Transformational and charismatic leadership: Assessing the convergent, divergent and criterion validity of the MLQ and the CKS

The Leadership Quarterly 18 (2007) 121 – 133 www.elsevier.com/locate/leaqua

Transformational and charismatic leadership: Assessing the convergent, divergent and criterion validity of the MLQ and the CKS Jens Rowold a,⁎, Kathrin Heinitz b a

University of Münster, Psychologisches Institut II, Fliednerstrasse 21, 48149 Münster, Germany b Free University of Berlin, Germany

Abstract This study aimed at empirically clarifying the similarities and differences between transformational, transactional, and charismatic leadership. More specifically, the convergent, divergent, and criterion validity of two instruments, the Multifactor Leadership Questionnaire (MLQ-5X) and the Conger and Kanungo Scales (CKS), was explored. It was found that transformational and charismatic leadership showed a high convergent validity. Moreover, these leadership styles were divergent from transactional leadership. With regard to criterion validity, subjective (e.g. satisfaction) as well as objective (profit) performance indicators were assessed. Firstly, results indicated that transformational as well as charismatic leadership augmented the impact of transactional leadership on subjective performance. In addition, transformational and charismatic leadership both contribute unique variance to subjective performance, over and above the respective other leadership style. Secondly, transformational leadership had an impact on profit, over and above transactional leadership. This augmentation effect could not be confirmed for charismatic leadership. Furthermore, transformational leadership augmented the impact of both transactional and charismatic leadership on profit. Implications for leadership theory and practice are discussed. © 2007 Elsevier Inc. All rights reserved. Keywords: Transformational leadership; Transactional leadership; Charismatic leadership; Profit

Over the last two decades a new genre of leadership theory, alternatively labeled as “charismatic” (Conger & Kanungo, 1998; Hunt, Boal, & Dodge, 1999; Shamir, House, & Arthur, 1993), “visionary” (Sashkin, 1988), or “transformational” (Avolio, Bass, & Jung, 1999; Bass, 1985), has emerged. Among these, two fields of research have gained considerable interest. First, within transformational leadership, leaders emphasize higher motive development, and arouse followers' motivation by means of creating and representing an inspiring vision of the future (Bass, 1997). Second, charismatic leadership (Conger & Kanungo, 1998) describes why followers identify with their respective leader. The positive effects of transformational and charismatic leadership on several organizational outcomes underscore their relevance (Bass, 1998; Yammarino, Spangler, & Bass, 1993). While these positive effects have been proven in a wide range of applied settings (Dumdum, Lowe, & Avolio, 2002; Judge & Piccolo, 2004; Lowe, Kroeck, & Sivasubramaniam, 1996), the elementary field of research concerning the constructs of transformational and charismatic ⁎ Corresponding author. Tel.: +49 251 8331377; fax: +49 251 8334104. E-mail address: [email protected] (J. Rowold). 1048-9843/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.leaqua.2007.01.003

122

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

leadership still needs further attention (Avolio & Yammarino, 2002; Hunt & Conger, 1999; Shamir et al., 1993). The urgent call for research in this area is summarized by Yukl (1999): One of the most important conceptual issues for transformational and charismatic leadership is the extent to which they are similar and compatible. […] The assumption of equivalence has been challenged by leadership scholars […] who view transformational and charismatic leadership as distinct but partially overlapping processes. (p. 298 ff.). The instruments to assess these constructs add to the confusion about the underlying meaning of transformational and charismatic leadership. As an example, the Multifactor Leadership Questionnaire (MLQ-5X; Bass & Avolio, 2000), which is used to assess transformational leadership, includes five subscales of transformational leadership. Of these five, three subscales were combined into one factor called charisma in earlier versions of the instrument. In turn, the empirical leadership literature uses the terms transformational and charismatic leadership inconsistently and interchangeably. Following the arguments made by Yukl (1999) and other leadership researchers (Judge, 2005), the present study aimed at explicating and demonstrating the relationships of the constructs of transformational and charismatic leadership, as well as their effect on individual (subjective) and organizational (objective) outcomes. In the following section, we compare and contrast two widely used instruments for assessing these leadership styles, namely the MLQ-5X and the Conger–Kanungo Scales (CKS). We focus on these instruments as they both represent elaborated theories and have adequate psychometric properties. We are aware, however, that other approaches to transformational and charismatic leadership may result in a more elaborated understanding of the subject as well (Rafferty & Griffin, 2004). 1. Comparison of MLQ and CKS 1.1. Similarities Both the MLQ and CKS belong to what has been labeled “neo-charismatic” leadership theories (Antonakis & House, 2002). Fundamental to the theories of Bass (1985) and Conger & Kanungo (1998) is the representation and articulation of a vision by the leader (Sashkin, 2004). As a long-term attempt to change followers' attitudes, selfconcepts (House & Shamir, 1993) and motives, this vision is rooted in commonly-held ethics and values (Bass & Steidlmeier, 1999). The ethical foundation of the vision is fundamental to both Bass' and Conger and Kanungo's theories. Thus, they focus on socialized as opposed to personalized charisma (Howell & Avolio, 1992). Socialized charismatic leaders use their abilities to achieve benefits for all followers, and not just for their own benefit. Both transformational and charismatic leaders are agents of change. In addition to the formulation of a vision, strong emotional ties between the leader and the led are necessary in order to change followers' belief systems and attitudes. In addition, if the leader is a trustworthy model and represents a code of conduct, transformation occurs more easily. As a consequence of the leader's charismatic qualities and behaviors, followers identify with the leader. In turn, values and performance standards are more likely to be adapted by followers. Finally, transformational and charismatic leaders foster performance beyond expectations. These similarities between the theories proposed by Bass (1985) and by Conger & Kanungo (1998) highlight the fact that they share at least one basic assumption (cf. Antonakis & House, 2002). In his original work on transformational leadership, Bass (1985) proposed that charisma is the main component of transformational leadership. However, several important differences between these theories are clearly evident. 1.2. Differences The main difference between the MLQ and CKS is that each is based on its own conceptualization of charismatic/ transformational leadership (Antonakis & House, 2002; House & Shamir, 1993). As a consequence, these measures include different sets of leadership scales. We turn to each of these instruments in the following paragraphs. Central to the theory of Bass (1985) is the distinction between transactional and transformational leadership. Over the last two decades, empirical research resulted in several subscales for the assessment of these two multifaceted constructs. Nine different subscales of transformational and transactional leadership are measured by the MLQ-5X (Avolio & Bass, 2004; Bass & Avolio, 2000). Whereas in the last decades the factorial validity of

