen as the unit of analysis for the present study such that items measuring social control, emotional expressivity, emotional control, and organiza- tional change were averaged to create scale scores for each leader participant, whereas fol- lower ratings of charismatic leadership, openness to organizational change, and leadership effectiveness were averaged and aggregated. To appropriately conceptualize and measure the extent to which leaders demonstrate social and emotional skills, charismatic behaviors, and also affect followers’attitudes toward the changes occurring in their respective work units, the present study adopted the close charismatic level of analysis described by Shamir (1995). Because immediate followers have day-to-day interaction with their leader and are privy to direct observation of their leader’s interpersonal behavior, they are much more likely to pro- vide accurate ratings of charisma, attitudes toward change, and leadership effectiveness than distant followers. Furthermore, concerns about followers’ romance of leadership and implicit leadership theories are significantly lessened when immediate followers are providing ratings of charismatic leadership and leadership effectiveness rather than distant followers. To justify the aggregation of follower ratings to create group-level measures, within-group reliability statistics (rwg; James et al., 1984) were calculated. The interrater reliability coefficients for char- ismatic leadership, follower openness to change, and leadership effectiveness were .82, .75, and .70, respectively. Overall, these results provide support for the aggregation of follower ratings to create group measures of charismatic leadership, openness to organizational change, and leadership effectiveness.
Control for common method variance. Lowe et al. (1996) and Dumdum et al.’s (2002) meta-analyses of transformational and charismatic leadership studies concluded that follower
266 Journal of Management / April 2005
ratings of charismatic leadership and leadership effectiveness are strongly related to one another, with mean corrected correlations of .81 and .85, respectively. However, as noted by several authors (e.g., Dvir, Eden, Avolio, & Shamir, 2002), mono-method bias is particularly problematic when assessing the relationship between follower ratings of charisma and leader- ship effectiveness, and many of the studies in the Lowe and Dumdum meta-analyses used fol- lower ratings of both charismatic leadership and leadership effectiveness. In short, using fol- lower ratings of both charisma and leadership effectiveness may result in artificially inflated correlations between the two variables.
In the present study, Rousseau’s (1985) split sample technique was used to control for the common-method variance associated with follower ratings of charismatic leadership and leadership effectiveness. To split the sample according to Rousseau’s technique, which has been used in other charismatic leadership studies (Agle & Sonnenfeld, 1994; Shamir et al., 1998) as a means to control for mono-method bias, half of the follower surveys for a given leader were used to measure leader charisma (n = 163), whereas the other half were used to measure leadership effectiveness and follower openness to change (n = 162). All follower sur- veys for a given leader were shuffled and assigned a number (e.g., 1-4 for a leader with four followers). For leaders with an even number of followers, odd-numbered follower surveys were used to measure charisma, and even-numbered follower surveys were used to measure leadership effectiveness and follower openness to change. For leaders with an odd number of followers, the “extra” follower was assigned to the charisma sample or leadership effective- ness/follower openness-to-change sample on an alternating basis. For example, five leaders in the sample had five follower participants, which generated 25 follower surveys. Thirteen of these surveys were used to measure charismatic leadership, and 12 were used to measure lead- ership effectiveness and follower openness to change. Overall, 163 followers were used to measure charisma, and 162 were used to measure leadership effectiveness and follower openness to change.
Results
Preliminary Analyses
Prior to testing the integrated model of charismatic leadership and the associated hypothe- ses, confirmatory factor analyses were conducted to demonstrate the construct validity of the leader and follower variables. Specifically, two measurement models were tested using Amos software (Arbuckle & Wothke, 1999) and the following conventional indicators of fit: chi- square (χ2), Goodness-of-Fit Index (GFI), Adjusted GFI (AGFI), Root Mean Square Error of Approximation (RMSEA), Incremental Fit Index (IFI), and Comparative Fit Index (CFI). The first measurement model tested the leader survey variables, which included leader social con- trol, emotional expressivity, emotional control, and organizational-change magnitude, postu- lating that each item would load significantly onto their associated scales. On the basis of the conventional standards, the model demonstrated an acceptable level of fit (χ2 = 137.89 [df = 157, p < .01], GFI = .91, AGFI = .86, RMSEA = .06, IFI = .91, CFI = .90).
