Annotated Bibliography and Outline – Leadership Style

PERSONALITY PROCESSES AND INDIVIDUAL DIFFERENCES

The Double-Edged Sword of Leader Charisma: Understanding the Curvilinear Relationship Between Charismatic Personality

and Leader Effectiveness

Jasmine Vergauwe Ghent University

Bart Wille University of Antwerp

Joeri Hofmans Vrije Universiteit Brussel

Robert B. Kaiser Kaiser Leadership Solutions, Greensboro, North Carolina

Filip De Fruyt Ghent University

This study advanced knowledge on charisma by (a) introducing a new personality-based model to conceptualize and assess charisma and by (b) investigating curvilinear relationships between charismatic personality and leader effectiveness. Moreover, we delved deeper into this curvilinear association by (c) examining moderation by the leader’s level of adjustment and by (d) testing a process model through which the effects of charismatic personality on effectiveness are explained with a consideration of specific leader behaviors. Study 1 validated HDS charisma (Hogan Development Survey) as a useful trait-based measure of charisma. In Study 2 a sample of leaders (N � 306) were assessed in the context of a 360-degree development center. In line with the too-much-of-a-good-thing effect, an inverted U-shaped relationship between charismatic personality and observer-rated leader effectiveness was found, indicating that moderate levels are better than low or high levels of charisma. Study 3 (N � 287) replicated this curvilinear relationship and further illustrated the moderating role of leader adjustment, in such a way that the inflection point after which the effects of charisma turn negative occurs at higher levels of charisma when adjustment is high. Nonlinear mediation modeling further confirmed that strategic and operational leader behaviors fully mediate the curvilinear relationship. Leaders low on charisma are less effective because they lack strategic behavior; highly charismatic leaders are less effective because they lack operational behavior. In sum, this work provides insight into the dispositional nature of charisma and uncovers the processes through which and conditions under which leader charisma translates into (in)effectiveness.

Keywords: charismatic personality, curvilinear relationships, leader effectiveness, process model, too-much-of-a-good-thing

The topic of charisma is characterized both by ambiguity and intense debate. Vivid questions about this intriguing and everyday- life construct involve its underlying nature (e.g., Bono & Judge, 2004; Resick, Whitman, Weingarden, & Hiller, 2009) as well as its consequences, particularly in, but not limited to, organizational

contexts. Is charisma something that can be measured indepen- dently from those perceiving a person as charismatic? If it is, can we identify a cluster of personality characteristics that meaning- fully predicts others’ ratings of charisma? And finally, is it always beneficial for leaders in organizational contexts to demonstrate

This article was published Online First May 29, 2017. Jasmine Vergauwe, Department of Developmental, Personality, and Social

Psychology, Ghent University; Bart Wille, Department of Training and Education Sciences, University of Antwerp; Joeri Hofmans, Research Group of Work and Organizational Psychology, Vrije Universiteit Brussel; Robert B. Kaiser, Kaiser Leadership Solutions, Greensboro, North Carolina; Filip De Fruyt, Department of Developmental, Personality, and Social Psychology, Ghent University.

We thank Hogan Assessments, and Jeff Foster in particular, for collab- orating on this research project. Moreover, we thank James LeBreton for

his methodological advice regarding the aggregation of observer ratings. This research was conducted with support from the Ghent University research fund (01J01513). The fourth author, Robert B. Kaiser, developed the Leadership Versatility Index (LVI), which is used in his consultancy business.

Correspondence concerning this article should be addressed to Jasmine Vergauwe, Department of Developmental, Personality, and Social Psychol- ogy, Ghent University, H. Dunantlaan 2, B-9000 Gent, Belgium. E-mail: [email protected]

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

Journal of Personality and Social Psychology, 2018, Vol. 114, No. 1, 110 –130 © 2017 American Psychological Association 0022-3514/18/$12.00 http://dx.doi.org/10.1037/pspp0000147

110

high levels of charisma? The overall objective of our work was to investigate these open questions.

Although most of us can easily imagine a charismatic person, and are able to tell whether someone is charismatic or not, to date, charisma is still a fuzzy construct in the scientific literature. At the core of the debate lies the question: Does charisma represent a personal characteristic of the leader (e.g., Judge, Piccolo, & Ko- salka, 2009; Riggio, 2009) or is it an attribution based on relational processes (e.g., Conger, Kanungo, & Menon, 2000; Howell & Shamir, 2005)? Traditional models of charismatic leadership, such as Conger and Kanungo’s (1987) model, conceptualize charisma as an attribution based on follower perceptions of their leader’s behavior. Stated differently, according to these models, charisma only exists “in the eye of the beholder.” More recently, however, increased attention is being devoted to trait-perspectives on lead- ership (e.g., Judge et al., 2009; Zaccaro, 2012), referring to cha- risma as a constellation of personal characteristics that allows an individual to influence other people by affecting their feelings, opinions, and behaviors (Riggio, 2009). As a compromise, the literature now acknowledges that charismatic leaders have certain characteristics that distinguish them from noncharismatic leaders (DuBrin, 2012). In other words, individual differences in person- ality play an important role in the level of charisma that is attrib- uted to a specific leader. Previous efforts to uncover this disposi- tional nature of charisma have mainly focused on Big Five personality traits (Bono & Judge, 2004), showing only modest associations. The starting point of our work was the aim to provide an in-depth investigation of the dispositional nature of charisma, by establishing a trait-based model of charisma that can be as- sessed independently from the observer’s perspective.

Turning to the outcomes of charisma, we can say that organi- zational research has generally shown that charisma is positively related to individual-, group-, and firm-level outcomes. Charis- matic leaders have the ability to inspire followers toward higher levels of performance and to instill deep levels of commitment, trust, and satisfaction (e.g., Conger et al., 2000; Dvir, Eden, Avolio, & Shamir, 2002; Shamir, House, & Arthur, 1993). As a result, they are generally perceived as more effective by their subordinates compared with less charismatic leaders (Amirul & Daud, 2012; Lowe, Kroeck, & Sivasubramaniam, 1996). However, in the light of recent theoretical advances in organizational- behavior and management literatures, it can be questioned whether this positive association between charisma and leader effectiveness is appropriately represented by a continuous and linear relation- ship. Specifically, the now widely established too-much-of-a- good-thing (TMGT) effect (Pierce & Aguinis, 2013) has chal- lenged the assumption that more of a desirable trait is always better. The alternative to this linear model is a perspective in which ordinarily beneficial antecedents are no longer advantageous when taken too far. Studies have indeed indicated that, after a certain point, too much leader assertiveness (Ames & Flynn, 2007), too much leader-member exchange (Harris & Kacmar, 2006), and too much contingent-reward leadership (Harris & Russell, 2013) can be detrimental for leadership outcomes. In the context of charisma, the critical question arises whether a leader can be too charismatic, meaning that from a certain point more charisma may no longer be advantageous or may even become a hindrance with respect to his or her effectiveness. Therefore, our work extended the available

literature in this domain by investigating curvilinear relationships between charismatic personality and leader effectiveness.

A central tenet in the TMGT principle is that the inflection point— or the point after which further increases in the “desirable trait” become counterproductive— is context-specific or depends on specific boundary conditions (Pierce & Aguinis, 2013). With regard to this context, charismatic leaders are more likely to emerge in situations of crises (Pillai & Meindl, 1998) and in environments characterized by a high degree of challenge and opportunities for change (Shamir & Howell, 1999). In this respect, a leader’s typical way of coping with stressful situations has been put forward as a boundary condition that influences the likelihood that charisma also translates into beneficial outcomes (Hogan & Hogan, 2007). Our work therefore investigated the role of leaders’ levels of adjustment as a condition under which the curvilinear relationship between charisma and effectiveness may vary.

Finally, an overview of the literature indicates that the mecha- nisms that explain any relationship between leader charisma and effectiveness are still unclear. Moreover, because the nature of the metatheoretical TMGT principle is more descriptive than explor- atory, the presence of a curvilinear relationship would not explain why charisma can backfire. That is, whereas the TMGT principle offers a prediction about the functional form of the association between charisma and effectiveness, it provides no account for the specific mechanisms that intervene in this relationship. Thus, explanatory frameworks underlying the TMGT effect become in- creasingly important (Busse, Mahlendorf, & Bode, 2016). As a final objective, our work also investigated specific leader behav- iors as mechanisms through which leader charisma can result in leader (in)effectiveness.

In sum, our aim was to enhance the understanding of charisma and its role in leader contexts in four different ways. In Study 1 we made a case for HDS charisma (Hogan Development Survey; Hogan & Hogan, 2009), as a new personality-based model to conceptualize and assess charisma (objective 1). In Study 2, this measure of charisma was related to leader effectiveness, with particular attention to curvilinear relationships, as this may signal a too-much-of-a-good-thing effect (objective 2). Besides replicat- ing this curvilinear relationship, Study 3 delved deeper into this association by examining adjustment as a potential moderator (objective 3) and by testing a process model in which the effects of charismatic personality on effectiveness can be explained through specific leader behaviors (objective 4). An integrative research model, including these four objectives, is presented in Figure 1.

Charismatic personality Overall leader effectiveness

Interpersonal behavior: Forceful / Enabling

Adjustment

Charismatic behavior/ Charismatic leadership

Organizational behavior: Strategic / Operational

2

3

4 4

4 4 1

Figure 1. Research model: The (curvilinear) relationship between char- ismatic personality and (observer-rated) overall leader effectiveness, as moderated by leader adjustment and mediated through leader behaviors.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

111CHARISMATIC PERSONALITY

A Trait-Based Perspective on Charisma

Conceptualizing charisma in terms of personality raises the question of which traits to consider. Investigating the relationship between charisma and the five-factor model of personality, Bono and Judge (2004) found that the highly charismatic leader tends to score high on extraversion and low on neuroticism. Nevertheless, their results also showed that the Big Five explained only 12% of the variability in charisma, which made the authors conclude that charisma might have dispositional antecedents that cannot be captured by the Big Five.

The current study proposes the HDS charismatic cluster, named after the personality instrument used to assess the personality of leaders (i.e., the Hogan Development Survey; Hogan & Hogan, 2009), as a useful trait-based measure of charisma. The HDS is an empirically validated personality instrument grounded in socioana- lytic theory (Hogan, 2007). A central premise of the theory is that personality is conceptualized as an individual’s reputation—that is, in terms of attributions observers make about that person’s char- acteristic behavior. In addition, socioanalytic theory identifies a dark side to reputation, referring to attributes that may be benefi- cial in some contexts but counterproductive in other contexts. The four personality tendencies constituting this charismatic cluster— that is, Bold, Mischievous, Colorful, and Imaginative— have been selected based on their conceptual overlap with the construct of charisma and have previously been referred to as the “charismatic cluster” (Kaiser & Hogan, 2007; Kaiser, LeBreton, & Hogan, 2015; VanBroekhoven, 2011). Generally, however, it has been labeled the “moving against” people factor (Horney, 1950), refer- ring to the tendency to overwhelm, coopt, intimidate, manipulate, and persuade as a technique for managing insecurities and self- doubts (Hogan & Hogan, 2009). This factor resembles what Tel- legen (1985) calls “positive affectivity” and has been related to management potential (Furnham, Trickey, & Hyde, 2012), leader- ship performance (Benson & Campbell, 2007), and innovative potential (Zibarras, Port, & Woods, 2008). Moreover, there is convincing empirical evidence linking each of these traits sepa- rately to charismatic leadership.

A first crucial feature of the charismatic personality concerns self-confidence (Bass, 1998; Bono & Judge, 2004; House & Howell, 1992). Self-confidence allows leaders to convey that they are credible in their conviction that high-performance expectations can be achieved (Dóci & Hofmans, 2015; Judge & Bono, 2000). This feature is captured in the HDS Bold scale. Second, charismatic persons are captivating, and this relates to a tendency to be expressive, energetic, and optimistic about the future (Bono & Judge, 2004). Charismatic leaders are extraverted and inspirational, with excellent rhetoric abil- ities (e.g., Emrich, Brower, Feldman, & Garland, 2001), which allow them to evoke enthusiasm, confidence, and commitment in their followers (Bass, 1998). This second dimension is captured in the HDS Colorful scale. Third, charismatic persons stand out because of their tendency to explore the unknown, persuading themselves and others to keep on pushing the limits. Charismatic leaders usually enjoy challenging the status quo and taking risks (Conger, Kanungo, Menon, & Mathur, 1997; House & Howell, 1992; Shamir et al., 1993), which is captured in the Mischievous scale of the HDS. Fourth and finally, charismatic leaders are visionary (Judge & Bono, 2000; House & Howell, 1992) and are seen as thinking in creative ways

(Mueller, Goncalo, & Kamdar, 2011). This is captured in the HDS Imaginative scale.