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

123

the MLQ-5X was under consideration (Den Hartog & Van Muijen, 1997; Yukl, 1999), a recent large sample analysis (Antonakis, Avolio, & Sivasubramaniam, 2003) supported the nine-factor model of the MLQ-5X, which is described in Table 1. Whereas the MLQ assesses leadership behavior at a single point in time, the CKS views leadership as a process over time. Thus, the two theories focus on different time frames. Conger & Kanungo (1998) developed a model of charismatic leadership which focuses on three distinct stages of the leadership process. In the first stage, the leader assesses the environment as to growth opportunities for the respective organization (cf. Table 1, Sensitivity to the Environment, SE) and carefully evaluates his/her followers' needs (Sensitivity to Members' Needs, SMN). In stage two, the respective leader formulates a strategic vision which is constantly presented in an inspiring way (Strategic Vision and Articulation, SVA). Finally, in the third stage, the leader provides a role model by demonstrating Personal Risk (PR) and Unconventional Behavior (UB) to his/her followers. By means of PR and UB, the leader builds up followers' trust and commitment (Conger, Kanungo, Menon, & Mathur, 1997). In sum, these five subscales form the Conger and Kanungo Scale (CKS) of charismatic leadership. A combined measure of charisma will be labeled “CKS-CH ” in the present study. At the subscale level, additional similarities between transformational (MLQ) and charismatic (CKS) leadership can be noted (cf. Table 1). In both the MLQ scale IS and the CKS scale SE, the leader questions the status quo and seeks out new ways to solve problems. Furthermore, articulating a vision and inspiring the followers are contents of both the MLQ scale IM and the CKS scale SVA. Thirdly, understanding and evaluating the followers' needs is a subject of both the MLQ IC and the CKS SMN scales. Unique to the CKS are the subscales of PR and UB. In contrast to the MLQ, no Table 1 Comparison of the subscales of MLQ-5X and CKS MLQ

CKS

Transformational leadership

Charismatic leadership

Inspirational motivation (IM). Articulation and representation of a vision; leaders' optimism and enthusiasm.

Sensitivity to the environment (SE). The leader assesses the environment for growth opportunities for his/her respective organization, criticizes the status quo and proposes radical changes in order to achieve organizational goals.

Idealized influence attributed (IIa). Instilling pride in and respect for the leader; the followers identify with the leader.

Sensitivity to members' needs (SMN). The leader carefully evaluates his/her followers' needs.

Idealized influence behavior (IIb). Representation of a trustworthy and energetic role model for the follower.

Strategic vision and articulation (SVA). The leader formulates a strategic vision for the respective organization. It is constantly presented to followers in an inspiring way.

Intellectual stimulation (IS). Followers are encouraged to question established ways of solving problems.

Personal risk (PR). Presenting self-confidence, demonstrating belief in the potential outcome of the vision.

Individualized consideration (IC). Understanding the needs and abilities of each follower; developing and empowering the individual follower.

Unconventional behavior (UB). Leaders build trust and commitment in followers; provides a role model for followers.

Transactional leadership Contingent reward (CR). Defining the exchanges between what is expected from the follower and what the follower will receive in return. Active management-by-exception (AMbE). In order to maintain current performance status the focus is on detecting and correcting errors, problems or complaints. Management-by-exception passive (MbEP). Addressing problems only after they have become serious. Nonleadership Laissez-faire (LF). Absence of leadership behavior.

124

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

subscale of the CKS assesses transactional leadership behaviors. All in all, discussing the MLQ and CKS scales reveals that these instruments do cover different but also similar facets of leadership behavior (Yukl, 1999). 2. Leadership styles and performance: The augmentation effect Several meta-analyses provided evidence for the criterion-related validity of transformational and charismatic leadership (DeGroot, Kiker, & Cross, 2000; Dumdum et al., 2002; Fuller, Patterson, Hester, & Stringer, 1996; Judge & Piccolo, 2004). These two constructs consistently showed a positive impact on both subjective (Lowe, Kroeck, & Sivasubramaniam, 1996) and objective (Barling, Weber, & Kelloway, 1996; Geyer & Steyrer, 1998) performance criteria. In addition, transactional leadership was positively related to outcome criteria, although to a lesser degree than transformational and charismatic leadership. A recurring theme within transformational leadership research literature is the augmentation hypothesis. In order to evaluate the two diverging leadership styles, it was hypothesized by Bass (1985) that transformational leadership adds unique variance beyond that of transactional leadership for predicting organizational outcomes. This augmentation effect of transformational leadership was found when predicting subjective (e.g., Waldman, Bass, & Yammarino, 1990) and objective measures of performance (e.g., Geyer & Steyrer, 1998), providing support for the criterion validity of transformational leadership. However, the augmentation effect has not been tested for charismatic leadership (CKS). Thus, implementing hierarchical regression analyses, the present study tested the augmentation effect for both charismatic and transformational leadership. 3. Overview of the present study In combination, we intended to provide insights into both the interrelationship as well as the effects of partially rivaling leadership constructs (Avolio & Yammarino, 2002; Yukl, 1999). More specifically, the present study had three objectives. First, using data from an empirical investigation, we investigated the nomological network of charismatic, transformational, and transactional leadership. We proposed that transformational and charismatic leadership share some attributes, are highly redundant, and show a convergent validity (H1a). In comparison, these two leadership styles show a divergent validity to transactional leadership (i.e. transactional leadership shows weaker correlations with transformational and charismatic leadership than these styles show with each other [H1b]). Second, we focused on the criterion validity of transformational, transactional, and charismatic leadership. On the side of subjective performance, we assessed subordinates' perception of effectiveness, their extra effort, and their satisfaction. On the side of objective performance, we focused on profit. Transactional leadership shows negative (Howell, Neufeld, & Avolio, 2005) as well as positive correlations to objective outcomes (Geyer & Steyrer, 1998). However, the meta-analytic results of Lowe et al. (1996) as well as Judge & Piccolo (2004) suggest that the correlation of transactional leadership and objective outcomes are positive, but to a lesser degree than with transformational leadership. Therefore, the following hypotheses will be tested: Transactional (H2a), transformational (H2b), and charismatic (H2c) leadership are positively correlated to subjective performance. Transactional (H3a), transformational (H3b), and charismatic (H3c) leadership are positively correlated to objective performance. In both cases, transactional leadership shows weaker correlations with the performance criteria than transformational and charismatic leadership. Third, for a more rigorous test of the impact these constructs have on both subjective and objective performance indicators, regression analysis was implemented. We proposed that both transformational and charismatic leadership augment the impact of transactional leadership on subjective and objective performance. Specifically, we began by testing if transformational leadership has a stronger relationship to subjective and objective performance criteria than transactional leadership. In the next analysis, we tested if the augmentation effect can be found for charismatic leadership as well. We proposed that charismatic leadership augments the effect of transactional leadership. Finally, we compared charismatic and transformational leadership directly. We tested if transformational leadership augments charismatic leadership and vice versa. This last step in the criterion validity analysis is an extension to convergent validity: If the results show that transformational and charismatic leadership are not completely redundant, it will be interesting to know which leadership style shows a stronger relationship to performance. In sum, with these analyses we aimed at clarifying the ways in which different leadership models relate to each other as well as to important organizational outcomes. It has been noted by leadership scholars that contextual influences are to be taken into consideration when examining the relationship between leadership styles and organizational outcomes (Bass, 1985; Waldman et al., 1990). More specifically, it was hypothesized that contextual variables such as size of department, type of organization, and