Groves / Charismatic Leadership 267
The second measurement model tested the follower-report variables, including charismatic leadership (six subscales), openness to change (eight items), and leadership effectiveness (six items). The conventional fit indices showed that the model demonstrated an acceptable level of fit (χ2 = 172.27 [df = 114, p < .01], GFI = .93, AGFI = .88, RMSEA = .05, IFI = .98, CFI = .97), and all charismatic leadership paths were highly significant except for DNMSQ. Given that the sample includes leaders at several organizational levels, it can be reasonably argued that leaders at lower levels in the organization have much less leeway in voicing discontent with the status quo than CEOs or presidents. Furthermore, after dropping this path from the measurement model, the resulting fit indices were minimally improved (χ2 = 162.18 [df = 118, p < .05], GFI = .94, AGFI = .89, RMSEA = .04, IFI = .98, CFI = .98), and the ∆χ2 (10.09, df = 4) was not significant. Thus, the originally specified measurement model was retained for hypothesis testing.
Hypothesis Testing
Table 1 presents the means, standard deviations, and correlations of the leader and follower variables in this study. Consistent with previous research (e.g., Dumdum et al., 2002; Lowe et al., 1996), leaders seen as charismatic by their followers were also rated higher on leader- ship effectiveness (r = .41, p < .01). Although highly significant, this relationship is consider- ably weaker than the results of the Lowe and Dumdum meta-analyses because the present study’s ratings of charisma and leadership effectiveness were from different followers. Not surprisingly, analysis of the charisma-leadership effectiveness relationship using the full fol- lower sample demonstrated a much stronger correlation (r = .71, p < .001) that is consistent with Lowe’s and Dumdum’s mean corrected correlations of .81 and .85, respectively. Charis- matic leadership was also related to social control (r = .27, p < .01) and emotional expressivity (r = .24, p < .05) but unrelated to emotional control (r = .11, n.s.). Furthermore, charismatic leadership correlated with organizational-change magnitude (r = .26, p < .01) and follower openness to change (r = .40, p < .01). Surprisingly, both social control and emotional expressivity were related to organizational-change magnitude (r = .33, p < .01, and r = .25, p < .01, respectively).
The overall fit of the integrated model of charismatic leadership was tested using Amos software for structural equation modeling (Arbuckle & Wothke, 1999). Analysis of fit indices revealed an adequate degree of absolute fit for the specified model: χ2 = 46.98 (df = 30, p < .05), GFI = .96, AGFI = .90, RMSEA = .05, IFI = .93, CFI = .92. Subsequent analysis of the structural model and path coefficients, which represent standardized regression weights, yielded results that support the integrated model’s hypotheses. Figure 1 illustrates the fit esti- mates and path coefficients for the integrated model.
Hypotheses 1-3. As predicted, social control (β = .20, p < .05) and emotional expressivity (β = .22, p < .05) were strongly related to charismatic leadership; however, emotional control did not reach significance (β = .15, p < .10). Not surprisingly, leader gender was strongly related to both social control (β = .20, p < .05) and emotional expressivity (β = .33, p < .01), as
268 Journal of Management / April 2005
269
Ta bl
e 1
D es
cr ip
ti ve
S ta
ti st
ic s
an d
C or
re la
ti on
s
V ar
ia bl
e M
SD 1
2 3
4 5
6 7
8 9
10 11
12 13
14 15
1. L
ea de
r te
nu re
a 5.
59 5.
89 —
2. L
ea de
r ge
nd er
a, e
1. 40
0. 49
.0 1
3. N
on pr
of it/
fo r-
pr of
itf 1.
21 0.
41 .0
4 .1
5 4.
O rg
an iz
at io
na l-
ch an
ge m
ag ni
tu de
d 3.
18 0.
78 .1
0 .0
3 –.
09 5.
S oc
ia l c
on tr
ol a
3. 80
0. 73
.1 9*
.2 0*
.0 9
.3 3*
* 6.
E m
ot io
na l e
xp re
ss iv
ity a
3. 32
0. 64
.3 3*
* .3
3* *
.1 2
.2 5*
* .2
1* 7.
E m
ot io
na l c
on tr
ol a
3. 96
0. 73
–. 13
–. 09
–. 01
–. 02
.1 1
–. 24
* 8.
C ha
ri sm
at ic
le ad
er sh
ip b
4. 72
0. 54
.0 6
.1 6
.0 3
.2 6*
* .2
7* *
.2 4*
.1 1
9. V
is io
n an
d ar
tic ul
at io
nb 4.
79 0.
72 .1
5 .1
1 .0
9 .1
7 .2
3* .2
4* .1
6 .6
5* *
10 . E
nv ir
on m
en ta
l s en
si tiv
ity b
4. 91
0. 52
–. 01
.2 1*
–. 03
.2 2*
.2 1*
.1 3
.2 6*
.6 0*
* .5
9* *
11 . S
en si
tiv ity
to m
em be
r ne
ed sb
5. 11
0. 66
.1 4
.2 0*
.1 6
.1 6
.2 6*
–. 01
.0 1
.5 5*
* .5
3* *
.5 2*
* 12
. P er
so na
l r is
kb 4.