In the light of the “trait versus attribution” debate described above, evidence for the construct validity of the HDS charismatic cluster can be obtained by linking people’s self-reports on this personality cluster to observers’ perceptions of charisma levels. More specifically, if HDS charisma truly captures charismatic personality, then we should find positive associations between HDS charisma self-reports and observers’ perceptions of charisma-related tendencies such as self- confidence, expressiveness, energy, optimism about the future, rhe- torical ability, being inspirational, risk taking, challenging the status quo, and creativity. Similarly, when HDS charisma is applied to a leadership context, we should observe positive associations between leaders’ self-reports on HDS charisma and followers’ attributions of charismatic leadership.

The Curvilinear Effect of Charismatic Personality

Turning to the outcomes of charisma, we note that a key question driving the current research is whether the association between people’s charismatic personality and their levels of effec- tiveness, particularly in a leadership context, is best represented by a curvilinear (cf. too-much-of-a-good-thing) instead of a linear relationship (cf. more is better).

Closer inspection of the four personality traits constituting the charismatic cluster already signals curvilinearity, given that each of these traits can be linked to dysfunctional tendencies when they are taken too far (Hogan & Hogan, 2009). Specifically, self- confidence (i.e., Bold) can translate into overconfidence, hubris, and narcissism in highly charismatic leaders (Deluga, 1997; House & Howell, 1992; Popper, 2002; Sankowsky, 1995), posing valid threats to their overall effectiveness. In line with these thoughts, a curvilinear relationship was found between the Bold scale and leader effectiveness (Grijalva, Harms, Newman, Gaddis, & Fraley, 2015). Similarly, the enthusiastic and entertaining nature of char- ismatics (i.e., Colorful) may turn into attention-seeking behaviors that distract the organization from its mission. In this context, Gardner and Avolio (1998) described highly charismatic leaders as “the epitome of drama” (p. 33). Further, risk tolerance and per- suasiveness of charismatics (i.e., Mischievous), the third corner- stone of the charismatic cluster, may turn into manipulative and exploitative behavior. This is in line with research showing that high charisma in leaders is also associated with Machiavellianism (Deluga, 2001). Finally, at the extreme of creativity (i.e., Imagi- native), highly charismatic leaders have also been described to think and act in fanciful, eccentric ways (Kaiser & Hogan, 2007), which may represent a final threat to their level of effectiveness in organizational settings. In sum, it can be expected that a certain degree of charismatic tendencies is indeed desirable and associated with higher effectiveness, whereas too much causes harm. Very low levels of charisma should manifest as a lack of the confidence, strategic vision, and dynamism often associated with effective leadership (Den Hartog et al., 1999; Lord, Foti, & DeVader, 1984). At very high levels, on the other hand, the self-absorbed tendencies associated with charisma—arrogance, manipulation, grandiose vi- sions, and dramatic attention seeking—may negatively affect ob- servers’ evaluation of leader effectiveness (Benson & Campbell, 2007; Epitropaki & Martin, 2004). These effects are expected to give shape to a curvilinear relationship between charisma and

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

112 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

leader effectiveness. To further understand these curvilinear ef- fects, we needed to take a closer look at the specific behaviors displayed by charismatic leaders.

Charismatic Personality and Leader Behaviors

An important objective of our work was to enhance our under- standing of the (curvilinear) association between charisma and leader effectiveness by investigating specific leader behaviors. To this end, we considered four leader-behavior dimensions, which serve as mediating mechanisms in our research model. Specifi- cally, we drew on the versatile leadership model (Kaiser, Over- field, & Kaplan, 2010) in which leader behaviors are covered by two pairs of opposing leadership dimensions: Forceful versus enabling leadership, representing the interpersonal side, or how one leads; and strategic versus operational leadership, representing the organizational side, or what one leads. Forceful leadership includes assuming authority and using power to push for perfor- mance, while enabling leadership concerns creating conditions for others to contribute, through empowerment, participation, and support. Strategic leadership can be defined as positioning the organization for the future by setting direction, expanding capa- bility, and supporting innovation, whereas operational leadership includes guiding the team to get things done in the near term by managing the tactical details of execution, focusing resources, and managing with process discipline (Kaiser et al., 2010, 2015). Although each of the two classes of leader behaviors are concep- tualized as opposing dimensions (i.e., highly forceful leaders are usually low on enabling), a small percentage of “versatile” leaders can use opposing leader behaviors with equal ease. This leadership model overlaps with other taxonomies of leader behavior (e.g., DeRue, Nahrgang, Wellman, & Humphrey, 2011; Yukl, 2006). For instance, in terms of Yukl’s (2006) taxonomy, forceful and enabling cover the relation-oriented category of leader behavior, strategic taps into the change-oriented category, and operational covers the task-oriented category of leader behavior. Importantly, each of these dimensions has clear conceptual associations with charismatic personality.

Interpersonal Leader Behavior

A forceful leader takes charge by assuming authority and giving direction, is decisive, speaks up, and doesn’t back down easily. Moreover, forceful leaders express high performance expectations and push people hard to get there (Kaiser et al., 2010)—features that are also characteristic of charismatic leaders (e.g., Conger & Kanungo, 1987; Shamir et al., 1993; Waldman, Ramirez, House, & Puranam, 2001). In addition, the high need for power, along with manifestations of authoritarian behavior that have been observed in charismatic leaders (House & Howell, 1992), suggests that charismatic personalities will be more likely to be forceful in their interpersonal style. Enabling behaviors, on the other hand, include listening to others, seeking their input, and supporting others by showing appreciation and being sensitive to people’s feelings (Kaiser et al., 2010). In this regard, the leadership literature has demonstrated an extensive overlap between charismatic leadership and narcissistic tendencies (e.g., Deluga, 1997; Galvin, Waldman, & Balthazard, 2010; Howell, 1988; Sankowsky, 1995), which are assumed to make charismatic leaders poor listeners and highly

sensitive to criticism (Maccoby, 2004). Narcissism is not only associated with an inflated sense of self-importance and a preoc- cupation with fantasies of unlimited success but also with inter- personal exploitation, a lack of empathy, and indifference toward others (House & Howell, 1992). As such, we expected highly charismatic leaders to be perceived as more forceful and less enabling by coworkers, compared with less charismatic leaders.

Organizational Leader Behavior

In addition to the effects at the interpersonal level, charisma has also been described to influence behavior at the organizational or business-related level. Most obviously, one of the hallmarks of charismatic leadership involves displaying exceptional strategic vision and articulation (Bass, 1985; Conger & Kanungo, 1998; Maccoby, 2004). Charismatic leaders are believed to engage in behaviors such as referring to collective history, emphasizing collective identity, communicating a collective vision or mission, and pursuing collective goals and interests. On the other hand, the realization of this vision requires leadership that fosters goal setting, planning, and task execution (Kaiser et al., 2010). It is here, at the operational level, that highly charismatic leaders may underachieve compared with those with lower charismatic tenden- cies. For instance, Conger (1990) noticed that charismatic leaders can become so excited by their ideas that they can lose touch with reality and get stuck in the process of implementing these visions. Operational behavior involves the short-term handling and moni- toring of daily tasks, and this may appear less appealing to highly charismatic leaders, who are mainly interested in the bigger picture and long-term objectives. Taken together, we expect highly char- ismatic leaders to be more strategic and less operational compared with less charismatic leaders.

How can these expected behavioral manifestations of leader charisma explain lower effectiveness ratings for the highest cha- risma levels? Drawing on the Antecedent-Benefit-Cost (ABC) framework (Busse et al., 2016), we theorized that the explanatory mechanism underlying the curvilinear relationship was to be found in inadequately proportioned patterns of leader behaviors associ- ated with various charisma levels. The ABC framework adopts a competitive-mediation perspective (e.g., Hayes & Preacher, 2010) to explain the TMGT effect, by stating that an overall effect results from the aggregation of multiple opposed effects. Specifically, a special case is investigated in which a dependent variable is affected by two (or more) mediators with opposite directionalities of influence, which are caused by a common antecedent variable. The competing mediators can be understood as benefits and costs of the antecedent. From an ABC perspective, a decrease in effec- tiveness (i.e., TMGT effect) occurs at higher levels of a desired antecedent variable, when the costs associated with the desired variable outweigh its benefits. When applied to the research model presented in Figure 1, the costs associated with operational leader behavior may outweigh the benefits delivered by strategic behav- ior when a certain level of charisma is exceeded. Highly charis- matic leaders may be strategically ambitious, but at the expense of getting day-to-day work activities executed in a proper manner, with detrimental effects on perceived effectiveness. Similarly, the costs associated with enabling behavior may outweigh the benefits that can be ripped from forceful behavior. Even when there are benefits of giving direction and expressing high performance ex-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

113CHARISMATIC PERSONALITY

pectations, beyond certain charisma levels leaders might be less capable to meet their followers’ needs because of a lack of en- abling behavior. Ultimately, this would also result in decreased ratings of effectiveness.

The Moderating Role of Adjustment

The central idea in our work was that charismatic tendencies become maladaptive, particularly in relation to leader effective- ness, when taken too far. Importantly, however, a core tenet in the TMGT principle is that the inflection point— or the point after which further increases in the “desirable” trait are no longer beneficial—is context-specific (Pierce & Aguinis, 2013). Consis- tent with these thoughts, previous work in this area, studying for instance the curvilinear association between conscientiousness and job performance (Le et al., 2011), has indeed indicated that situ- ational features play a crucial role in determining this inflection point. Findings particularly indicate that there needs to be a match between a person’s trait levels and the requirements that are imposed in a certain environment or situation. In this regard, a leadership context can be thought of as an environment that typically combines high pressure with high discretion. High pres- sure indicates that leaders often face difficult decisions with po- tentially far-reaching implications for themselves, their subordi- nates, and their entire organizations. Charismatic leaders in particular often encounter such stressful conditions, as they are more likely to emerge in situations of crises and in environments characterized by a high degree of challenge and opportunities for change (Pillai & Meindl, 1998; Shamir & Howell, 1999). High discretion means that they can and are even required to take responsibility for their actions. Kaiser and Hogan (2007) have described both conditions as situations in which derailment is more likely to occur.

A crucial element that can help leaders to cope with these high levels of pressure is the leader’s ability to remain self-composed and adjusted (Hogan & Hogan, 2007). For instance, Kaiser et al. (2015) argued that low adjustment or increased reactivity to dif- ficult circumstances diminishes the resources needed to self- regulate, and the resulting experience of threat triggers self- protective strategies. It is these self-protective strategies that define the dark side of charisma and which have the potential to render charisma dysfunctional in terms of leaders’ effectiveness. In other words, it can be predicted that a leader’s level of adjustment, or his or her general ability to cope with stressful events, plays an important role in determining at which level charisma loses its beneficial effects.

Predictions and Plan of Study

Our account of the current investigation of charisma and its outcomes in a leadership context led to four sets of predictions, which are also summarized in our research model (see Figure 1). The first set of predictions relates to the construct validity of HDS charisma as a trait-based measure of charisma. We expected HDS charisma to relate positively to self and observers’ perceptions of charisma-related personality tendencies (Hypothesis 1a) and to followers’ attributions of charismatic leadership (Hypothesis 1b).

The second prediction focused on the link between charismatic personality and leader effectiveness. In line with the TMGT prin-

ciple, a curvilinear effect was expected for charismatic personality and observer-ratings of leader effectiveness: Leaders with both low and high charismatic personalities would be perceived as being less effective than leaders with moderate levels of charisma (Hypothesis 2).

The third prediction addressed the potential moderating effect of the leader’s level of adjustment. Specifically, adjustment was expected to moderate the curvilinear effect of charisma on leader effectiveness, in such a way that the inflection point after which the relation turns asymptotic and negative occurs at higher levels of charisma when adjustment is high (Hypothesis 3). As such, the decrease in effectiveness (cf. the right part of the inverted U-shape) would present itself at higher levels of charisma when adjustment is high.