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

125

hierarchical level moderated the relationship between leadership styles and outcome criteria (Antonakis & House, 2002; Yukl, 2002). In the present study, we were able to include the focal leaders' hierarchical level as one of the more important contextual conditions (cf. Lowe et al., 1996), as for example managerial functions and experiences may vary with hierarchical levels of the leader, likewise the experiences of the respondents (Bass, Waldman, Avolio, & Bebb, 1987). In turn, within all the regression analyses, we controlled for focal leaders' hierarchical level. Compared to transactional leadership it was hypothesized that transformational and charismatic leadership explains additional variance in subjective (H4a) and objective (H4b) performance, whilst controlling for focal leaders' hierarchical level. Also, it was hypothesized that both transformational and charismatic leadership augment the impact of the respective other leadership style on subjective (H5a) and objective (H5b) performance whilst controlling for focal leaders' hierarchical level. 4. Method 4.1. Participants Participants were employees in a large public transport company in Germany. From a total of N = 298 employees, 220 responded (response rate = 73.8%). These employees assessed the leadership style of their respective direct leader (i.e. supervisor) who led one of the companys' 45 branches. At least two employees reported to their respective leader. The branches are hierarchically nested so that leaders of four hierarchical levels were rated. Seven employees (3.4%) reported to top executives, 37 (17.9%) to upper management, 88 (42.5%) to middle management, and 75 (33.8%) to first-level supervisors. The mean age of the participants was 40 years (SD = 7.1); 73% were male. 4.2. Questionnaire design 4.2.1. Leadership style Transformational and transactional leadership was assessed using a German translation of the MLQ 5-X (Bass & Avolio, 2000; cf. Note 1; Yammarino, Spangler, & Dubinsky, 1998). In contrast to earlier German translations of the MLQ-5X (Felfe & Goihl, 2002; Geyer & Steyrer, 1998), this translation proved to measure the nine dimensions of transformational and transactional leadership described above with good psychometric properties (Rowold, 2004, 2005). For assessing charismatic leadership style, the Conger and Kanungo Scales (CKS; Conger & Kanungo, 1998) were carefully translated from English to German by a professional and then backtranslated by an English native speaker, both experts in the field of organizational psychology (Behling & Law, 2000). For each of the MLQ and CKS items, participants rated the frequency of leaders' behavior on a 5-point rating scale (1 = not at all, 5 = frequently, if not always). The subscale values represent the unweighted means of the respective items; combined measures represent the mean values of the respective subscales. In addition, participants indicated their focal leaders' hierarchical level within the organization, ranging from 1 (upper management) to 4 (second-level supervisors). As we did not have any a priori expectation concerning the differential affect of the transformational or charismatic subscales, we used combined measures that were consistent with recent empirical studies (Avolio, Zhu, Koh, & Bhatia, 2004). Transformational, transactional and charismatic leadership show satisfactory α-levels (see Table 3). We therefore assume that our results are based on reliable construct assessment. In order to test the factorial validity of the leadership instruments, confirmatory factor analyses were performed. For both instruments several alternative models found in prior research (cf. Antonakis et al., 2003; Conger & Kanungo, 1998) were tested. Several fit indices were used to evaluate the model fit (Medsker, Williams, & Holahan, 1994) and the χ2-Difference-Test was implemented to compare rivaling factor models. Results from Table 2 demonstrate that for the MLQ-5X, the nine factor model showed the best fit, replicating recent research (Rowold, 2005). As for the CKS, the five factor model advocated by Conger and colleagues was supported. These results seem to justify the use of the two leadership instruments. 5. Performance indicators 5.1. Subjective measures The present study included three subjective performance measures. Respondents were asked to rate their Extra Effort (EEF ), the Effectiveness of leaders' behavior (EFF ), and their Satisfaction (SAT ) with the respective

126

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

Table 2 Results of confirmatory factor analyses

MLQ Null-model One factor modela Three factor modelb Six factor modelc Nine factor modeld CKS Null-model One factor modela Five factor modeld

χ2

df

AIC

TLI

CFI

RMSEA

Δχ2

Δdf

21701 1797 1677 1350 1247

666 594 591 579 558

21773 2013 1899 1596 1535

NA 0.94 0.94 0.96 0.96

NA 0.94 0.95 0.96 0.97

0.38 0.10 0.09 0.08 0.07

20454⁎⁎⁎ 550⁎⁎⁎ 430⁎⁎⁎ 103⁎⁎⁎

108 36 33 21

14509 997 533

276 230 220

14555 1135 691

NA 0.94 0.97

NA 0.95 0.98

0.46 0.12 0.08

13976⁎⁎⁎ 464⁎⁎⁎

56 10

Note: a) All items load on one leadership factor; b) transformational, transactional, and laissez-faire leadership; c) in contrast to the nine factor model, inspirational motivation and both idealized influence factors were combined into one “charisma” factor; d) factor model as described in Table 1 (target model); NA = not available; Δχ2 was calculated by subtracting model χ2 from that of the nine-factor (MLQ) or five-factor (CKS) model; likeways, Δdf was calculated by subtracting model df from that of the nine-factor (MLQ) or six-factor (CKS) model; ⁎⁎⁎ p b .001.