10 0.
87 .0
8 .1
9 .0
8 .1
0 .2
1* .1
5 –.
04 .5
9* *
.4 0*
* .4
1* *
.3 8*
* 13
. U nc
on ve
nt io
na l b
eh av
io rb
3. 27
0. 89
–. 04
–. 02
–. 01
.0 1
.0 2
.0 8
.0 3
.5 5*
* .2
8* *
.2 2*
.1 2
.3 4*
* 14
. D N
M SQ
b 2.
77 0.
81 –.
16 .0
3 –.
24 *
.0 7
.0 9
.0 1
.0 4
.2 4*
.0 1
.1 2
.2 0
.0 4
.3 9*
* 15
. O pe
nn es
s to
c ha
ng ec
5. 78
0. 74
.0 7
.0 1
.1 4
.1 2
.2 1*
.1 5
.1 2
.4 0*
* .4
1* *
.2 2*
.1 2
.2 1*
.2 6*
.1 1
16 . L
ea de
rs hi
p ef
fe ct
iv en
es sc
5. 52
0. 97
.0 6
.1 6
.1 7
.0 8
.1 2
.2 7*
* .0
5 .4
1* *
.3 8*
* .3
2* *
.2 2*
.3 1*
* .2
0* .0
1 .4
1* *
N ot
e: D
N M
SQ =
D oe
s N
ot M
ai nt
ai n
St at
us Q
uo .
a. B
as ed
o n
le ad
er s
ur ve
y (n
= 1
08 ).
b. B
as ed
o n
sp lit
s am
pl e
fo llo
w er
s ur
ve y
(n =
1 63
). c.
B as
ed o
n sp
lit s
am pl
e fo
llo w
er s
ur ve
y (n
= 1
62 ).
d. C
om po
si te
m ea
su re
b as
ed o
n le
ad er
a nd
f ol
lo w
er s
ur ve
ys (
n =
1 08
). e.
1 =
m al
e; 2
= f
em al
e. f.
1 =
n on
pr of
it; 2
= f
or -p
ro fi
t. *p
< .0
5 **
p <
.0 1
ever, leader gender was unrelated to emotional control. Thus, social control and emotional expressivity explained unique variance in charismatic leadership beyond the effects of gender and the other variables in the integrated model. Overall, there was strong support for Hypothe- ses 1 and 2, but no support for Hypothesis 3.
Hypotheses 4 and 5. The hypothesized relationships among follower openness to change, charismatic leadership, and leadership effectiveness were analyzed using the structural equa- tion model presented in Figure 1 and Baron and Kenny’s (1986) suggested mediation analy- ses. The model shows that charismatic leadership was positively related to follower openness to change (β = .40, p < .01), providing strong support for Hypothesis 4. Furthermore, follower openness to change strongly predicted leadership effectiveness (β = .42, p < .01), whereas charismatic leadership demonstrated a weaker relationship with leadership effectiveness (β = .22, p < .05). As anticipated in Hypothesis 5, follower openness to change significantly medi- ated the relationship between charismatic leadership and leadership effectiveness. The pres- ence of follower openness to change in the model considerably reduced the strength of the relationship between charismatic leadership and leadership effectiveness. Barron and Kenny’s suggested mediation analyses also showed that follower openness to change satisfied all three conditions for mediation: (a) Charismatic leadership strongly predicted follower openness to change (β = .36, p < .01), (b) charismatic leadership strongly predicted leadership effectiveness (β = .39, p < .001), and (c) the relationship between charismatic leadership and leadership effectiveness lessened significantly (β = .39, p < .001, to β = .20, p < .05) when fol- lower openness to change was entered into the regression model. Sobel’s (1982) method, which tests the significance and stability of such changes, found that the beta weight change for charismatic leadership was statistically significant (z = 3.15, p < .01). Thus, there was strong support for Hypothesis 5.
Hypothesis 6. The final hypothesis of the integrated model is the moderating effect of orga- nizational-change magnitude on the relationship between charismatic leadership and leader- ship effectiveness. This effect was not supported by the structural model, as the path coeffi- cient for the interaction term was not significant (β = .16, p < .10). Thus, Hypothesis 6 was not supported.