Our final set of predictions addressed the underlying mecha- nisms of the curvilinear relationship between leader charisma and effectiveness. With regard to interpersonal leadership, we expected charisma to be positively associated with forceful behavior (Hy- pothesis 4a) and negatively associated with enabling behavior (Hypothesis 4b). With regard to organizational leadership, we expected charisma to be positively associated with strategic be- havior (Hypothesis 5a) and negatively associated with operational behavior (Hypothesis 5b). Moreover, we expected these leader behaviors to mediate the curvilinear relationship between charis- matic personality and leader effectiveness (Hypothesis 6). Beyond a certain optimal level (i.e., the inflection point), further increases in charismatic personality might reduce the effectiveness of lead- ers in two important ways (Busse et al., 2016): Enabling costs may outweigh forceful benefits, and operational costs may outweigh strategic benefits.

These hypotheses were tested in three studies. In Study 1, evidence of construct validity was provided for the HDS charis- matic cluster as a trait-based measure of charisma. In Study 2, ratings of leader effectiveness (i.e., from self, subordinates, peers, and superiors) were collected to test for the relationship between charismatic personality and leader effectiveness. Finally, in Study 3, a second sample of 360-degree-rated leaders was used to rep- licate the curvilinear relationship between charismatic personality and observer-rated leader effectiveness, as well as to examine interaction-effects with the leader’s level of adjustment (i.e., mod- eration) and to explore the underlying mechanisms (i.e., media- tion).

Study 1

In Study 1, we used two different samples to provide construct validity evidence for HDS charisma as a useful trait-based measure of charisma: Goldberg’s (2008) Eugene-Springfield community sample (Sample 1) and a Belgian sample of leaders (Sample 2). The first goal of this validation study was to empirically test whether HDS charisma relates to charisma-related tendencies that have been described in the literature, such as self-confidence, expressiveness, energy, optimism about the future, rhetorical abil- ity, being inspirational, risk taking, challenging the status quo, and creativity (cf. Hypothesis 1a). To this end, the Eugene-Springfield Community sample was used; with it HDS charisma could be related to a list of self- and observer-rated personality descriptions (i.e., Big Five Inventory; John & Srivastava, 1999). By relating HDS charisma to a set of fine-grained behavioral descriptions

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

114 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

reflecting personality tendencies, we gained an in-depth under- standing of its content.

The second question we addressed in this validation study was whether charismatic personality, as operationalized by leaders’ scores on HDS charisma, related to followers’ attributions of charismatic leadership. This question was answered using data from actual leaders, who were rated by subordinates in terms of charismatic leadership (Sample 2). As charisma pertains to a constellation of personal characteristics that allow an individual to influence other people by affecting their feelings, opinions, and behaviors (Riggio, 2009), charismatic personality should be re- flected in followers’ attributions of charisma, a point also made by socioanalytic theory (Hogan, 2007). Hence, if HDS charisma re- ally captures charismatic personality, it should be positively re- lated to charismatic leadership attributions (Hypothesis 1b).

Method

All research was conducted according to the ethical rules pre- sented in the General Ethical Protocol of the Faculty of Psychol- ogy and Educational Sciences of Ghent University.

Procedure and participants. Sample 1. Data were used from Goldberg’s (2008) Eugene-

Springfield community sample. Previous research has, for in- stance, used this sample to investigate personality structure (DeY- oung, Quilty, & Peterson, 2007), well-being (Naragon-Gainey & Watson, 2014), and vocational interests (Pozzebon, Visser, Ashton, Lee, & Goldberg, 2010). For the current study, a sub- sample (N � 156) was used, from which we have self- and observer ratings of five-factor model personality in 1998 (Time 1: Big Five Inventory) and self-ratings of charismatic personality in 2007 (Time 2: Hogan Development Survey). At Time 1, an aver- age of three peers provided observer ratings of personality. Targets were on average 47.67 years old (SD � 11.27), and 44% were male (see Goldberg, 2008, for additional details about this sample).

Sample 2. In the context of a course assignment, third-year psychology undergraduate students were asked to recruit one tar- get leader. Students were only responsible for recruiting the target and for delivering the informed consent. Three inclusion criteria were imposed: Targets had to be (a) at least 25 years old, (b) responsible for at least three subordinates, and (c) have at least 3 years of working experience. All target leaders received an email including a personal login and a link to an online survey. In total, 204 Belgian leaders participated in the study by providing self- ratings on their personality (HDS and NEO-Five Factor Inven- tory). Fifty-seven percent of the leaders were male and the mean age of the sample was 45.96 years (SD � 8.62). The majority of the leaders had completed a higher education program (89.4% had a bachelor’s degree or higher) and the average job tenure was 24.01 (SD � 8.50) years. Each of the targets was asked to nomi- nate one subordinate deemed willing and able to evaluate their direct superior on charismatic leadership (Conger-Kanungo Scale). Thirty-eight percent of the subordinates were male and their mean age was 39.87 years (SD � 10.24). Subordinates reported frequent personal contacts with their respective leaders (60.1% reported to have daily contact or more) and indicated that they were familiar with their target’s behavior at work (M � 4.08, SD � .78; on a 5-point Likert scale). On average, subordinates and leaders indi-

cated that they had been working together for an average duration of 71.49 months (SD � 68.60).

Measures. Charismatic personality. In both samples the participants

completed the 56 items comprising the Bold, Mischievous, Col- orful, and Imaginative scales of the HDS (Hogan & Hogan, 2009). Participants responded by indicating whether they agreed or dis- agreed with the items. Consequently, the raw scale scores ranged from 0 to 56, with higher scores indicating higher charisma levels. Cronbach alphas of the combined HDS charisma scale were .84 (Sample 1) and .85 (Sample 2). Correlations between the four scales ranged between r � .23 (Bold-Imaginative) and r � .45 (Bold-Colorful) in Sample 1 and between r � .19 (Bold- Imaginative) and r � .53 (Mischievous-Imaginative) in Sample 2.

Big Five traits. In Sample 1 both self-reports and observer reports were provided on the 44-item Big Five Inventory (BFI; John & Srivastava, 1999) and two additional items measuring physical attractiveness (see Goldberg, 2008). For each of these 46 descriptions, we obtained an observer score by averaging the separate peer ratings. The average rwg(j) interrater agreement co- efficient (James, Demaree, & Wolf, 1984) among the peers was relatively high (rwg(j) � .65), justifying this aggregation approach. In Sample 2, leaders completed the 60-item NEO Five-Factor Inventory (Hoekstra, Ormel, & De Fruyt, 2007) to measure their standing on the Big Five traits (i.e., Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness). The internal consistencies of the five scales were acceptable to good, ranging between .71 (Openness) and .88 (Neuroticism).

Charismatic leadership. In Sample 2, subordinates rated their leader using the 20-item Conger-Kanungo Scale (CKS; Conger et al., 1997) of charismatic leadership. The CKS consists of five subscales: strategic vision and articulation (7 items), personal risk (3 items), sensitivity to the environment (4 items), sensitivity to members’ needs (3 items), and unconventional behavior (3 items). Items were rated on a 5-point response format ranging from 1 (not characteristic) to 5 (very characteristic). Example items are: “Consistently generates new ideas for the future of the organiza- tion” and “Uses non-traditional means to achieve organizational goals.” A high level of internal consistency was obtained for the entire charismatic leadership scale (� � .92). The Cronbach alphas for the separate subscales were also acceptable to good: � � .92 for strategic vision and articulation, � � .85 for personal risk, � � .84 for sensitivity to the environment, � � .78 for sensitivity to members’ needs, and � � .63 for unconventional behavior. All descriptive statistics, correlations, and internal consistencies of the study variables in Sample 2 are reported in Table 1.

Results

Correlations between HDS charisma and self-rated and observer-rated BFI descriptions were examined in the Eugene- Springfield sample (cf. Hypothesis 1a). Results in Table 2 confirm that, across rater sources, behavioral indicators tapping into extraversion are highly relevant for describing charismatic personalities. For self- and observer ratings respectively, posi- tive associations were found with the following items: “Is talkative” (r � .36 and .26), “Is full of energy” (r � .30 and .28), “Is outgoing, sociable” (r � .34 and .31), and “Has an assertive personality” (r � .36 and .32); negative associations

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

115CHARISMATIC PERSONALITY

were found with the following: “Is reserved” (r � �.28 and �.30), “Tends to be quiet” (r � �.24 and �.28), and “Is sometimes shy, inhibited” (r � �.22 and �.29). Tapping right into one of the core features of charisma, “Generates a lot of enthusiasm” also had significant associations with HDS cha- risma (r � .38 and .30 for self- and observer ratings respec- tively, p � .001). Results further confirmed that charismatic

personalities are usually perceived as open and creative minds. For self- and observer ratings respectively, positive correlations were found between HDS charisma and personality descrip- tions, including: “Has an active imagination” (r � .28 and .31), “Is inventive” (r � .37 and .29), “Is original, comes up with new ideas” (r � .37 and .32), “Likes to reflect, play with ideas” (r � .34 and .20), and “Is curious about many different things” (r �

Table 1 Descriptive Statistics and Variable Intercorrelations in Study 1 (Sample 2: N � 204)

Variable M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1. Sexa — — 2. Age 45.96 8.62 �.09 3. Experienceb 24.01 8.50 �.10 .96���

4. HDS charismac 26.64 7.92 .10 �.02 �.04 .85 5. Neuroticism 2.26 .65 .08 �.17� �.17� �.14 .88 6. Extraversion 3.82 .53 .13 �.10 �.07 .44��� �.50��� .84 7. Openness 3.21 .51 .15� .13 .11 .19� �.03 .09 .71 8. Agreeableness 3.82 .44 .11 .22�� .25�� �.12 �.27��� .34��� .12 .84 9. Conscientiousness 4.09 .42 .07 .13 .17� .02 �.33��� .41��� �.03 .31��� .81

10. SVA 3.53 .82 .10 �.01 .01 .27��� �.05 .27��� .10 .20� .21�� .92 11. PR 2.39 1.03 .05 .04 .05 .28��� �.12 .27��� .14 .06 .06 .47��� .85 12. SE 3.69 .75 .11 .05 .09 .10 �.03 .16� .11 .25��� .25��� .67��� .34��� .84 13. SMN 3.67 .80 .14 �.02 .02 .13 .03 .16� .10 .22�� .21�� .65��� .31��� .68��� .78 14. UB 2.49 .84 .02 �.11 �.09 .29��� �.13 .29��� .17� �.07 .01 .43��� .61��� .33��� .32��� .63 15. CKS_total 3.26 .66 .11 �.01 .02 .29��� �.08 .30��� .15� .18� .21�� .90��� .69��� .79��� .76��� .66��� .92

Note. Bold values on the diagonal show the internal consistency of the relevant variable. Conger-Kanungo subscales are SVA � strategic vision and articulation; PR � personal risk; SE � sensitivity to the environment; SMN � sensitivity to members’ needs; and UB � unconventional behavior. a Sex is dummy coded such that 0 � male and 1 � female. b Work experience in years. c HDS � Hogan Development Survey; maximal score is 56 (raw scores). � p � .05. �� p � .01. ��� p � .001.