leader. These scales ranged from (1) very low to (5) very high and are part of the MLQ-5X (Bass & Avolio, 2000). 5.2. Objective measures The participating transport company had a clearly defined system of evaluating the financial performance of the 45 branches. Prior to each year, the top management carefully defined the budget for each branch individually. This process was carried out in close collaboration with the respective branch leader and took into account —parameters such as the size of the respective branch and its annual performance goal. At the end of each year, the gains and losses of each branch were calculated, yielding the final financial performance. The difference between these two measures (profit) was an ideal objective performance indicator and made it easy to communicate each branch's success in financial terms. In sum, each branch was responsible for its budget and could directly contribute to the company's financial performance goals. For N = 39 branches, profit data was available from company records and standardized for further analyses. 6. Procedure The questionnaires were administered to participants during work time; full anonymity was assured. All N = 220 participants filled out both the MLQ (which included the subjective performance criteria) and the CKS. The survey took place during the last 3 months of the same year for which the objective performance measures were collected. 7. Results 7.1. Convergent and divergent analysis Using correlational analysis, we explored the relationships between transformational (TF), transactional (TA — both measured with the MLQ) and charismatic leadership (as measured by the CKS, CKS-CH). TA is seen as divergent, TF and CKS-CH as convergent leadership styles. Table 3 shows the intercorrelations of constructs (individual level, N = 220). 7.1.1. Convergent validity The high correlation between TF and CKS-CH (r = .88, p b .01) supports convergent validity between these constructs (hypothesis 1a).

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

127

Table 3 Means, standard deviations, intercorrelations, and internal consistency reliabilities for key study variables at the individual level (N = 220)

1. Level 2. Transactional leadership 3. Transformational leadership 4. CKS-CH 5. Extra effort 6. Effectiveness 7. Satisfaction

M

SD

1

2

3

4

5

6

7

3.11 2.90 3.35 3.26 3.19 3.56 3.47

0.94 0.48 0.65 0.69 0.84 0.76 0.86

– − .04 − .31⁎⁎ − .32⁎⁎ − .25⁎⁎ − .31⁎⁎ − .31⁎⁎

.85 .57⁎⁎ .52⁎⁎ .45⁎⁎ .41⁎⁎ .33⁎⁎

.87 .88⁎⁎ .77⁎⁎ .79⁎⁎ .74⁎⁎

.87 .76⁎⁎ .81⁎⁎ .75⁎⁎

.81 .74⁎⁎ .65⁎⁎

.70 .79⁎⁎

.83

Note: Level = focal leaders' hierarchical level; TA = transactional leadership; TF = transformational leadership; CKS-CH = charismatic leadership as measured by the Conger and Kanungo Scales, EEF = extra effort; EFF = effectiveness; SAT = satisfaction. Values along the diagonal represent internal consistency estimates (Cronbach's alpha). ⁎⁎p ≤ .01 (one-tailed).

7.1.2. Divergent validity Both the correlations between TA and TF (r = .57, p b .01) as well as between TA and CKS-CH (r = .52, p b .01) were statistically significant. Further statistical analysis was deemed necessary to check if these divergent validity coefficients were significantly smaller than convergent validity. Meng, Rosenthal, & Rubin's (1992) approach was utilized to test differences between dependent correlations. It was found that 1) rTF,CH N rTF,TA (Z = 1.90, p b .05) and 2) rTF,CH N rCH,TA (Z = 2.35, p b .01). Thus, divergent validity coefficients were significantly smaller than convergent validity, supporting hypothesis 1b. 8. Leadership style and performance Table 3 reveals that all of the leadership styles were significantly associated with indicators of subjective performance. Thus, hypotheses 2a, 2b, and 2c were supported by the data. It should be noted that focal leaders' hierarchical level was significantly correlated with transformational and charismatic leadership style as well as the three subjective performance indicators. Thus, higher-level leaders exhibit more of these leadership behaviors which are, in turn, associated with subjective performance. These results highlight the importance of contextual variables such as respondents' level (Lowe et al., 1996). As was stated in the introduction, we controlled for level in the regression analyses which will be presented next. 9. Augmentation analyses The results reported above show that transformational and charismatic leadership are highly convergent. However, as already mentioned, the two questionnaires (MLQ and CKS) both measure facets of leadership behavior that are not part of the respective other instrument. Therefore, on the one hand, it was examined if the facets of transformational and charismatic leadership augment transactional leadership and, on the other hand, if they augment each other, i.e. if they explain genuine variance in subjective and objective performance indicators. 9.1. Subjective performance Hierarchical regression analysis was performed in order to assess the impact of the hierarchical level and different leadership styles on subjective performance indicators. Each step within these hierarchical regressions is represented by a separate row in Table 4. Throughout the analyses, the hierarchical level was entered first into the regression. In the next steps, we separately entered different leadership styles into the regression analysis. The results show that, for the subjective performance indicators, transformational leadership adds between 29 and 39% variance beyond that of transactional leadership (analysis #1). Also, charismatic leadership adds comparable amounts of variance (29–40%) to transactional leadership (analysis #2). Taken together, these results support hypothesis 4a.

128

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

Table 4 Summary of simultaneous regression analyses (standardized Betas) for leadership styles predicting subjective measures of performance at the individual level (N = 220) Analysis No.