Comparing Plausible Alternative Models
Although the results presented above offer mostly strong support for the hypotheses of the integrated model, they do not rule out the possibility that competing models provide a stronger description of the data. To provide comparative support for the integrated model, several plau- sible theoretical models of charismatic leadership were developed and tested using structural equation modeling. The results of these tests are provided in Table 2. The first step in develop- ing plausible alternatives was to create more parsimonious models of charismatic leadership. To this end, the integrated model (Model 1) was compared with a direct-effects model (Model 2) in which social control, emotional expressivity, and emotional control skills, follower open- ness to change, and leadership effectiveness were postulated to predict charismatic leadership.
270 Journal of Management / April 2005
This model is consistent with the view that follower perceptions of leadership effectiveness may influence follower ratings of leadership behavior (Meindl & Ehlrich, 1987). By compari- son of the fit indices, this model was rejected in favor of the integrated model. Another plausi- ble theoretical model is a simpler alternative to the integrated model in which social control, emotional expressivity, and emotional control skills predict charismatic leadership, which alone is posited to predict leadership effectiveness (Model 3). This model excludes the fol- lower openness-to-change and organizational-change magnitude variables and was also rejected in favor of the integrated model given inadequate fit statistics.
The integrated model was also compared with alternative models of charismatic leadership that postulate theoretically different relationships among the variables of interest. Specifi- cally, Model 4 adopts the perspective that leader social control, emotional expressivity, and emotional control skills predict charisma, whereas the relationship between charismatic lead- ership and leadership effectiveness is mediated by organizational-change magnitude. By com- parison with the integrated model, Model 4 excludes follower openness to change and postu- lates that charismatic leaders must successfully initiate and implement organizational change to be perceived as effective leaders. Although Model 4 provided greater fit than all other com- peting alternatives (Models 2, 3, and 5), it was also rejected in favor of the integrated model due to the significant chi-square change between the two models, ∆χ2 (∆df) = 33.39 (2), p < .01. Finally, Model 5 tests leader social control, emotional expressivity, and emotional control skills as predictors of charismatic leadership and follower openness to change as a contin- gency factor that moderates the relationship between charismatic leadership and leadership effectiveness. This model demonstrated particularly poor fit to the data and was rejected in
Groves / Charismatic Leadership 271
Table 2 Results of Competing Models Tested Against the Integrated Model of
Charismatic Leadership
Model χ2 df GFI AGFI RMSEA IFI CFI ∆χ2 (∆df)a
Model 1: Integrated model 46.98* 30 .96 .90 .05 .93 .92 — Model 2: Total direct effects on
charismatic leadership 864.40*** 37 .71 .65 .16 .60 .65 817.42*** (7) Model 3: Social/emotional
skills → Charismatic leadership → Leadership effectiveness 164.33*** 40 .89 .82 .12 .85 .83 117.35*** (10)
Model 4: Organizational-change magnitude mediation 80.37** 32 .90 .83 .10 .75 .72 33.39** (2)
Model 5: Follower openness-to- change moderation 120.75*** 34 .89 .84 .11 .74 .71 73.77** (4)
Note: GFI = Goodness-of-Fit Index; AGFI = Adjusted Goodness-of-Fit Index; RMSEA = Root Mean Square Error of Approximation; IFI = Incremental Fit Index; CFI = Comparative Fit Index. a. ∆χ2 is calculated by subtracting the chi-squared associated with the integrated model (Model 1) from the chi- squared value associated with the competing model. *p < .05 **p < .01 ***p < .001
favor of the integrated model. Overall, these results provide further support for the hypothe- sized relationships of the integrated model.
Discussion
The purpose of this study was to address several limitations of charismatic leadership the- ory and research, including the exclusion of follower and contextual variables, limited empiri- cal support for mediating processes, and mono-method bias, by testing an integrated model of leader skills, follower attitudes, and contextual variables. Overall, the results of this study sug- gest that charismatic leadership may be best explained by a theoretical model that postulates leader social and emotional skills, followers’attitude toward organizational change, and orga- nizational-change magnitude. The finding that these factors may collectively determine the emergence and efficacy of charismatic leadership in organizations is important for several rea- sons. First, no other empirical study to date has linked these variables into an integrated model and tested it against alternative models of charismatic leadership. Although leadership schol- ars agree that leaders, followers, and contextual factors play critical roles in the charismatic leadership process, the present study represents the first empirical effort to connect these variables via an integrated model.