Table 2 Correlations Between HDS Charisma and Self-Rated and Observer-Rated Personality Descriptions in Study 1 (Sample 1: N � 156)

Description

HDS charisma (2007)

Description

HDS charisma (2007)

Observers (1998)

Self (1998)

Observers (1998)

Self (1998)

Is talkative .26�� .36��� Is emotionally stable, not easily upset �.09 .02 Tends to find fault with others �.01 .03 Is inventive .29��� .37���

Does a thorough job �.03 .03 Has an assertive personality .32��� .36���

Is depressed, blue �.11 �.12 Is original, comes up with new ideas .32��� .37���

Is reserved �.30��� �.28�� Can be cold and aloof .00 .09 Can be somewhat careless .17� .10 Not good-looking �.09 �.22��

Is relaxed, handles stress well .04 .20� Perseveres until the task is finished �.03 .17�

Is full of energy .28�� .30��� Values artistic, aesthetic experiences .17� .15 Starts quarrels with others .04 .11 Is sometimes shy, inhibited �.29�� �.22��

Can be moody �.02 .01 Is considerate and kind to almost everyone �.06 .02 Is a reliable worker �.06 �.14 Does things efficiently .10 .14 Can be tense .07 .10 Remains calm in tense situations �.05 .11 Is ingenious, a deep thinker .09 .24�� Prefers work that is routine �.34��� �.24��

Generates a lot of enthusiasm .30��� .38��� Is helpful and unselfish with others �.09 �.04 Has a forgiving nature �.08 .05 Is outgoing, sociable .31��� .34���

Physically attractive .03 .33��� Is sometimes rude to others .03 .14 Tends to be disorganized .03 �.03 Makes plans and follows through with them .12 .11 Worries a lot �.05 �.19� Likes to reflect, play with ideas .20� .34��

Has an active imagination .31��� .28�� Has few artistic interests �.16 �.12 Tends to be quiet �.28�� �.24�� Likes to cooperate with others �.05 .00 Is generally trusting �.09 �.07 Is easily distracted .03 .05 Tends to be lazy �.08 �.08 Is sophisticated in art, music, literature .09 .19�

Gets nervous easily �.08 �.24�� Is curious about many different things .26�� .23��

Note. Big Five Inventory (John & Srivastava, 1999) descriptions; HDS � Hogan Development Survey. � p � .05. �� p � .01. ��� p � .001.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

116 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

.23 and .26); negative correlations included: “Prefers work that is routine” (r � �.24 and �.34). Interestingly, charismatic personalities are more likely to be perceived as “Somewhat careless” by observers (r � .17), which could reflect the risk-taking behavior that is associ- ated with charisma. Finally, associations were found between HDS charisma and self-rated emotional stability descriptions such as “Is relaxed, handless stress well” (r � .20), “Worries a lot” (r � �.19), and “Gets nervous easily” (r � �.24), as well as self-rated attrac- tiveness items such as “Physically attractive” (r � .33) and “Not good-looking” (r � �.22).

In the sample of actual leaders (Sample 2), correlations between HDS charisma and charismatic leadership attributions were exam- ined. Consistent with our expectations (Hypothesis 1b), Table 1 demonstrates that leaders’ self-rated charismatic personality (HDS) was positively related to the subordinate-rated charismatic leadership composite (CKS), r � .29, p � .001. Regarding the CKS subscales, the expected positive relationship was confirmed for strategic vision and articulation, r � .27, p � .001, personal risk, r � .28, p � .001, and unconventional behavior, r � .29, p � .001. No significant correlations were found between HDS cha- risma and both sensitivity scales, r � .10 and .13, p � .05.

Finally, the availability of information about the leaders’ stand- ing on the Big Five domains allowed us to investigate the incre- mental validity of the HDS charismatic personality cluster to predict followers’ charismatic attributions above and beyond the Big Five personality traits. As such, a hierarchical regression analysis was conducted in which the Big Five traits were entered in a first step, followed by HDS charisma in a second step. The charismatic leadership composite (CKS), as rated by subordinates, served as the dependent variable. Results indicated that the set of Big Five traits was significantly related to ratings of charismatic leadership (R2 � .13, p � .01) and that HDS charisma demon- strated incremental validity over and above the Big Five person- ality traits (�R2 � .04, p � .01).

Discussion

In summary, the results of Study 1 speak for the significance of HDS charisma as a useful, trait-based measure of charisma. By relating HDS charisma to a set of fine-grained behavioral descrip- tions that reflect personality tendencies, a more in-depth under- standing of its content was obtained. Consistent with other re- search (Bass, 1998; Bono & Judge, 2004; De Vries, 2008), charismatic persons are typically described as energetic, assertive, talkative people who inspire others by generating a lot of enthu- siasm. Moreover, inventiveness, imaginativeness, and originality reflect their creative minds, whereas their carelessness may reflect risk-taking behaviors. Interestingly, stress-coping is perceived to be good by the participants themselves, while this is not necessar- ily the case for observers. However, it is possible that descriptions that reflect emotional stability are judged less accurately by peers because of a lower level of “trait visibility” (Funder & Dobroth, 1987). Finally, some associations between HDS charisma and self-rated— but not observer-rated—personality descriptions may indicate a self-enhancement bias, which is particularly character- istic for people with high levels of self-esteem (e.g., Judge, LePine, & Rich, 2006). For instance, their attractiveness and ingenious levels might be slightly overrated because observer reports do not reflect these characteristics. Note that HDS charisma and BFI were

administered with a 9-year time difference in the Eugene- Springfield sample, making the correlations around r � .30 quite substantial.

Moreover, a positive correlation was found between HDS cha- risma based on leaders’ self-reports and subordinate-rated charis- matic leadership styles. The observed relationship of r � .29 between HDS charisma and the CKS measure of charismatic leader behavior needs to be interpreted keeping in mind that different constructs (i.e., personality and leader behavior) were rated by different raters (cf. De Vries, 2012). Provided that other studies report levels of self-other agreement among leaders and subordinates on the exact same variables of r � .16 (e.g., for transformational leadership; see Judge et al., 2006), we consider this as convincing evidence that HDS charisma is a valid measure of charismatic personality.

Finally, we provided incremental validity evidence for HDS charisma, which accounts for an additional proportion of the variance in charismatic leadership beyond Big Five traits. Control- ling for Big Five traits was relevant in this context given that prior research had already established their association with charismatic leadership (Bono & Judge, 2004). We found that, despite the conceptual and empirical overlap with the FFM domains, most importantly with extraversion (i.e., r � .44 in the current study), the observed positive association between HDS charisma and charismatic leadership cannot be explained by Big Five traits, including extraversion, that has previously been found to be the most important personality correlate of charismatic leadership (Bono & Judge, 2004; De Vries, 2008).

Study 2

Having provided evidence for the validity of HDS charisma as a trait-based operationalization of charisma, the purpose of Study 2 is to test for the expected effects of charismatic personality on leader effectiveness. In line with the metatheoretical TMGT prin- ciple (Pierce & Aguinis, 2013), we hypothesized a curvilinear relationship between charismatic personality and leader effective- ness (Hypothesis 2). Such a perspective challenges the existing theories of charismatic leadership that advocate the “more is better” idea (e.g., Bass, 1985; Bass & Avolio, 1994), and is consistent with increasing evidence in the organizational and ap- plied personality literature in support of curvilinear relationships (e.g., Debusscher, Hofmans, & De Fruyt, 2014; Grijalva et al., 2015; Le et al., 2011).

Method

Procedure and participants. This study used data for 306 leaders, all employed by the same international aerospace com- pany. The data, including demographics, experience, and ratings of effectiveness, were gathered as part of an assessment process conducted for a training-and-development program. Participants went through the program in cohorts of approximately 25 to 30 leaders each, spaced out over 3 years. Most of the leaders were North American (95%) men (65.4%), and the mean age was 47.64 years (SD � 6.39). An average of 14 raters (with a minimum of 7 and a maximum of 31 raters) rated each leader in terms of overall effectiveness, including at least one subordinate, one peer, and one superior. Taken together, 4,345 coworkers participated in this

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

117CHARISMATIC PERSONALITY

study, comprising 666 superiors, 1,659 peers, and 2,020 subordi- nates. The leaders had on average 16.01 years (SD � 7.23) of managerial experience and had a mean tenure in the current job of 2.51 years (SD � 2.54). Leaders occupied different managerial levels ranging from supervisors (30%) to general managers (20%).

Measures. All descriptive statistics, correlations, and internal consistencies of the study variables are reported in Table 3.

Demographic and control variables. Based on significant correlations with the study criteria (see Table 3), leader sex and managerial experience were used as relevant control variables in statistical tests of the hypotheses.

Charismatic personality. Leaders completed the 56 items from the Bold, Mischievous, Colorful, and Imaginative scales of the HDS (Hogan & Hogan, 2009). HDS charisma scores were expressed in terms of percentiles ranging from 1 to 100 (i.e., relative to the general population of working adults in the U.S.). Percentile scores help to interpret the relationship between char- ismatic personality and leader effectiveness by referencing person- ality scores to a normative population (e.g., Is the optimal level of charisma near the normative mean, slightly elevated, or highly elevated?). The internal consistency of the HDS charisma scale was .85.

Overall leader effectiveness. A single-item of the Leadership Versatility Index (LVI; Kaiser et al., 2010) was used to measure overall leader effectiveness. The item reads: “Please rate this individual’s overall effectiveness as a leader on a ten-point scale where 5 is adequate and 10 is outstanding.” Leaders (N � 306), along with their subordinates (N � 2,020), peers (N � 1,659), and superiors (N � 666), provided overall leader-effectiveness ratings. An average of seven subordinates, five peers, and two superiors rated each of their respective leaders. Based on a composite of the ratings from superiors, peers, and subordinates, an aggregated observer rating was computed for overall leader effectiveness. This aggregate score represents the grand mean of the mean ratings for the observer rating groups, excluding self-ratings. To obtain this aggregated score, the mean ratings across raters within the supe- rior, peer, and subordinate groups were calculated separately. For example, to obtain an aggregate score of overall effectiveness for

a particular leader, the mean ratings of the superiors (6.50), peers (7.71), and subordinates (8.75) were summed (22.96), and divided by three (7.65). This procedure results in an overall score that unit-weights each observer-rater perspective and, according to Oh and Berry (2009), is the most valid way to aggregate ratings from coworkers to an overall score. To provide additional justification for this aggregation method, the rwg(j) interrater agreement coef- ficient (James et al., 1984) and the one-way random effects intra- class correlation coefficient (ICC; McGraw & Wong, 1996) were computed within superior, peer, and subordinate groups, as well as across these three sources (LeBreton, Burgess, Kaiser, Atchley, & James, 2003). The results in the Appendix indicate that the level of similarity across superior, peer, and subordinate ratings is suffi- ciently high to support aggregation (LeBreton & Senter, 2008).

Kaiser et al. (2010) summarized validity and reliability evidence for the single item overall effectiveness measure, showing that it has substantial correlations with other, multiitem scales of leader effectiveness (e.g., r � .86 with Quinn, Spreitzer, & Hart’s, 1991, managerial effectiveness scale; r � .73 with Tsui’s, 1984, mana- gerial reputational effectiveness scale). In the current study, the correlation between different rater sources was r � .34 for superior-peer, r � .20 for superior-subordinate, and r � .22 for peer-subordinate ratings, demonstrating a modest degree of con- vergent validity of the single-item measure that is similar in magnitude to meta-analytic estimates of cross-source correlations on multiitem scales (Conway & Huffcut, 1997).

Results

To test for curvilinearity in the relationship between charismatic personality and leader effectiveness (Hypothesis 2), we conducted a hierarchical regression analysis. Prior to the analysis, we cen- tered the charismatic personality scores and then computed the squared term based on the centered scores. The control variables (i.e., sex and managerial experience) were entered in a first step, followed by charismatic personality (centered) in a second step, and the squared term for charismatic personality was entered in a

Table 3 Descriptive Statistics and Variable Intercorrelations in Study 2 (N � 306)

Variable M SD 1 2 3 4 5 6 7 8

1. Sexa — — 2. Age 47.64 6.39 �.19��

3. Experienceb 16.01 7.23 �.24��� .70���

4. Charismatic personalityc 48.82 18.13 �.01 �.01 .12� .85 5. Leader effectiveness

(self) 7.62 .77 �.04 .07 .15� .29���

6. Leader effectiveness (observers) 8.22 .43 �.14� .08 .17�� .05 .21���

7. Leader effectiveness (subordinates) 8.31 .64 �.11 .06 .14� .05 .14� .69���

8. Leader effectiveness (peers) 8.03 .53 �.14� .03 .08 �.03 .10 .69��� .22���

9. Leader effectiveness (superiors) 8.31 .66 �.05 .08 .13� .08 .20��� .75��� .20��� .34���

Note. Bold values on the diagonal show the internal consistency of the relevant variable. a Sex is dummy coded such that 0 � male and 1 � female. b Managerial experience in years. c Maximal score is 100 (percentiles). � p � .05. �� p � .01. ��� p � .001.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

118 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

third and final step. The aggregated observer rating for overall effectiveness served as the dependent variable.