(1)

(2)

(3)

(4)

Extra effort

Level TA TF Level TA CKS-CH Level TA TF CKS-CH Level TA CKS-CH TF

Effectiveness

Satisfaction

Beta

R

ΔR

Beta

R

ΔR

Beta

R2

ΔR2

− .02 .03 .74⁎⁎ .42⁎ .01 .38 .00 .01 .43⁎⁎ .37⁎⁎ .00 .01 .37⁎⁎ .43⁎⁎

.12 .29 .59 .12 .29 .58 .12 .29 .59 .62 .12 .29 .58 .62

.12⁎⁎ .18⁎⁎ .29⁎⁎ .12⁎⁎ .18⁎⁎ .29⁎⁎ .12⁎⁎ .18⁎⁎ .29⁎⁎ .03⁎⁎ .12⁎⁎ .18⁎⁎ .29⁎⁎ .04⁎⁎

− .06 − .04 .79⁎⁎ − .08 − .01 .78⁎⁎ − .04 − .07 .38⁎⁎ .49⁎⁎ − .04 − .07 .49⁎⁎ .38⁎⁎

.16 .30 .63 .16 .30 .66 .16 .30 .63 .69 .16 .30 .66 .69

.16⁎⁎ .14⁎⁎ .33⁎⁎ .16⁎⁎ .14⁎⁎ .36⁎⁎ .16⁎⁎ .14⁎⁎ .33⁎⁎ .05⁎⁎ .16⁎⁎ .14⁎⁎ .36⁎⁎ .03⁎⁎

.08⁎⁎ −.15⁎⁎ .86⁎⁎ .05 −.10 .82⁎⁎ .10 −.17⁎⁎ .46⁎⁎ .48⁎⁎ .10⁎ −.17⁎⁎ .48⁎⁎ .46⁎⁎

.08 .17 .57 .08 .17 .58 .08 .17 .57 .62 .08 .17 .58 .62

.08⁎⁎ .10⁎⁎ .39⁎⁎ .08⁎⁎ .10⁎⁎ .40⁎⁎ .08⁎⁎ .10⁎⁎ .39⁎⁎ .05⁎⁎ .08⁎⁎ .10⁎⁎⁎ .40⁎⁎ .04⁎⁎

2

2

2

2

Note: Level = focal leaders' hierarchical level; TF = transformational leadership; TA = transactional leadership; CKS-CH = charismatic leadership as measured by the Conger and Kanungo Scales. ⁎⁎ p ≤ .01; ⁎ p ≤ .05.

Furthermore, the results show that both transformational (analysis #3) and charismatic leadership (analysis #4) add unique variance (3–5%) for subjective performance beyond that of the respective other leadership style. These results lend support to hypothesis 5a. 9.2. Objective performance The impact of leadership styles on profit is analyzed at the branch level. Before the data was aggregated at the branch level, it was tested if different subordinates attributed the same leadership styles to their mutual supervisor (i.e. branch leader). Following the recommendations made by McGraw & Wong (1996), interrater agreement (ICC1 and ICC2) and within-group agreement indices (rwg) were calculated (James, Demaree, & Wolf, 1984). Table 5 indicates that the raters highly agreed on the three leadership scales. Consequently, leadership ratings were aggregated at the branch level. Descriptive statistics of key study variables are presented in Table 5. As can be seen, profit is significantly correlated to transformational leadership. While this result supports hypothesis 3b, the hypotheses 3a and 3c were not supported. With regard to the augmentation hypothesis, the analyses of objective performance paralleled the analyses of subjective performance. Hierarchical regression analyses were implemented to assess the impact of different leadership styles on profit, while controlling for the focal leaders' hierarchical level (cf. Table 6). Transformational leadership augmented the impact of transactional leadership on profit: 14% variance in profit is due to transformational leadership, over and above transactional leadership (analysis #1). In contrast, charismatic leadership did not augment transactional Table 5 Descriptive statistics for key study variables at the branch level (N = 39)

1. Level 2. TA 3. TF 4. CKS-CH 5. Profit

M

SD

α

ICC1

ICC2

Mean rwg

SD rwg

3.03 2.89 3.34 3.20 0.00

0.96 0.37 0.55 0.50 1.00

– .79 .93 .94 –

– .51 .67 .46 –

– .67 .80 .63 –

– .95 .96 .95 –

– .07 .06 .07 –

1

2

3

4

− .15 − .52⁎⁎ − .55⁎⁎ .21

.69⁎⁎ .69⁎⁎ .21

.93⁎⁎ .26⁎

.15

Note: Level = focal leaders' hierarchical level within the organization; TA = transactional leadership; TF = transformational leadership; CKS-CH = charismatic leadership as measured by the Conger and Kanungo Scales; EEF = extra effort; EFF = effectiveness; SAT = satisfaction, Profit = (standardized) objective profit. ⁎ p ≤ .05; ⁎⁎ p ≤ .01 (one-tailed).

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

129

Table 6 Summary of simultaneous regression analysis (standardized Betas) for leader-ship styles predicting objective measures of performance at the branch level (N = 39) Analysis No.

(1)

(2)

(3)

(4)

Leadership style

Level TA TF Level TA CKS-CH Level TA TF CKS-CH Level TA CKS-CH TF

Profit Beta

R2

ΔR2

.52⁎⁎ − .15 .63⁎ .42⁎ .01 .38 .47⁎ − .10 .92⁎ − .38 .47⁎ − .10 − .38 .92⁎

.05 .10 .24⁎ .05 .10 .15 .05 .10 .24 .26 .05 .10 .15 .26

.05 .06 .14⁎ .05 .06 .05 .05 .06 .14⁎ .02 .05 .06 .05 .11⁎

Note: Level = focal leaders' hierarchical level; TF = transformational leadership; TA = transactional leadership; CKS-CH = charismatic leadership as measured by the Conger and Kanungo Scales. ⁎⁎ p ≤ .01; ⁎ p ≤ .05.