Second, the results provide much-needed insight regarding the role of followers in charis- matic leadership, particularly the manner in which charismatic leaders influence followers’ attitude toward organizational change and subsequent outcomes. Although the Gardner and Avolio (1998) and Shamir et al. (1993) models provide important theoretical contributions, charismatic leadership theory lacks sufficient attention to how follower characteristics influ- ence the emergence and effectiveness of charismatic leaders. Of all the streams of charismatic leadership research, issues surrounding the role of follower characteristics and how they affect the charismatic relationship between leaders and followers have been largely ignored (Con- ger, 1999). As predicted, followers who rated their leaders as charismatic also reported being open and accepting of the changes occurring in their organization, a finding that is strength- ened by the present study’s use of a split follower sample to control for mono-method bias. Specifically, followers of charismatic leaders were much more open to their work role changes and had a positive outlook on the benefits of the changes to their organization. Given the nature of organizational change and the pervasiveness of employee resistance to change, this finding suggests that charismatic leadership behaviors may engender follower attitudes and beliefs that promote rather than resist organizational change.
Pawar and Eastman (1997) maintained that a necessary ingredient of successful organiza- tional change is leadership behavior capable of overcoming resistance to change by causing followers to experience a “felt need” for the proposed changes. Indeed, the results suggest that charismatic leadership behaviors, including the ability to powerfully articulate an inspiring vision and communicate to followers a sense of ownership of the vision, may affect followers’ openness to organizational change and perceptions of leadership effectiveness. Whereas the preponderance of previous research has examined the direct effects of charismatic leadership (Dumdum et al., 2002; Lowe et al., 1996), the current investigation advances charismatic lead- ership theory by providing empirical support for a key mediating variable. Furthermore, the
272 Journal of Management / April 2005
results from comparing alternative models of charismatic leadership demonstrated far greater support for follower openness to change as mediating the relationship between charisma and leadership effectiveness than for moderating this relationship. This finding represents an important link to the organizational-change literature and related research on employee resis- tance to change (Dirks et al., 1996; Piderit, 2000), which documents the critical role of employee attitudes during organizational change. Given the rate of change facing many of today’s organizations, the finding that follower openness to change is a key mediating variable suggests that leaders should consistently manage their followers’ perceptions of the change process to ensure effective facilitation of organizational change.
Third, the results suggest that charismatic leaders use specific social and emotional skills to enact behaviors that influence followers during periods of organizational change. Although leadership theory has long emphasized the importance of leader interpersonal skills, very little empirical research has established the relationship between possession of basic social and emotional skills and leadership behavior, and no study to date has linked specific skill sets to charismatic leadership. The present study found that leaders with greater social control skills, including self-presentation and social role-playing abilities, were more likely to be perceived as charismatic by their followers. This finding is consistent with previous research on self- monitoring (Snyder, 1974) and behavioral flexibility (Zaccaro, 2002) that suggests effective leaders must be adept in social situations and adjust their behavior to the changing dynamics of their work group. Given charismatic leaders’ reliance on follower needs and values as cues for constructing and articulating their vision (Conger & Kanungo, 1994), social control skills appear to play a critical role in a leader’s ability to enact core charismatic behaviors. Further- more, the need for social control skills may be exacerbated during periods of large-scale orga- nizational change when leaders face the difficult task of convincing both allies and adversaries of their vision’s merits. Unexpectedly, leader social control skills were associated with high organizational-change magnitude, suggesting that a leader’s demonstration of behavioral flexibility may provide key role modeling for followers and create fertile conditions for the implementation of major organizational change. Interestingly, these results may be compared to Riggio and Friedman’s (1983) finding that social control contributes to perceptions of trust- worthiness and credibility. Certainly, perceptions of leader honesty and credibility are critical during periods of significant organizational change because followers stand to gain or lose much from a leader’s proposed changes, including status, power, control over resources, and promotional opportunities.
The present study’s results also suggest that charismatic leaders use key emotional skills to influence followers during periods of organizational change. Leader emotional expressivity skills, including nonverbal expression through eye contact, animated facial expressions, and body/posture gestures, were strongly related to charismatic leadership. This finding indicates that the motivating effects of vision and articulation behaviors, which serve as the core behav- ioral components of many prominent charismatic leadership theories (e.g., Gardner & Avolio, 1998; House, 1977; Shamir et al., 1993), may be dependent on charismatic leaders’ ability to demonstrate genuine commitment to their vision through emotional displays. Contrary to expectations, leader emotional control skills were not significantly related to charismatic lead- ership, suggesting that perceptions of charismatic leadership may not require constant regula-
Groves / Charismatic Leadership 273
tion of emotional displays and communications. Rather, followers may regard leaders who are capable of posing emotions on cue and constantly adjusting nonverbal behavior as chame- leons devoid of authentic feelings and emotions. This finding is consistent with research sug- gesting that follower attributions of leader integrity are partially determined by the consis- tency and sincerity of leader behaviors and whether such behaviors are perceived to benefit organizational goals rather than leader self-interests (Yorges, Weiss, & Strickland, 1999; Yukl, 2002). Overall, these findings are consistent with previous studies (Awamleh & Gardner, 1999; Holladay & Coombs, 1994; Howell & Frost, 1989) that concluded a strong vision delivery style characterized by nonverbal communication is a critical determinant of perceived charisma. However, the findings from these studies are somewhat limited in that vision delivery was manipulated through the use of professional actors in a laboratory setting, whereas the present study examined actual organizational leaders and used validated measures of nonverbal communication skills.