The results (Table 4, Model 1) first show that more experienced leaders were perceived as more effective (Step 1: � � .14, p � .05). Next, we added the linear term for charismatic personality (Step 2). This revealed that charismatic personality was not lin- early related to leader effectiveness (� � .04, p � .05). Relevant to Hypothesis 2, however, are the results of Step 3, where both the linear and the squared term were included. Consistent with the hypothesized inverted-U curvilinear effect, the squared term for charismatic personality was negative and significant (� � �.24, p � .001). The robustness of this finding was further illustrated by significant curvilinear effects in each of the three rater groups (see Models 2 to 4 in Table 4). In other words, leaders with both low and high charismatic personalities were perceived as being less effective than leaders with moderate levels of charisma, and this was true according to all three the rater groups.

On exploratory grounds, a similar regression analysis was con- ducted to test whether the same trend was observed for the asso- ciation between charismatic personality and self-perceived leader effectiveness. Again, control variables were entered (Step 1), fol- lowed by charismatic personality (Step 2), and the squared term for charismatic personality (Step 3). The only difference was that the dependent variable was not other-perceived but self-perceived overall effectiveness. Table 4 (Model 5) indicates that more expe- rienced leaders also perceived themselves as more effective (� � .15, p � .05). More importantly, however, Step 2 showed that charismatic personality was linearly related to self-perceived ef- fectiveness (� � .27, p � .001), whereas the squared term for charismatic personality in Step 3 was not significant (� � .02, p � .05). This indicates that higher charisma levels are consistently associated with higher self-perceived effectiveness. Figure 2 shows the regression lines for the significant quadratic and linear effects for observer and self-ratings of perceived leader effective- ness, respectively. In this figure, it can be seen that—according to relevant others—moderate, or slightly elevated, levels of charisma (i.e., around percentile 60) were associated with the highest effec- tiveness levels.

Discussion

In Study 2 the relationship between leaders’ charismatic per- sonality and overall effectiveness was examined. Consistent with our expectations, we found that leader charisma related to observer-rated effectiveness in a curvilinear way, with moderate levels being more effective than low or high levels of charismatic personality. Moreover, the curvilinear relationship held across the three observer groups (i.e., subordinates, peers, and supervisors). Next, we found that this relationship was different for self-rated overall effectiveness. Consistent with self-enhancement theories (e.g., Alicke & Govorun, 2005; Leary, 2007), a positive linear relationship was found, implying that higher charismatic tenden-

Table 4 Hierarchical Regression Analyses Examining the Associations Between Charismatic Personality and Overall Leader Effectiveness in Study 2 (N � 306)

Step

Overall leader effectiveness

Model 1: Aggregated observer rating

Model 2: Subordinates

Model 3: Peers

Model 4: Superiors

Model 5: Self

� SE (b) �R2 � SE (b) �R2 � SE (b) �R2 � SE (b) �R2 � SE (b) �R2

Step 1 .04�� .02� .02� .02 .02�

Sex �.10 .05 �.08 .08 �.13� .07 �.02 .08 .00 .10 Experience .14� .00 .12� .01 .05 .00 .12� .01 .15� .01

Step 2 .00 .00 .00 .00 .07���

Charisma .04 .00 .04 .00 �.04 .00 .07 .00 .27��� .00 Step 3 .05��� .06��� .02� .01� .00

Charisma .08 .00 .09 .00 �.02 .00 .09 .00 .27��� .00 Charisma2 �.24��� .00 �.24��� .00 �.14��� .00 �.12� .00 .02 .00

Note. Sex is dummy coded such that 0 � male and 1 � female. � p � .05. �� p � .01. ��� p � .001.

self-rating observer-rating

charismatic personality

O ve

ra ll e

ffe ct

ive ne

ss

Figure 2. Overall leader effectiveness as a function of charismatic per- sonality (percentiles): aggregated observer-ratings versus self-ratings of overall effectiveness (Study 2).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

119CHARISMATIC PERSONALITY

cies were consistently related to higher self-perceived effective- ness. This finding is also in line with other research demonstrating that leaders with high self-esteem typically overrate their perfor- mance on a variety of criteria (e.g., Judge et al., 2006).

Study 3

Study 2 showed that charismatic personality related in a curvi- linear way to observer-rated leadership effectiveness. In Study 3, we sought to replicate and extend these results in two important ways. First, the potential moderating role of the leader’s level of adjustment in this curvilinear relationship is tested (Hypothesis 3). Second, the mechanisms underlying this curvilinear association are explored. Specifically, a process model is tested describing the association between charismatic personality and perceived leader effectiveness, as mediated through leader behaviors. This is in line with recent calls to integrate trait and behavioral leadership theo- ries into process-type models which aim to clarify the effects of distal individual differences (e.g., traits and styles) on leader outcomes through more proximal leader behaviors (Antonakis, Day, & Schyns, 2012; DeRue et al., 2011; Dinh & Lord, 2012; Zaccaro, 2012). As charismatic leaders express high performance expectations and push people hard to get there (Waldman et al., 2001), potentially at the expense of being sensitive to followers’ feelings (Deluga, 1997), we expect charisma to be positively associated with forceful (Hypothesis 4a), and negatively associated with enabling behavior (Hypothesis 4b). Moreover, charismatic leaders display exceptional strategic vision (Bass, 1985), but they can become so excited about their ideas, that they get stuck in the process of implementing their big visions (Conger, 1990). There- fore, we expect charisma to be positively associated with strategic (Hypothesis 5a), and negatively associated with operational be- havior (Hypothesis 5b). We argue that the effectiveness of various levels of leader charisma is mediated by these leader behaviors (Hypothesis 6). From an ABC perspective (Busse et al., 2016), enabling costs may outweigh forceful benefits, and/or operational costs may outweigh strategic benefits, such that beyond a certain optimal level, further increases in charismatic personality might reduce the effectiveness of leaders.

Method

Procedure and participants. Development-center test data were obtained from an international consultancy firm specialized in leader assessment and executive coaching. Leaders (N � 287) from 23 different countries (e.g., 53% North America, 33% West- ern Europe, 8% Africa, 4% East Asia) participated in the study. To obtain a true 360 view of the leadership criteria (i.e., leader effectiveness and the four leader behaviors), only leaders who were rated at least once by each of three rater categories (i.e., superiors, peers, and subordinates) were included. An average of 11 raters (1 superior, 5 peers, and 5 subordinates; minimum of 5 and a maximum of 27 raters) rated each leader in terms of overall effectiveness and leader behaviors. Taken together, 3,052 cowork- ers participated in this study, comprising 309 superiors, 1,380 peers, and 1,363 subordinates. Most of the leaders were male (81%) and the mean age was 45.37 years (SD � 6.78). They reported an average of 15.78 years (SD � 7.77) managerial expe- rience and had a mean tenure in their current job of 2.99 years (SD � 3.40). Leaders occupied different managerial levels—from supervisors (12.2%) to general managers (15%)—and most of them worked in business organizations. Part of the data were also used in Kaiser et al. (2015). Different than Kaiser et al. (2015), which focused on relationships between the 11 individual HDS traits and the four LVI leader behaviors, the current study focused on the HDS “charismatic cluster” as a measure of charismatic personality and its relation to overall leader effectiveness. Further, although the four LVI leader behaviors served as the main criteria in Kaiser et al. (2015), they are examined as mediators in the current study.

Measures. All descriptive statistics, correlations, and internal consistencies of the study variables are reported in Table 5.

Demographic and control variables. Based on significant correlations with the study criteria (see Table 5), age and mana- gerial experience qualified as relevant control variables. However, because of the strong correlation between age and experience (r � .72, p � .001), and because the impact of experience on leader effectiveness has already been established (Avery, Tonidandel, Griffith, & Quiñones, 2003), only managerial experience was

Table 5 Descriptive Statistics and Variable Intercorrelations in Study 3 (N � 287)

Variable M SD 1 2 3 4 5 6 7 8 9 10

1. Sexa — — 2. Age 45.37 6.67 .07 3. Experienceb 15.98 7.73 �.02 .72���

4. Charismatic personalityc 59.16 20.62 �.02 �.12� �.06 .84 5. Forceful �.06 .47 .10 �.01 �.01 .08 .93 6. Enabling �.24 .38 �.05 �.10 �.06 .06 �.70��� .92 7. Strategic �.29 .36 �.10 �.16�� �.06 .28��� .45 �.06 .92 8. Operational �.12 .25 .07 .10 .04 �.31��� .01 �.01 �.30��� .80 9. Leader effectiveness (observers) 7.73 .84 �.08 �.10 .06 .03 .06 .34��� .45��� .18�� —

10. Leader effectiveness (self) 7.72 .85 �.10 .01 .14� .17�� .08 .12� .18�� .05 .32��� —

Note. Bold values on the diagonal show the internal consistency of the relevant variable. a Sex is dummy coded such that 0 � male and 1 � female. b Managerial experience in years. c Maximal score is 100 (percentiles). � p � .05. �� p � .01. ��� p � .001.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

120 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

included as a control variable. As in Study 2, we also controlled for sex.

Charismatic personality. As in Study 2, leaders completed the 56 items from the Bold, Mischievous, Colorful, and Imaginative scales of the HDS (Hogan & Hogan, 2009). The internal consis- tency of HDS charisma was .84.

Adjustment. Leaders completed the 37-item adjustment scale of the Hogan Personality Inventory (HPI; Hogan & Hogan, 2007), which corresponds to the FFM Emotional Stability dimension and can be described as the degree to which a person appears calm and self-accepting or, conversely, self-critical and tense. The internal consistency of the adjustment scale was .85.

Leadership criteria. The Leadership Versatility Index (LVI; Kaiser et al., 2010) was used to measure both overall leader effectiveness (see Study 2) and specific leader behaviors. Within the LVI, leader behaviors can be covered by two pairs of opposing dimensions in leadership: Forceful versus enabling leadership, and strategic versus operational leadership. Each of the four dimen- sions are surveyed by means of 12 items using the “too little/too much” response format ranging from �4 (much too little), to 0 (the right amount), to 4 (much too much). Methodologically, the LVI provides a unique operationalization of these behaviors that goes beyond traditional rating scales and complements our theoretical grounds that emphasize the need for balance between deficiency and excess. Sample items are: “Takes charge—in control of his/her area of responsibility” (Forceful), “Participative—includes people in making decisions” (Enabling), “Spends time and energy on long-term planning—future-oriented” (Strategic), and “Orga- nized—takes a methodical approach to getting things done” (Op- erational; Kaiser et al., 2010).

Both leaders (N � 287) and their respective coworkers (N � 3,052) completed the LVI. To compute aggregated observer rat- ings for overall leader effectiveness and each of the four leader behaviors, a similar procedure was followed as in Study 2, such that each rating group (i.e., subordinates, peers, and superiors) was equally weighted in the observer score (Oh & Berry, 2009). Based on interrater agreement (rwg(j)) and interrater reliability (ICC) coefficients (James et al., 1984; McGraw & Wong, 1996) within and across these three sources, support was found to justify this aggregation (see the Appendix).

Cronbach alphas of the aggregated LVI leader behavior dimen- sions were .93 for forceful behavior, .92 for both enabling and strategic leader behavior, and .80 for operational behavior. Fre- quencies of leaders being perceived as doing “too little,” doing “the right amount,” and doing “too much” of each of the four

leader behaviors are displayed in Table 6, along with the mean charismatic personality score within each group of leaders. Gen- erally, the frequencies of leaders underdoing leader behaviors are the highest (e.g., 74% of the leaders perform too little strategic behavior), compared with leaders overdoing and leaders doing the right amount of each of the leader behaviors.

Results

Relating charismatic personality to leader effectiveness. To test for curvilinearity in the relationship between charismatic personality and leader effectiveness, we followed the same ana- lytical procedure as in Study 2. The results in Table 7 indicate a positive and linear relationship between charismatic personality and self-perceived effectiveness (Step 2: � � .18, p � .01), whereas no quadratic effect was found (Step 3: � � .06, p � .05). Conversely, when testing the relationship with observer-rated leader effectiveness, the linear term for charismatic personality was not significant (Step 2: � � .04, p � .05), whereas the squared term was negative and significant (� � �.15, p � .05). The regression lines for the significant quadratic and linear effects of respectively observer-rated and self-rated leader effectiveness are highly similar to those reported in Figure 2.