leadership (analysis #2). Thus, hypothesis 4b was only partially supported. Furthermore, charismatic leadership did not augment the impact of transformational leadership on profit (analysis #3). However, transformational leadership augments charismatic and transactional leadership (analysis #4) and adds 11% of unique variance. These results partially confirmed hypothesis 5b. Although these results echo the results from Table 5, where only transformational leadership was associated with profit, we performed these analyses as only the regression analyses provide additional information about the augmenting effect of certain leadership styles. The β-weights, however, should not be interpreted due to multicollinearity problems (high correlations between the leadership styles, see Table 5; cf. Cohen, Cohen, West, & Aiken, 2002). Summarizing the results for objective performance, hierarchical regression analyses show that transformational leadership augments the impact of transactional as well as charismatic leadership on profit. 10. Discussion Although they are often compared and used interchangeably, charismatic (CKS) and transformational (MLQ) leadership have a differential focus on the leadership phenomenon and its outcomes on top of the many components they share. This study provides evidence for convergent validity between transformational (MLQ) and charismatic (CKS) leadership. From the results reported in Table 3, we estimate the shared variance between these constructs to be 78%. While this supports the idea that charismatic and transformational leadership are to a larger degree overlapping, the still remaining 22% variance gives us an idea about the uniqueness of these constructs. Furthermore, analyses show that transformational and charismatic leadership are differential to transactional leadership, which provides, on the side of transformational leadership, further support for the divergent validity, showing that transformational leadership is not redundant to transactional leadership. According to our analyses, the same counts for charismatic leadership as measured with the CKS (cf. Note 2). However, the correlations between transactional and transformational leadership are high and significant. While these results limit divergent validity of the MLQ to some extent, they replicate findings from earlier research. Drawing on large samples from diverse organizational settings, independent analyses showed that transformational and transactional leadership are highly correlated (Antonakis et al., 2003; Bass & Avolio, 2000). With regard to criterion validity of transformational and charismatic leadership, the conducted hierarchical regressions provide more insight into the convergence (or overlapping) of these two concepts. First of all, the augmentation hypothesis was confirmed in this study. Transformational leadership augments transactional leadership for subjective as well as objective criteria. With respect to the objective criterion, our results confirm earlier results from Geyer & Steyrer (1998). Their results showed that transformational leadership uniquely accounts for 6% of the

130

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

variance of long-term objective performance ratings (profit). In comparison, in our study 14% of variance in profit is explained by transformational leadership, over and above transactional leadership. Furthermore, charismatic leadership (as measured with the CKS) explains additional variance in the subjective performance measures. However, charisma does not augment transactional leadership with regard to the objective performance measure. The key finding in this study is that although transformational (MLQ) and charismatic (CKS) leadership share a considerable portion of variance (78%), they have a differentiated impact on profit. When comparing transformational and charismatic leadership within the hierarchical regressions, it becomes obvious that these two concepts explain approximately the same amount of extra variance in subjective performance measures above the respective other concept. Therefore, with regard to subjective performance criteria, the augmentation effect can be confirmed for charismatic leadership as measured with the CKS scale as well. In order to learn more about which subscales of the CKS augment both transactional and transformational leadership, we conducted additional hierarchical regression analyses. It was found that all subscales of the CKS have a significant impact on subjective performance indicators such as Extra Effort. In sum, the additional regression analyses put further emphasis on the importance of analyzing a wide range of different leadership styles, as assessed by different leadership instruments. 10.1. Implications for theory The two terms of charismatic and transformational leadership are often used interchangeably. Although our empirical results suggest that they are highly convergent, both the MLQ and the CKS capture their own piece of charisma (cf. Table 1). While both approaches to leadership explain unique variance in subjective performance criteria, only the MLQ assesses parts of the leader–follower-relationship that are more directly related to profit. Several reasons might account for this phenomenon. First, in addition to behaviors, the MLQ includes followers' attributions which have been hypothesized to play an important role for the leadership process and related outcomes (Avolio, Gardner, Walumbwa, Luthans, & May, 2004). Second, whereas the MLQ focuses on the effect of focal leaders' perceived behavior on subordinates, Conger and Kanungo provide a more leader-centered theory. Thus, it might be speculated that in the CKS theory, follower processes (e.g. their attribution, motivation) that are related to profit are missing. Third, as the CKS implies three stages of the charismatic leadership process, future theoretical modifications should explicate effects of the stage model. Therein, it should be specified how the three stages are to be combined (e.g. sum or product of stages' subscales) to predict outcome criteria. Also, the effect of the CKS subscales of UB and PR should be specified for different contexts. From our results it might be speculated that within transaction-based organizations, these extraordinary and/or risky behaviors might be irrelevant to objective outcomes. 10.2. Implications for practice Our results have implications for managerial selection and training. Both charismatic and transformational leadership behavior are important to subjective performance. Hence, both approaches to leadership are valuable and should be the focus of leadership development interventions. It can be proposed that, depending on the organizational relevant performance indicator, certain distinct leadership styles will show a unique impact on subjective indicators. Given the context sensitivity of leadership phenomena (Antonakis et al., 2003; cf. Pawar & Eastman, 1997), the implementation of different leadership approaches within a single organization seems warranted. In sum, both theories are valuable and complement each other concerning the clarification of the influence of leadership on subjective indicators of organizational behavior. With regard to objective criteria such as profit, transformational leadership augments the impact of other (e.g. transactional and charismatic) leadership styles. In turn, evaluation and development of managers' transformational leadership abilities will help organizations to accomplish business goals (Avolio, 1999). 10.3. Limitations Several limitations should be noted. First of all, our sample is limited to a single organization. The obtained results may therefore be context-specific. This is even more of importance as the organization is German and Germany and charisma have a difficult relation. Research with transformational leadership in the German culture, however, has shown