Implications for Organizations
The findings from the present study suggest key implications for executive selection, pro- motion, and development practices. Rather than relying exclusively on interviews for selec- tion of executive-level employees, findings from the present investigation suggest that organi- zations would benefit by augmenting their selection and promotion processes with social and emotional skill assessments. Social control and emotional expressivity skill assessments may also improve assessment center techniques as a means to promote the best managers to execu- tive-level positions. Thus, organizations will capture a more accurate evaluation of a partici- pant’s ability to perform the role of a change agent at the highest levels of the organization. Moreover, many scholars and practitioners have identified social competence and interper- sonal skill development as primary reasons for implementing executive development pro- grams (Fulmer, 1997). Findings from the current study suggest that key social and emotional skills contribute to a leader’s ability to enact the charismatic leadership behaviors that create follower openness to change and affect leadership effectiveness during organizational change. Consequently, businesses and consulting firms stand much to gain by tailoring their executive development programs to the specific interpersonal skill needs of participants, such as creat- ing communication skill profiles for their executive personnel and providing training and development opportunities for individuals lacking certain skill sets.
Limitations and Future Research
The present study’s relatively small sample size limits the strength of the findings dis- cussed above, although the diversity of participating organizations and their respective indus- tries suggests that hypothesized relationships of the integrated model may be a robust phe- nomenon. Certainly, further empirical research is needed to replicate these findings and test the integrated model on different samples of organizational leaders. Another potential limita- tion is follower response bias because leaders identified the direct reports that participated in the study. Thus, leader participants may have selected only those direct reports most likely to
274 Journal of Management / April 2005
provide favorable ratings of leadership effectiveness and charismatic qualities. Future research would benefit from research procedures that ensure each direct report has the oppor- tunity to participate, although requiring participation from all direct reports may decrease the leader response rate. Furthermore, some research suggests that people tend to hold overly favorable views of their abilities and skills (Kruger & Dunning, 1999) and that skill self- assessment data tend to be inflated. However, this concern was partially mitigated by the fact that leaders in the present study were unaware that they were assessing their social and emo- tional skills. Nevertheless, the use of alternative social and emotional skill assessments in future studies would improve our understanding of charismatic leadership.
In addition, the cross-sectional nature of the study’s research design precludes causal con- clusions regarding the relationships among the variables of interest. Several researchers have questioned the causal relationship between leadership and key outcomes (Meindl & Ehrlich, 1987) such that leadership effectiveness may cause attributions of charisma to the leader. However, Shamir’s (1995) analysis of close and distant charismatic leadership argues that the attribution of charisma to a leader by immediate followers cannot be simply explained by fol- lowers’ knowledge of performance cues. Compared with distant followers of charismatic leaders, the followers in the present study were direct reports of their respective leaders and had other sources of information (e.g., direct observation of leader behavior) on which to judge their leader’s qualities and behaviors. Thus, concerns regarding follower romance of leadership and implicit leadership theories are minimized when using direct reports as followers in charismatic leadership studies.
Overall, future investigations of charismatic leadership would greatly benefit from improved measurement of leadership effectiveness, contextual variables, and the charisma– leadership effectiveness relationship. The use of objective measures of leadership effective- ness offers much-needed balance to the field’s dependence on follower perceptual measures of leadership effectiveness (Dumdum et al., 2002; Lowe et al., 1996). As the present study demonstrated, mono-method bias significantly exaggerates the relationship between follower perceptions of leader charisma and leadership effectiveness, suggesting that the Lowe and Dumdum meta-analytic estimates of this relationship may be grossly inflated. Furthermore, future empirical studies stand to gain much by developing more precise measures of the scope and intensity of organizational change. Although results from the present study indicate that charismatic leadership is associated with work groups undergoing large-scale changes, orga- nizational-change magnitude did not significantly moderate the charisma–leadership effec- tiveness relationship. One explanation for this unexpected result is that the valid measurement of highly significant organizational transformations such as culture change may necessitate rigorous qualitative methods such as in-depth interviews and focus group testing rather than survey methodology. Also, results from comparing the integrated model with rival alternative models demonstrated some evidence for organizational-change magnitude mediating the cha- risma–leadership effectiveness relationship (Model 4). Although this model was rejected in favor of the integrated model, leadership and organizational-change research may greatly ben- efit from longitudinal studies in which ratings of leader charisma and leadership effectiveness are obtained at successive stages of large-scale organizational change.