Adjustment as a moderator. A hierarchical regression anal- ysis was conducted to test whether the quadratic relationship between charismatic personality and leader effectiveness is mod- erated by adjustment. This answers the question whether the in- flection point in the curvilinear relationship between charismatic personality and observer-rated effectiveness depends on the lead- er’s level of self-regulation. Sex and managerial experience were first entered in the regression as control variables (Step 1), fol- lowed by charismatic personality (centered) and adjustment (cen- tered; Step 2), and the squared term for charismatic personality (Step 3). In a final step, interaction terms between (a) adjustment and charismatic personality and (b) adjustment and the quadratic term of charismatic personality were entered (Step 4).

Relevant to Hypothesis 3 are the two interaction terms reported in Step 4 of the regression. As can be seen in Table 8, the interaction term between adjustment and the linear effect of char- ismatic personality approaches conventional levels of significance (� � .13, p � .06), while the interaction term between adjustment and the quadratic effect of charismatic personality is not signifi- cant (� � �.04, p � .05). To further interpret this relationship, we plotted the moderation effect in Figure 3. In line with our hypoth- esis, this figure shows that the inflection point after which the

Table 6 Frequencies (%) of the Three Categories of Leaders in Study 3 (N � 287), and Mean Charismatic Personality Scores (MCharisma) Within the Three Categories of Leaders

Category

Forceful Enabling Strategic Operational

% Mcharisma % Mcharisma % Mcharisma % Mcharisma

Too little 54 59.29 61 58.68 74 57.34 57 63.01 The right amount 13 58.61 23 60.31 15 63.67 19 59.23 Too much 33 59.16 16 59.33 11 65.65 24 49.90

Note. Leaders were categorized as “the right amount” when the LVI scores were within plus/minus three Standard Errors of Measurement around 0, because scores within this range are statistically indistinguishable from “0” at p � .001 (Ghiselli, Campbell, & Zedeck, 1981); LVI scores exceeding this range � “too much”; below this range � “too little.”

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

121CHARISMATIC PERSONALITY

charisma-effectiveness relationship turns asymptotic and negative is lower for leaders who score low on adjustment, compared with leaders scoring high on adjustment.

Leader behaviors as mediating mechanisms. Prior to testing the mediation hypothesis, we investigated the relationships be- tween charismatic personality and each of the four leader behav- iors (i.e., the mediators). Four hierarchical regressions were con- ducted, with sex and managerial experience entered in the first step (i.e., the control variables) and charismatic personality in the second step. Consistent with our expectations (Hypothesis 5), we found charismatic personality to be positively associated with strategic leadership (� � .27, p � .001) and negatively with operational leadership (� � �.31, p � .001), indicating that higher charisma scores are associated with a tendency to do more strate- gic behavior and less operational behavior. No significant associ- ations were found between charismatic personality and the two interpersonal leadership dimensions (� � .08 and .06 for forceful

and enabling respectively, p � .05), which is inconsistent with Hypothesis 4.

A visualization of the significant effects helps in refining the interpretation of these associations (see Figure 4). The point where the regression line crosses zero (i.e., the right amount) on the leader behavior scale corresponds to the percentile score on char- ismatic personality associated with the optimal amount of the leader behavior. The positive relation between charismatic person- ality and strategic behavior (Panel A) illustrates that, whereas low charisma levels correspond with a tendency to underdo strategic behavior, high charisma levels correspond with doing the right amount of strategic behavior. This figure thus reveals that leaders low in charisma are more inclined to show too little strategic behavior, rather than that highly charismatic leaders are more inclined to show too much strategic behavior. With regard to the negative association between charisma and operational behavior, Figure 4 (Panel B) shows that high charisma corresponds with a tendency to underdo operational behavior, whereas low charisma levels correspond with an optimal amount of operational behavior. Hence, high charisma levels are associated with higher strategic behaviors (but not too much) and a lack of operational behaviors (i.e., too little).

Hypothesis 6 predicted that LVI leader behaviors (i.e., the mediators M) mediate the curvilinear relationship between charis- matic personality (i.e., the independent variable X) and observer- rated leader effectiveness (i.e., the dependent variable Y). The mediation hypothesis was tested using path modeling in Mplus version 7.3. Because charismatic personality was related to stra- tegic and operational leadership but not to forceful and enabling leadership, only the two business-related behaviors were included in the model. In particular, strategic and operational behavior were predicted by the linear and squared effect of charismatic person- ality, while leader effectiveness was predicted by the linear and squared effects of strategic and operational leader behavior and the linear and squared effect of charismatic personality (see Figure 5). In this model, the predictor and mediators were centered before computing the squared effects, and the linear and squared effects of strategic and operational leader behavior were allowed to cor- relate. Together, this yields the following set of equations:

Mstrategic � istrategic � a1 strategicX � a2 strategicX 2 � estrategic

Table 7 Hierarchical Regression Analyses Examining the Associations Between Charismatic Personality and Overall Leader Effectiveness in Study 3 (N � 287)

Step

Overall leader effectiveness

Model 1: Aggregated observer rating

Model 2: Self

� SE (b) �R2 � SE (b) �R2

Step 1 .01 .03�

Sex �.08 .13 �.10 .13 Experience .06 .01 .14� .01

Step 2 .00 .03��

Charisma .04 .00 .18�� .00 Step 3 .02� .00

Charisma �.01 .00 .20 .00 Charisma2 �.15� .00 .06 .00

Note. Sex is dummy coded such that 0 � male and 1 � female. � p � .05. �� p � .01.

Table 8 Hierarchical Regression Analyses Examining the Moderating Effect of Adjustment in the (Curvilinear) Relationship Between Charismatic Personality and Overall Leader Effectiveness in Study 3 (N � 287)

Step

Observer-rated leader effectiveness

� SE (b) �R2

Step 1 .01 Sex �.08 .13 Experience .06 .01

Step 2 .01 Charisma .04 .00 Adjustment .08 .00

Step 3 .02�

Charisma2 �.14� .00 Step 4 .02†

Adjustment Charisma .13† .00 Adjustment Charisma2 �.04 .00

Note. Sex is dummy coded such that 0 � male and 1 � female. † p � .06. � p � .05.

4

5

6

7

8

9

Low charisma High charisma

)srevresbo( ssenev itceffe llar ev O

Low adjustment High adjustment

Figure 3. Moderating effect of adjustment in the curvilinear relationship between charismatic personality and observer-rated leader effectiveness (Study 3).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

122 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

Moperational � ioperational � a1 operationalX � a2 operationalX 2 � eoperational

Y � iY � b1Mstrategic � b2Mstrategic 2 � b3Moperational � b4Moperational

2

� c1X � c2X 2 � eY

With respect to the relationship between the predictor and the mediators (i.e., the a-paths in Figure 5), we found that charis- matic personality was positively related to strategic (� � .26, p � .001) and negatively to operational leader behavior (� � �.33, p � .001), whereas no curvilinear effects were found (� � �.07 and .01 for strategic and operational leader behavior respectively, p � .05). Regarding the relationships between the mediators and the outcome (i.e., the b-paths in Figure 5), we found that both the linear (� � .49, p � .001, and � � .27, p � .001) and the curvilinear components (� � �.16, p � .05, and � � �.13, p � .05) of strategic and operational leader behavior related to leader effectiveness. This suggests that higher levels of strategic and operational behavior posi- tively relate to perceived effectiveness but only up to a point that there is no additional benefit of more strategic and opera- tional behaviors (i.e., positive flattening curves). Finally, the

direct effect of charisma on leader effectiveness (i.e., the c-paths in Figure 5) was nonsignifcant (� � �.06 and �.07 for the linear and quadratic effect respectively, p � .05), which indicates that the relationship between charismatic personality and leader effectiveness is fully mediated by strategic and operational behaviors.

To formally test the indirect mediation effects of charismatic personality on leader effectiveness via strategic and operational leader behavior, we tested the indirect effects following the approach of Hayes and Preacher (2010), which was specifically developed for testing nonlinear mediation. Because the a-path is linear, whereas the b-path is quadratic, the mediation— or indirect— effect was computed as follows: � � a(b1 2b2(i aX)); see Hayes and Preacher (2010), p. 633. As can be seen in this formula, the mediation effect depends on the value of the predictor (i.e., X is part of the formula), which means that the effect of charismatic personality on leader effectiveness through strategic and operational leader behavior depends on the leader’s level of charisma. For this reason, Hayes and Preacher (2010) referred to the indirect effect as the instanta-

A B

charismatic personality

Strategic

To o

lit tle

To o

m uc

h Th

e r ig

ht

am ou

nt To

o lit

tle To

o m

uc h

Th e r

ig ht

am

ou nt

Operational

To o

lit tle

To o

m uc

h Th

e r ig

ht

am ou

nt To

o lit

tle To

o m

uc h

Th e r

ig ht

am

ou nt

charismatic personality

Figure 4. Regression lines for charismatic personality predicting strategic (Panel A) and operational (Panel B) leader behavior. The regression lines are drawn for the range in which we have charisma observations (Study 3).

charismatic personality

charismatic personality 2

Strategic behavior

Strategic behavior 2

Operational behavior

Operational behavior 2

Overall leader effectiveness

c2

c1

a1(op)

a2(op)

a1(st)

a2(st)

b1

b2

b3

b4

ey

eop est

Figure 5. Path model for testing the nonlinear mediation between charismatic personality and overall leader effectiveness through strategic (st) and operational (op) leader behavior (Study 3).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

123CHARISMATIC PERSONALITY

neous indirect effect, which is the effect of the predictor on the outcome through the mediator(s) at a specific value of the predictor. This instantaneous indirect effect was tested for different levels of charismatic personality using nonparametric bootstrapping (N � 1,000). A graphical representation of the instantaneous indirect effects, together with their 95% confi- dence intervals, is shown in Figure 6.

As can be seen in Panel A of Figure 6, and in line with the positive linear a- and b-paths, we found a positive instantaneous indirect effect of charismatic personality on leader effectiveness through strategic leader behavior. Moreover, combining the posi- tive linear and negative quadratic b-paths yields a positive flatten- ing curve, implying that the positive effect of charisma on effec- tiveness weakens at higher levels of charisma. In turn, Panel B shows that the instantaneous indirect effect of charismatic person- ality on leader effectiveness through operational leader behavior is negative (which is in line with the negative linear a- and positive linear b-path). Moreover, because of the negative curvilinear b-path, the effect becomes more negative when charismatic per- sonality increases.

Combined, these findings clearly reveal the mechanisms that un- derlie the curvilinear relationship between charismatic personality and leader effectiveness. For example, for leaders with a centered cha- risma score of �30 (i.e., low charismatic personality), the predicted instantaneous indirect effect for strategic behavior is .005, which

translates into a negative effect of low charisma on effectiveness via strategic behavior (i.e., �30 .005 � �.150), whereas the predicted instantaneous indirect effect for operational behavior is �.002, which translates into a positive effect of low charisma on effectiveness via operational behavior (i.e., �30 �.002 � .060). Together, this yields a combined negative effect of low charisma on leader effec- tiveness of �.150 .060 � �.090, in which the negative effect is entirely due to the lack of strategic behavior. For leaders with an average charismatic personality (i.e., a centered score of 0), the predicted instantaneous indirect effect for strategic behavior is .005, while the predicted instantaneous indirect effect for operational be- havior is �.003. Together, the effect of average charisma on leader effectiveness equals 0 [(.005 0) (�.003 0)]. Finally, for leaders with a centered charisma score of 30 (i.e., high charismatic personality), the predicted instantaneous indirect effect for strate- gic behavior is .004, which translates into a positive effect of high charisma on effectiveness via strategic behavior (i.e., 30 .004 � .120), whereas the predicted instantaneous indirect effect for op- erational behavior is �.005, which translates into a negative effect of high charisma on effectiveness via operational behavior (i.e., 30 �.005 � �.150). Together, this yields a combined negative effect of high charisma on leader effectiveness of .120 (�.150) � �.030, in which the negative effect is entirely due to the lack of operational behavior.