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

131

that there are barely any differences in the correlational structure of transformational leadership and several outcomes to American studies (Felfe, Tartlerm, & Liepmann, 2004; Geyer & Steyrer, 1998; Kuchinke, 1999). We therefore expect that the here obtained results are to a certain degree transferable to other cultures as well. Furthermore, on the side of the subjective criteria, the data might underlie a common source and monomethod bias (Avolio, Yammarino, & Bass, 1991). Thus, the correlations reported here may be inflated. For example, Brown & Keeping (2005) demonstrated that the relationship between transformational leadership and subjective outcomes are “highly influenced by the interpersonal affect raters feel towards the targets being rated” (p. 245). However, with respect to transformational leadership behaviors, their influence on performance is supported by the results of the objective criteria as well. Thus, one of the most important findings of this study is not due to common source bias. Thirdly, for the analysis of the impact of leadership styles on objective performance criteria, the effect of early performance data should be controlled for. As an example, Tosi, Misangyi, Fanelli, Waldman, & Yammarino (2004) found no significant effect of CEO charismatic leadership style on organizational performance, when the effect of early performance data was accounted for. Fourthly, as already mentioned, several measurements and theories of transformational and charismatic leadership exist. Most of them agree on core facets of transformational/charismatic leadership. However, it can not be excluded that other measures would lead to different results. Fifthly, we use a limited set of variables for predicting performance. Additional variables, such as value congruence (Jung & Avolio, 2000) or self-efficacy (Kirkpatrick & Locke, 1996), may impact performance. Finally, we assessed constructs at a single point in time. Thus, we cannot be sure if leadership affects performance or the other way around. 10.4. Perspectives for future research Future research should try to further clarify the similarities and differences between – as well as the augmentation effect of – transformational and charismatic leadership in order to add further understanding to the variety of effective leadership. Therein, other measurements should also play a role in order to narrow down the effective leadership behaviors contained in one or the other concept. Analyses of convergent and divergent validity could draw on multitrait multirater matrices (cf. Conway, 1996) and should implement divergent measures such as personality. The present study was conducted in a German, transaction-based context (cf. Howell et al. 2005). Thus, empirical studies extending this research onto other cultures, countries, and organizational contexts such as nonprofit or research and development organizations (cf. Keller, 1992) are needed. Also, the impact of rivaling leadership styles on additional subjective performance measures such as organizational citizenship behavior (OCB; cf. Podsakoff, MacKenzie, & Bommer, 1996) and commitment (Mathieu & Zajac, 1990) should be explored. Longitudinal and experimental designs should also be part of research that focuses on the relationship between transformational and charismatic leadership styles with and the direction of influence of these leadership styles on performance and other criteria. Acknowledgement The assistance of L.H. Laukamp with data collection is gratefully acknowledged. The authors would like to thank D. Liepmann, the senior editor, and three anonymous reviewers for constructive criticism on earlier drafts of this paper. Notes (1) Research Edition Translation performed by Dr. Jens Rowold on September 25th, 2003. Translated and reproduced by special permission of the Publisher, MIND GARDEN, Inc., Redwood City, CA 94061 www.mindgarden.com from Multifactor Leadership Questionnaire. Copyright 1995, 2000 by Bernhard M. Bass and Bruce J. Avolio et al. All rights reserved. Further reproduction is prohibited without the Publisher's written consent. (2) Convergent and divergent validity of the MLQ and CKS was also analyzed on the subscale level. For the interested reader, the corresponding author will provide a full correlation matrix.

References Antonakis, J., Avolio, B. J., & Sivasubramaniam, N. (2003). Context and leadership: An examination of the nine-factor full-range leadership theory using the Multifactor Leadership Questionnaire. Leadership Quarterly, 14, 261−295.

132

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

Antonakis, J., & House, R. J. (2002). The full-range leadership theory: The way forward. In B. J. Avolio & F. J. Yammarino (Eds.), Transformational and charismatic leadership: The road ahead ( pp. 3−34). Amsterdam: JAI. Avolio, B. J. (1999). Full leadership development: Building the vital forces in organizations. Thousand Oaks: Sage Publications. Avolio, B. J., & Bass, B. M. (2004). Multifactor leadership questionnaire. Manual and sampler set, 3rd ed. Redwood City: Mind Garden, Inc. Avolio, B. J., Bass, B. M., & Jung, D. I. (1999). Re-examining the components of transformational and transactional leadership using the Multifactor Leadership Questionnaire. Journal of Occupational and Organizational Psychology, 72, 441−462. Avolio, B. J., Gardner, W. L., Walumbwa, F. O., Luthans, F., & May, D. R. (2004). Unlocking the mask: A look at the process by which authentic leaders impact follower attitudes and behaviors. Leadership Quarterly, 15, 801−823. Avolio, B. J., & Yammarino, F. J. (2002). Reflections, closing thoughts, and future directions. In B. J. Avolio & F. J. Yammarino (Eds.), Transformational and charismatic leadership: The road ahead ( pp. 385−406). Avolio, B. J., Yammarino, F. J., & Bass, B. M. (1991). Identifying common method variance with data collected from a single source: An unresolved sticky issue. Journal of Management, 17, 571−587. Avolio, B. J., Zhu, W., Koh, W. L., & Bhatia, P. (2004). Transformational leadership and organizational commitment: Mediating role of psychological empowerment and moderating role of structural distance. Journal of Organizational Behavior, 25, 951−968. Barling, J., Weber, T., & Kelloway, E. K. (1996). Effects of transformational leadership training on attitudinal and financial outcomes: A field experiment. Journal of Applied Psychology, 81, 827−832. Bass, B. M. (1985). Leadership and performance beyond expectations. New York: Free Press. Bass, B. M. (1997). Does the transactional-transformational leadership paradigm transcend organizational and national boundaries? American Psychologist, 52, 130−139. Bass, B. M. (1998). Transformational leadership: Industrial, military and educational impact. Mahway, NJ: Lawrence Erlbaum Associates. Bass, B. M., & Avolio, B. J. (2000). MLQ Multifactor leadership questionnaire. Redwood City: Mind Garden. Bass, B. M., & Steidlmeier, P. (1999). Ethics, character, and authentic transformational leadership behavior. Leadership Quarterly, 10, 181−217. Bass, B. M., Waldman, D. A., Avolio, B. J., & Bebb, M. (1987). Transformational leadership and the falling dominoes effect. Group & Organization Studies, 12, 73−87. Behling, O., & Law, K. S. (2000). Translating questionnaires and other research instruments: Problems and solutions. Thousand Oakes, CA: Sage. Brown, D. J., & Keeping, L. M. (2005). Elaborating the construct of transformational leadership: The role of affect. Leadership Quarterly, 16, 245−272. Cohen, P., Cohen, J., West, S. G., & Aiken, L. S. (2002). Applied multiple regression/correlation analysis for the behavioral sciences. New Jersey: Lawrence Erlbaum Associates. Conger, J. A., & Kanungo, R. N. (1998). Charismatic leadership in organizations. Thousand Oaks: Sage. Conger, J. A., Kanungo, R. N., Menon, S. T., & Mathur, P. (1997). Measuring charisma: Dimensionality and validity of the Conger-Kanungo scale of charismatic leadership. Canadian Journal of Administrative Sciences, 14, 290−302. Conway, J. M. (1996). Analysis and design of multitrait–multirater performance appraisal studies. Journal of Management, 22, 139−162. DeGroot, T., Kiker, D. S., & Cross, T. C. (2000). A meta-analysis to review organizational outcomes related to charismatic leadership. Canadian Journal of Administrative Sciences, 17, 356−371. Den Hartog, D. N., & Van Muijen, J. J. (1997). Transactional versus transformational leadership: An analysis of the MLQ. Journal of Occupational and Organizational Psychology, 70, 19−34. 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. In B. J. Avolio & F. J. Yammarino (Eds.), Transformational and charismatic leadership: The road ahead ( pp. 35−66). Amsterdam: JAI. Felfe, J., & Goihl, K. (2002). Deutsche überarbeitete und ergänzte Version des Multifactor Leadership Questionnaire (MLQ) [German revised and endorsed Version of the Multifactor Leadership Questionnaire (MLQ)]. In A. Glöckner-Rist (Ed.), ZUMA-Informationssystem. Elektronisches Handbuch sozialwissenschaftlicher Erhebungsinstrumente, 6th ed. Mannheim: Zentrum für Umfragen, Methoden und Analysen. Felfe, J., Tartler, K., & Liepmann, D. (2004). Advanced research in the field of transformational leadership. Zeitschrift für Personalforschung, 18, 262−288. Fuller, J. B., Patterson, C. E. P., Hester, K., & Stringer, S. Y. (1996). A quantitative review of research on charismatic leadership. Psychological Reports, 78, 271−287. Geyer, A. L. J., & Steyrer, J. M. (1998). Transformational leadership and objective performance in banks. Applied Psychology—An International Review, 47, 397−420. House, R. J., & Shamir, B. (1993). Toward the integration of transformational, charismatic, and visionary theories. In M. M. Chemers & R. Ayman (Eds.), Leadership theory and research: Perspectives and directions ( pp. 167−188). San Diego: Academic Press. Howell, J. M., & Avolio, B. J. (1992). The ethics of charismatic leadership: Submission or liberation? Academy of Management Executive, 6, 43−54. Howell, J. M., Neufeld, D. J., & Avolio, B. J. (2005). Examining the relationship of leadership and physical distance with business unit performance. Leadership Quarterly, 16, 273−285. Hunt, J. G., Boal, K. B., & Dodge, G. E. (1999). The effects of visionary and crisis-responsive charisma on followers: An experimental examination of two kinds of charismatic leadership. Leadership Quarterly, 10, 423−448. Hunt, J. G., & Conger, J. A. (1999). From where we sit: An assessment of transformational and charismatic leadership research. Leadership Quarterly, 10, 335−343. 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. Judge, T. A. (2005). Psychological leadership research: The state of the art. Paper presented at the 4th Conference of the German Psychological Association, section I/O Psychology, Bonn, Germany.