Clearly, the complex and robust nature of charismatic leadership necessitates further empirical research to improve our understanding of this powerful phenomenon. As illustrated
Groves / Charismatic Leadership 275
in the present study, a worthwhile challenge for future researchers is to examine the leader skills and behaviors, follower characteristics, and contextual conditions that collectively explain the efficacy of charismatic leadership in organizations.
References
Agle, B. R., & Sonnenfeld, J. A. 1994. Charismatic chief executive officers: Are they more effective? An empirical test of charismatic leadership theory. Academy of Management Best Papers Proceedings, 2-6.
Arbuckle, J. L., & Wothke, W. 1999. Amos 4.0. Chicago: SmallWaters Corporation. Awamleh, R., & Gardner, W. L. 1999. Perceptions of leader charisma and effectiveness: The effects of vision content,
delivery, and organizational performance. Leadership Quarterly, 10: 345-373. Baron, R. M., & Kenny, D. A. 1986. The moderator-mediator variable distinction in social psychological research:
Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51: 1173- 1182.
Bass, B. M. 1985. Leadership and performance beyond expectations. New York: Free Press. Briggs, S. R., & Cheek, J. M. 1988. On the nature of self-monitoring: Problems with assessment, problems with valid-
ity. Journal of Personality and Social Psychology, 54: 663-678. Conger, J. A. 1999. Charismatic and transformational leadership in organizations: An insider’s perspective on devel-
oping streams of research. Leadership Quarterly, 10: 145-179. Conger, J. A., & Kanungo, R. N. 1994. Charismatic leadership in organizations: Perceived behavioral attributes and
their measurement. Journal of Organizational Behavior, 15: 439-452. Conger, J. A., & Kanungo, R. N. 1998. Charismatic leadership in organizations. Thousand Oaks, CA: Sage. DeVellis, R. F. 1991. Scale development: Theory and applications. Newbury Park, CA: Sage. Dirks, K. T., Cummings, L. L., & Pierce, J. L. 1996. Psychological ownership in organizations: Conditions under
which individuals promote and resist change. Research in Organizational Change and Development, 9: 1-23. Dumdum, U. R., Lowe, K. B., & Avolio, B. J. 2002. A meta-analysis of transformational and transactional leadership
correlates of effectiveness and satisfaction: An update and extension. Transformational and Charismatic Leader- ship, 2: 35-66.
Dvir, T., Eden, D., Avolio, B., & Shamir, B. 2002. Impact of transformational leadership on follower development and performance: A field experiment. Academy of Management Journal, 45: 735-744.
Fulmer, R. M. 1997. The evolving paradigm of leadership development. Organizational Dynamics, 73: 59-70. Gardner, W. L., & Avolio, B. J. 1998. The charismatic relationship: A dramaturgical perspective. Academy of Manage-
ment Review, 23: 32-58. Holladay, S. J., & Coombs, W. T. 1994. Speaking of visions and visions being broken: An exploration of the effects of
content and delivery on perceptions of leader charisma. Management Communication Quarterly, 8: 165-189. House, R. J. 1977. A 1976 theory of charismatic leadership. In J. G. Hunt & L. L. Larson (Eds.), Leadership: The cut-
ting edge: 189-207. Carbondale: Southern Illinois University Press. Howell, J. M., & Frost, P. 1989. A laboratory study of charismatic leadership. Organizational Behavior and Human
Decision Processes, 43: 243-269. James, L. R., Demaree, R. G., & Wolf, G. 1984. Estimating within-group interrater reliability with and without
response bias. Journal of Applied Psychology, 69: 85-98. Kruger, J., & Dunning, D. 1999. Unskilled and unaware of it: How difficulties in recognizing one’s own incompetence
lead to inflated self-assessments. Journal of Personality and Social Psychology, 6: 1121-1134. Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. 1996. Effectiveness correlates of transformational and
transactional leadership: A meta-analytic review. Leadership Quarterly, 7: 385-425. Mayer, J. D., & Salovey, P. 1997. In P. Salovey & D. Sluyter (Eds.), Emotional development and emotional intelli-
gence: Implications for educators: 3-31. New York: Basic Books. Meindl, J. R., & Ehlrich, S. B. 1987. The romance of leadership and the evaluation of organizational performance.