A B

Strategic Operational

charismatic personality

tceffe tceridni suoenatnatsni

in st

an ta

ne ou

s in

di re

ct e

ffe ct

charismatic personality

Figure 6. The instantaneous indirect effects of charismatic personality on overall leader effectiveness through strategic (panel A) and operational (panel B) leader behavior at specific values of charisma (centered), together with the 95% confidence intervals (i.e., dotted lines; Study 3).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

124 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

Discussion

In Study 3 we replicated the curvilinear relationship between charismatic personality and leader effectiveness. Moreover, con- ditions under which and processes through which this curvilinear relationship occurs were examined. Hogan and Hogan (2007) pointed to the crucial role of adjustment in professional contexts. The presented study showed that, when a leader’s level of adjust- ment is high, the inflection point after which the relation with effectiveness turns asymptotic and negative occurs at higher levels of charisma. This means that a high level of adjustment can alleviate the negative effects associated with high charisma levels.

Finally, the overall results of the mediation analysis revealed that leaders low on charisma are perceived to be less effective than leaders with an average charisma level because they lack strategic behaviors, while leaders high on charisma are perceived to be less effective because they lack operational behaviors. In line with the ABC framework (Busse et al., 2016), the instantaneous-indirect- effect approach showed that the TMGT effect results from two competing indirect effects: a positive indirect effect via strategic behavior (i.e., benefit variable) and a negative indirect effect via operational behavior (i.e., cost variable). At high charisma levels, the beneficial effect of highly strategic behavior is still there, but these benefits are offset by the operational costs associated with high charisma levels.

General Discussion

Our work aimed to advance the understanding of leader cha- risma by (a) introducing a trait-based model of charisma, (b) demonstrating curvilinear relationships between charismatic per- sonality and leader effectiveness, (c) studying the boundary con- ditions under which the nature of this relationship may change, and (d) examining the processes through which this relationship may occur.

The first objective was addressed in Study 1. Results of this study generally speak for the significance of HDS charisma as a useful, trait-based measure of charisma. In addition to conceptual arguments for the relevance of this constellation of personality traits, we found significant correlations between HDS charisma and fine-grained behavioral descriptions of charisma that were both self-rated and observer-rated 9 years earlier. Moreover, a significant correlation was found between HDS charisma based on leaders’ self-reports and subordinate-rated charismatic leadership styles (CKS charismatic leadership), and we provided incremental validity evidence for HDS charisma, which accounts for an addi- tional proportion of the variance in charismatic leadership beyond Big Five traits.

Delving deeper into the associations between the two charisma ratings further showed that no significant relationships were ob- served between HDS charisma and both CKS sensitivity subscales. This reinforces the idea that the current measure of charismatic personality focuses on communicating vision, unconventional be- havior, and personal risk taking but does not necessarily cover attention to other people’s needs or assessing events in the external environment. Although an extensive discussion of this finding transcends the purposes of this study, it is relevant to point out that this may shed light on the difference between charismatic and transformational leadership. Consistent with Bass’s (1985) con- ceptual differentiation between charisma and individualized con-

sideration, this finding might indicate that sensitivity or individual consideration is a critical factor distinguishing charismatic from transformational leaders. This idea was also expressed by Bono and Judge (2004), who stated that “the modesty and kindness of agreeable individuals is not the hallmark of charismatic leaders” (p. 903).

Moreover, existing theories and research on leader charisma have in common that they all tend to depart from a rather simplistic “more is better” perspective. With evidence increasing in favor of an alternative “too much of a good thing” perspective in the fields of applied personality, organizational behavior, and management science, a second objective of our work was to investigate whether leaders can be too charismatic. Consistent with our expectations, the results in both Study 2 and Study 3 revealed a nonlinear relationship between charismatic personality and observer-rated overall leader effectiveness, supporting the idea that moderate levels are better than low or high levels of charismatic personality. This finding aligns with leadership research demonstrating the dynamic of strengths becoming weaknesses when overusing them (e.g., Kaiser & Hogan, 2011; McCall, 2009). Striking in this regard is the divergent effect of charismatic personality on self-rated overall effectiveness, which was positive and linear in both stud- ies, implying that higher charismatic tendencies consistently go together with higher self-perceived effectiveness. The explanation for this finding may be found in self-enhancement theory (Leary, 2007), which states that people are motivated to protect their levels of self-esteem, especially in potentially threatening situations like self-assessment. This may explain why the highly charismatic, with typically high levels of self-esteem, might be blind to their weaknesses and exaggerate their strengths.

A third objective, which was addressed in Study 3, was to investigate the boundary conditions under which this curvilinear relationship may change. A central tenet in the TMGT principle is that the inflection point after which the relationship turns asymp- totic is context-specific (Pierce & Aguinis, 2013). Although we did not take situational variables into account, such as stressful situa- tions, we did take into account how one usually reacts to stressful situations (i.e., adjustment). Consistent with Hogan and Hogan (2007), we found that adjustment plays an important moderating role. When the leader’s level of adjustment is high, the inflection point after which the relation with effectiveness turns asymptotic and negative occurs at higher levels of charisma. This means that the “damage” of being highly charismatic depends on other traits that the leader has: A high level of adjustment can buffer the negative effects associated with high charisma levels.

A final objective of our work, also addressed in Study 3, was to explore the mechanisms that account for the nonlinear relationship between charismatic personality and observer-rated effectiveness. For this purpose, both interpersonal (i.e., forceful and enabling) and organizational (i.e., strategic and operational) behaviors were considered as potential outcomes of charismatic personality, but only the latter were significantly associated with charisma levels. Using path modeling, we found that strategic and operational leader behaviors fully mediate the curvilinear relationship between charismatic personality and overall leader effectiveness. More- over, the instantaneous-indirect-effect approach clearly provides insight into the mechanisms driving this curvilinear relationship. Specifically, it was found that different leader behaviors are accountable for the curvilinearity between charismatic personality

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

125CHARISMATIC PERSONALITY

and overall effectiveness at different charisma levels. At lower charisma levels, the lack of strategic leader behavior makes leaders less effective than moderately charismatic leaders (cf. the left part of the inverted U-shape in Figure 2). At higher charisma levels, on the other hand, a clear lack of operational leader behavior reduces leader effectiveness (cf. the right part of the inverted U-shape in Figure 2).

Research Implications

The current study departed from a research model, integrating leader characteristics, leader behaviors, and finally relevant out- comes. This kind of overarching framework may help to structure this field of study and, eventually, facilitate the accumulation of knowledge in this domain. The specific conceptualizations adopted in the present study for leader characteristics (i.e., charismatic personality, adjustment), leader behavior (interpersonal and orga- nizational), and leader outcomes (effectiveness) allowed us to shed light on the general but highly prevalent question: “What breaks a leader?” (cf. Ames & Flynn, 2007; Hogan, Hogan, & Kaiser, 2010). Linking charismatic tendencies to leadership behaviors revealed that charisma is most strongly associated to business- related behaviors. Whereas conventional wisdom suggests that highly charismatic leaders fail for interpersonal reasons like arro- gance, self-centeredness, and not caring about others (Blair, Hoff- man, & Helland, 2008; O’Boyle, Forsyth, Banks, & McDaniel, 2012), our findings suggest that business-related behaviors, more than interpersonal behavior, drive leader effectiveness ratings.

Interestingly, the mediation analysis pinpoints the exact role of these business-related behaviors in the curvilinear relationship between charismatic personality and leader effectiveness. For op- erational behavior, the results indicate that higher charisma scores are associated with a lack of operational behavior and that this impacts negatively on leader effectiveness. Moreover, this effect is curvilinear, meaning that the detrimental effects of this lack of operational behaviors become even stronger at higher levels of charisma. Regarding strategic behavior, it was found that higher charisma scores are associated with more strategic behavior and that this impacts positively on leader effectiveness. And this effect is also curvilinear, indicating that the beneficial effects of these higher levels of strategic behavior become weaker at higher levels of charisma. Taken together, although the decline in perceived effectiveness of highly charismatic leaders cannot be due to “stra- tegic overreach” or the tendency to do too much strategic behavior, high strategic levels are associated with a lack of operational behavior, which has a negative impact on the perceived effective- ness. Insufficient operational leadership refers to (a) an inability to attend day-to-day operations, (b) an inadequate focus and level of personal efficiency, and (c) a lack of process discipline to manage an orderly workflow. It seems that highly charismatic leaders overestimate what they can do and underestimate their limits, the risks, and the complex tangle of involvements. These findings align with management research (e.g., Chatterjee & Hambrick, 2011; Malmendier & Tate, 2005) that has related hubris/narcissism to bad business decisions (e.g., paying too much for acquisitions) and to erratic corporate financial performance. The underlying culprit seems to be a lack of self-discipline and insufficient atten- tion for the operational details of business management.

Taken together, our mediation results provide support for theo- retical models of leadership arguing for leader behaviors as mech- anisms through which individual leader traits influence leadership effectiveness (e.g., Antonakis et al., 2012; DeRue et al., 2011; Dinh & Lord, 2012; Zaccaro, 2012). Based on the fact that we found full mediation, it can be suggested that for the curvilinear relation between charismatic personality and leader effectiveness, it’s all in the behavior. From a broader perspective, the results of the current study support and expand the idea of the TMGT effect (Pierce & Aguinis, 2013) as a metatheoretical principle and pro- vide an explanation of how it works for charismatic personality from a cost-and-benefit perspective (Busse et al., 2016).

In terms of practical implications, our research findings may be useful in a leadership-selection context. Specifically, our findings suggest that organizations may want to consider selecting appli- cants with midrange levels of charisma into leadership roles, instead of extremely charismatic leaders. Besides their moderate charisma score, applicants preferably should score high on adjust- ment. Moreover, knowledge of charismatic tendencies could be useful for the purposes of coaching and development. For instance, one strategy could be to confront highly charismatic leaders with the potential gap between their own perception of effectiveness (i.e., being very effective) and the perceptions of their collabora- tors (i.e., being not so effective), along with the most prevalent pitfalls associated with their leadership style. Results of the me- diation analysis are particularly relevant in this regard, demonstrat- ing that highly charismatic leaders would probably gain the most from a coaching program focused on operational deficiencies. On the other hand, the developmental advice for leaders with low charismatic personalities might have a different focus—namely, on increasing strategic behavior. This training program could, for instance, focus on spending more time and energy on long-term planning, taking a broader perspective on the business as a whole, questioning the status quo, and creating a safe environment for trying new things (Kaiser et al., 2010).

Limitations and Strengths

Some limitations of the current work should be acknowledged. First, a single-item measure was used to assess the overall effec- tiveness of leaders (Kaiser et al., 2010), whereas some argue against the use of single-item measures (Pedhazur & Schmelkin, 1991). However, evidence is accumulating that single-item mea- sures can be reliable, certainly when it pertains to constructs that are sufficiently narrow and unambiguous, such as overall job satisfaction and effectiveness (Sackett & Larson, 1990; Wanous & Hudy, 2001). Moreover, we also included other leadership- effectiveness criteria that allowed us to map more specific leader- behavior dimensions. Nevertheless, future research is warranted that replicates our findings with other and multiple-item leadership outcomes.

A second limitation of this study is that no actual situational factors were included as influencers of the relationship between charismatic personality and overall effectiveness. Previous re- search has, for instance, revealed conditions of crisis and perceived uncertainty as relevant moderators in this relationship (e.g., De Hoogh, Den Hartog, & Koopman, 2005; House & Aditya, 1997; Waldman et al., 2001). In line with this stream of research, it could be, for instance, that under conditions of high environmental

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

126 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

uncertainty, the inflection point in the curvilinear association be- tween charisma and effectiveness occurs at higher levels of cha- risma than under conditions of environmental certainty. In fact, higher charisma scores may not always lead to derailment. In certain conditions, such as low-stress situations, the charisma- effectiveness relationship may be linear, rather than curvilinear. However, we believe that high-stress and pressure situations are rather typical for a “normal” leadership context, enhancing the likelihood of a curvilinear relationship.

We do want to point out, however, that we did test the idea that the curvilinear relationship between charisma and effectiveness is subject to boundary conditions. This was done by testing the moderating effect of adjustment, which reflects how one usually deals with stressful situations. Moreover, one of the advantages of the “too little/too much” response format adopted in the current study is that part of this situational variability is automatically taken into account. After all, coworkers rate the behaviors of their leaders as “too little,” “too much,” or “the right amount” given the specific situation that one is evaluated in. In other words, although this approach does not provide concrete information about the specific situational factors that might be influencing this associa- tion, situational variables are implicitly controlled for when using this particular measurement scale, provided that something is “the right amount” in a given situation. Nevertheless, future research should aim to uncover the specific circumstances in which the curvilinear relationship between charismatic personality and effec- tiveness can be obtained.