J. Rowold, K. Heinitz / The Leadership Quarterly 18 (2007) 121–133

133

Judge, T. A., & Piccolo, R. F. (2004). Transformational and transactional leadership: A meta-analytic test of their relative validity. Journal of Applied Psychology, 89, 755−768. Jung, D. I., & Avolio, B. J. (2000). Opening the black box: An experimental investigation of the mediating effects of trust and value congruence on transformational and transactional leadership. Journal of Organizational Behavior, 21, 949−964. Keller, R. T. (1992). Transformational leadership and the performance of research and development project groups. Journal of Management, 18, 489−501. Kirkpatrick, S. A., & Locke, E. A. (1996). Direct and indirect effects of three core charismatic leadership components on performance and attitudes. Journal of Applied Psychology, 81, 36−51. Kuchinke, K. P. (1999). Leadership and culture: Work-related values and leadership styles among one company's U.S. and German telecommunication employees. Human Resource Development Quarterly, 10, 135−154. Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. (1996). Effectiveness correlates of transformational and transactional leadership: A metaanalytic review of the MLQ literature. Leadership Quarterly, 7, 385−425. Mathieu, J. E., & Zajac, D. M. (1990). A review and meta-analysis of the antecedents, correlates and consequences of organizational commitment. Psychological Bulletin, 108, 171−194. McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intraclass correlation coefficients. Psychological Methods, 1, 30−46. Medsker, G. J., Williams, L. J., & Holahan, P. J. (1994). A review of current practices for evaluating causal models in organizational behavior and human resources management research. Journal of Management, 20, 439−464. Meng, X. L., Rosenthal, R., & Rubin, D. B. (1992). Comparing correlated correlation coefficients. Psychological Bulletin, 111, 172−175. Pawar, B. S., & Eastman, K. K. (1997). The nature and implications of contextual influences on transformational leadership: A conceptual examination. Academy of Management Review, 22, 80−109. Podsakoff, P. M., MacKenzie, S. B., & Bommer, W. H. (1996). Transformational leader behaviors and substitutes for leadership as determinants of employee satisfaction, commitment, trust, and organizational citizenship behaviors. Journal of Management, 22, 259−298. Rafferty, A. E., & Griffin, M. A. (2004). Dimensions of transformational leadership: Conceptual and empirical extensions. Leadership Quarterly, 15, 329−354. Rowold, J. (2004). MLQ-5X. German translation of bass and Avolio's multifactor leadership. Questionnaire Redwood City: Mind Garden. Rowold, J. (2005). Multifactor leadership questionnaire. Psychometric properties of the German translation by Jens Rowold. Redwood City: Mind Garden. Sashkin, M. (1988). The visionary leader. In J. A. Conger & R. N. Kanungo (Eds.), Charismatic leadership: The elusive factor in organizational effectiveness ( pp. 98−124). San Francisco: Jossey-Bass. Sashkin, M. (2004). Transformational leadership approaches. In J. Antonakis, A. T. Cianciolo, & R. J. Sternberg (Eds.), The nature of leadership ( pp. 171−196). Thousand Oaks: Sage Publications. 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. Tosi, H. L., Misangyi, V. F., Fanelli, A., Waldman, D. A., & Yammarino, F. J. (2004). CEO charisma, compensation, and firm performance. Leadership Quarterly, 15, 405−420. Waldman, D. A., Bass, B. M., & Yammarino, F. J. (1990). Adding to contingent-reward behavior — The augmenting effect of charismatic leadership. Group & Organization Studies, 15, 381−394. Yammarino, F. J., Spangler, W. D., & Bass, B. M. (1993). Transformational leadership and performance: A longitudinal investigation. Leadership Quarterly, 4(1), 81−102. Yammarino, F. J., Spangler, W. D., & Dubinsky, A. J. (1998). Transformational and contingent reward leadership: Individual, dyad, and group levels of analysis. Leadership Quarterly, 9, 27−54. Yukl, G. (1999). An evaluation of conceptual weaknesses in transformational and charismatic leadership theories. Leadership Quarterly, 10, 285−305. Yukl, G. (2002). Leadership in organizations, 5th ed. Upper Saddle River, NJ: Prentice Hall.