Academy of Management Journal, 30: 91-109. Miller, V. D., Johnson, J. R., & Grau, J. 1994. Antecedents to willingness to participate in a planned organizational
change. Journal of Applied Communication Research, 22: 59-80.
276 Journal of Management / April 2005
Nadler, D. A., & Tushman, M. L. 1990. Beyond the charismatic leader: Leadership and organizational change. Cali- fornia Management Review, Winter: 77-96.
Pawar, B. S., & Eastman, K. K. 1997. The nature and implications of contextual influences on transformational leader- ship: A conceptual examination. Academy of Management Review, 22: 80-109.
Piderit, S. K. 2000. Rethinking resistance and recognizing ambivalence: A multidimensional view of attitudes toward an organizational change. Academy of Management Review, 25: 783-794.
Riggio, R. E. 1986. Assessment of basic social skills. Journal of Personality & Social Psychology, 51(3): 649-660. Riggio, R. E. 1989. Social Skills Inventory manual. Palo Alto, CA: Consulting Psychologists Press. Riggio, R. E. 1992. Social interaction skills and nonverbal behaviors. In R. S. Feldman (Ed.), Applications of nonver-
bal behavioral theories and research: 3-30. Hillsdale, NJ: Lawrence Erlbaum. Riggio, R. E., & Friedman, H. S. 1983. Individual differences and cues to deception. Journal of Personality and Social
Psychology, 45: 899-915. Rousseau, D. M. 1985. Issues of levels in organizational research: Multi-levels and cross level perspectives. In L. L.
Cummings & B. M. Staw (Eds.), Research in organizational behavior: 43-72. Greenwich, CT: JAI. Shamir, B. 1995. Social distance and charisma: Theoretical notes and an exploratory study. Leadership Quarterly, 6:
19-47. Shamir, B., House, R. J., & Arthur, M. B. 1993. The motivational effects of charismatic leadership: A self-concept
based theory. Organization Science, 4: 577-594. Shamir, B., & Howell, J. M. 1999. Organizational and contextual influences on the emergence and effectiveness of
charismatic leadership. Leadership Quarterly, 10: 257-283. Shamir, B., Zakay, E., Breinin, E., & Popper, M. 1998. Correlates of charismatic leader behavior in military units: Sub-
ordinates’ attitudes, unit characteristics, and superiors’ appraisals of leader performance. Academy of Manage- ment Journal, 41: 387-409.
Snyder, M. 1974. The self-monitoring of expressive behavior. Journal of Personality and Social Psychology, 30: 526- 537.
Sobel, M. 1982. Asymptotic confidence intervals for indirect effects in structural equation models. In S. Leinhart (Ed.), Sociological methodology: 290-312. San Francisco: Jossey-Bass.
Sosik, J. J., & Dworakivsky, A. C. 1998. Self-concept based aspects of the charismatic leader: More than meets the eye. Leadership Quarterly, 9: 503-526.
Tucker, J. S., & Riggio, R. E. 1988. The role of social skills in encoding posed and spontaneous facial expressions. Journal of Nonverbal Behavior, 12: 87-97.
Waldman, D. A., Ramirez, G. G., House, R. J., & Puranam, P. 2001. Does leadership matter? CEO leadership attributes and profitability under conditions of perceived environmental uncertainty. Academy of Management Journal, 44: 134-143.
Weber, M. 1947. Max Weber: The theory of social and economic organization. Glencoe, IL: Free Press. Yorges, S. L., Weiss, H. M., & Strickland, O. J. 1999. The effect of leader outcomes on influence, attributions, and per-
ceptions of charisma. Journal of Applied Psychology, 84: 428-436. Yukl, G. 1999. An evaluation of conceptual weaknesses in transformational and charismatic leadership theories. Lead-
ership Quarterly, 10: 285-305. Yukl, G. 2002. Leadership in organizations (5th ed.). Upper Saddle River, NJ: Prentice Hall. Zaccaro, S. J. 2002. Organizational leadership and social intelligence. In R. E. Riggio, S. E. Murphy, & F. J. Pirozzolo
(Eds.), Multiple intelligences and leadership: 29-54. Mahwah, NJ: Lawrence Erlbaum.
Biographical Note
Kevin S. Groves is an assistant professor of management and faculty director of the PepsiCo/Frito-Lay Leadership Center at California State University, Los Angeles. He received his Ph.D. in organizational behavior from Claremont Graduate University. His current research interests include executive leadership development, leader social and emo- tional intelligence, and charismatic leadership.
Groves / Charismatic Leadership 277