Beyond these limitations, this study also has a number of nota- ble strengths that bolster its contribution to the extant literature. First, except for the Eugene-Springfield sample, participants were all actual leaders behaving in authentic leadership situations. Moreover, most of the data were collected in the context of large-scale and multinational leader-assessment programs (Study 2 and Study 3) that benefit from a number of methodological strengths: for instance, sufficiently large samples of leaders as- sessed using a multiinformant design in which large samples of coworkers participated. Finally, including multiple leader- effectiveness criteria (both behaviors and overall leader effective- ness) allowed us to delve deep into the exploratory mechanisms underlying the nonlinear charisma-effectiveness association, which can be considered as a robust methodological advancement (Antonakis et al., 2012; Hayes & Preacher, 2010) and is highly relevant for both theory and practice.

Conclusion

Our work tested a personality-based operationalization of cha- risma. In line with the TMGT effect, the picture that emerged from the presented set of studies suggests that leaders with average levels of trait-charisma are perceived as more effective by cowork- ers than those with either low or high charisma levels. However, higher charisma levels are less harmful for leaders having high adjustment levels as well. Our findings further clarified how and why charismatic personality impacts leader effectiveness, as we found that leaders low on charisma are less effective because they lack strategic behaviors, while highly charismatic leaders are less effective because they lack operational behaviors. These findings may stimulate further research on the specific conditions under which charismatic personality is something desirable— or not.

References

Alicke, M. D., & Govorun, O. (2005). The better-than-average effect. In M. D. Alicke, D. Dunning, & J. Krueger (Eds.), The self in social judgment (pp. 85–106). New York, NY: Psychology Press.

Ames, D. R., & Flynn, F. J. (2007). What breaks a leader: The curvilinear relation between assertiveness and leadership. Journal of Personality and Social Psychology, 92, 307–324. http://dx.doi.org/10.1037/0022- 3514.92.2.307

Amirul, S. R., & Daud, H. N. (2012). A study on the relationship between leadership styles and leadership effectiveness in Malaysian GLCs. Eu- ropean Journal of Business and Management, 4, 193–201.

Antonakis, J., Day, D. V., & Schyns, B. (2012). Leadership and individual differences: At the cusp of a renaissance. The Leadership Quarterly, 23, 643– 650. http://dx.doi.org/10.1016/j.leaqua.2012.05.002

Avery, D. R., Tonidandel, S., Griffith, K. H., & Quiñones, M. A. (2003). The impact of multiple measures of leader experience on leader effec- tiveness: New insights for leader selection. Journal of Business Re- search, 56, 673– 679. http://dx.doi.org/10.1016/S0148-2963(01)00312-5

Bass, B. M. (1985). Leadership and performance beyond expectation. New York, NY: Free Press.

Bass, B. M. (1998). Transformational leadership: Industry, military, and educational impact. Mahwah, NJ: Erlbaum.

Bass, B. M., & Avolio, B. J. (1994). Improving organizational effective- ness through transformational leadership. Thousand Oaks, CA: Sage.

Benson, M. J., & Campbell, J. P. (2007). To be, or not to be, linear: An expanded representation of personality and its relationship to leadership performance. International Journal of Selection and Assessment, 15, 232–249. http://dx.doi.org/10.1111/j.1468-2389.2007.00384.x

Blair, C. A., Hoffman, B. J., & Helland, K. R. (2008). Narcissism in organizations: A multisource appraisal reflects different perspectives. Human Performance, 21, 254 –276. http://dx.doi.org/10.1080/089 59280802137705

Bono, J. E., & Judge, T. A. (2004). Personality and transformational and transactional leadership: A meta-analysis. Journal of Applied Psychol- ogy, 89, 901–910. http://dx.doi.org/10.1037/0021-9010.89.5.901

Busse, C., Mahlendorf, M. D., & Bode, C. (2016). The ABC for studying the too-much-of-a-good-thing effect: A competitive mediation frame- work linking antecedents, benefits, and costs. Organizational Research Methods, 19, 131–153. http://dx.doi.org/10.1177/1094428115579699

Chatterjee, A., & Hambrick, D. C. (2011). CEO personality, capability cues, and risk-taking: How narcissists react to their successes and stumbles. Administrative Science Quarterly, 56, 202–237. http://dx.doi .org/10.1177/0001839211427534

Conger, J. A. (1990). The dark side of leadership. Organizational Dynam- ics, 19, 44 –55. http://dx.doi.org/10.1016/0090-2616(90)90070-6

Conger, J. A., & Kanungo, R. N. (1987). Toward a behavioral theory of charismatic leadership in organizational settings. The Academy of Man- agement Review, 12, 637– 647.

Conger, J. A., & Kanungo, R. N. (1998). Charismatic leadership in organizations. Thousand Oaks, CA: Sage.

Conger, J. A., Kanungo, R. N., & Menon, S. T. (2000). Charismatic leadership and follower effects. Journal of Organizational Behavior, 21, 747–767. http://dx.doi.org/10.1002/1099-1379(200011)21:7�747:: AID-JOB46�3.0.CO;2-J

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 Admin- istrative Sciences, 14, 290 –301. http://dx.doi.org/10.1111/j.1936-4490 .1997.tb00136.x

Conway, J. M., & Huffcut, A. I. (1997). Psychometric properties of multisource performance ratings: A meta-analysis of subordinate, super- visor, peer, and self-ratings. Human Performance, 19, 331–360. http:// dx.doi.org/10.1207/s15327043hup1004_2

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

127CHARISMATIC PERSONALITY

Debusscher, J., Hofmans, J., & De Fruyt, F. (2014). The curvilinear relationship between state neuroticism and momentary task perfor- mance. PLoS ONE, 9, e106989. http://dx.doi.org/10.1371/journal.pone .0106989

De Hoogh, A. H. B., Den Hartog, D. N., & Koopman, P. L. (2005). Linking the Big Five-Factors of personality to charismatic and transactional leadership; perceived dynamic work environment as a moderator. Jour- nal of Organizational Behavior, 26, 839 – 865.

Deluga, R. J. (1997). Relationship among American presidential char- ismatic leadership, narcissism, and rated performance. The Leader- ship Quarterly, 8, 49 – 67. http://dx.doi.org/10.1016/S1048- 9843(97)90030-8

Deluga, R. J. (2001). American presidential Machiavellianism: Implica- tions for charismatic leadership and rated performance. The Leadership Quarterly, 12, 339 –363. http://dx.doi.org/10.1016/S1048-9843(01) 00082-0

Den Hartog, D. N., House, R. J., Hanges, P. J., Ruiz-Quintanilla, S. A., Dorfman, P. W., Abdalla, I. A., . . . Zhou, J. (1999). Culture-specific and cross-culturally generalizable implicit leadership theories: Are attributes of charismatic/transformational leadership universally endorsed? The Leadership Quarterly, 10, 219 –256. http://dx.doi.org/10.1016/S1048- 9843(99)00018-1

DeRue, D. S., Nahrgang, J. D., Wellman, N., & Humphrey, S. E. (2011). Trait and behavioral theories of leadership: An integration and meta- analytic test of their relative validity. Personnel Psychology, 64, 7–52. http://dx.doi.org/10.1111/j.1744-6570.2010.01201.x

De Vries, R. E. (2008). What are we measuring? Convergence of leader- ship with interpersonal and non-interpersonal personality. Leadership, 4, 403– 417. http://dx.doi.org/10.1177/1742715008095188

De Vries, R. E. (2012). Personality predictors of leadership styles and the self– other agreement problem. The Leadership Quarterly, 23, 809 – 821. http://dx.doi.org/10.1016/j.leaqua.2012.03.002

DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and domains: 10 aspects of the Big Five. Journal of Personality and Social Psychology, 93, 880 – 896. http://dx.doi.org/10.1037/0022-3514 .93.5.880

Dinh, J. E., & Lord, R. G. (2012). Implications of dispositional and process views of traits for individual difference research in leadership. The Leadership Quarterly, 23, 651– 669. http://dx.doi.org/10.1016/j.leaqua .2012.03.003

Dóci, E., & Hofmans, J. (2015). Task complexity and transformational leadership: The mediating role of leaders’ state core self-evaluations. The Leadership Quarterly, 26, 436 – 447. http://dx.doi.org/10.1016/j .leaqua.2015.02.008

DuBrin, A. (2012). Leadership: Research findings, practice, and skills (7th ed.). Mason, OH: Cengage Learning.

Dvir, T., Eden, D., Avolio, B. J., & Shamir, B. (2002). Impact of trans- formational leadership on follower development and performance: A field experiment. Academy of Management Journal, 45, 735–744. http:// dx.doi.org/10.2307/3069307

Emrich, C. G., Brower, H. H., Feldman, J. M., & Garland, H. (2001). Images in words: Presidential rhetoric, charisma, and greatness. Admin- istrative Science Quarterly, 46, 527–557. http://dx.doi.org/10.2307/ 3094874

Epitropaki, O., & Martin, R. (2004). Implicit leadership theories in applied settings: Factor structure, generalizability, and stability over time. Jour- nal of Applied Psychology, 89, 293–310. http://dx.doi.org/10.1037/0021- 9010.89.2.293

Funder, D. C., & Dobroth, K. M. (1987). Differences between traits: Properties associated with interjudge agreement. Journal of Personality and Social Psychology, 52, 409 – 418. http://dx.doi.org/10.1037/0022- 3514.52.2.409

Furnham, A., Trickey, G., & Hyde, G. (2012). Bright aspects of dark side traits: Dark traits associated with work success. Personality and Indi-

vidual Differences, 52, 908 –913. http://dx.doi.org/10.1016/j.paid.2012 .01.025

Galvin, B. M., Waldman, D. A., & Balthazard, P. (2010). Visionary communication qualities as mediators of the relationship between nar- cissism and attributions of leader charisma. Personnel Psychology, 63, 509 –537. http://dx.doi.org/10.1111/j.1744-6570.2010.01179.x

Gardner, W. L., & Avolio, B. J. (1998). The charismatic leadership relationship: A dramaturgical perspective. The Academy of Management Review, 23, 32–58.

Ghiselli, E. E., Campbell, J. P., & Zedeck, S. (1981). Measurement theory for the behavioral sciences. San Francisco, CA: Freeman & Company.

Goldberg, L. R. (2008). The Eugene-Springfield community sample: Infor- mation available from the research participants (ORI Tech. Rep. No., Vol. 48, number 1). Retrieved from http://scottbarrykaufman.com/wp- content/uploads/2014/06/ESCS-RepMar08.pdf

Grijalva, E., Harms, P. D., Newman, D. A., Gaddis, B. H., & Fraley, R. C. (2015). Narcissism and leadership: A meta-analytic review of linear and nonlinear relationships. Personnel Psychology, 68, 1– 47. http://dx.doi .org/10.1111/peps.12072

Harris, K. J., & Kacmar, K. M. (2006). Too much of a good thing: The curvilinear effect of leader-member exchange on stress. The Journal of Social Psychology, 146, 65– 84. http://dx.doi.org/10.3200/SOCP.146.1 .65-84

Harris, K. J., & Russell, L. M. (2013). An investigation of the curvilinear effects of contingent reward leadership on stress-related and attitudinal outcomes. International Journal of Business and Social Science, 4, 26 –35. Retrieved from http://ijbssnet.com/journals/Vol_4_No_ 10_Special_Issue_August_2013/3.pdf

Hayes, A. F., & Preacher, K. J. (2010). Quantifying and testing indirect effects in simple mediation models when the constituent paths are nonlinear. Multivariate Behavioral Research, 45, 627– 660. http://dx.doi .org/10.1080/00273171.2010.498290

Hoekstra, H. A., Ormel, J., & De Fruyt, F. (2007). Handleiding: NEO persoonlijkheids-vragenlijsten/Manual: NEO personality question- naires. Amsterdam, The Netherlands: Hogrefe.

Hogan, J., Hogan, R., & Kaiser, R. B. (2010). Management derailment. In S. Zedeck (Ed.), American Psychological Association handbook of in- dustrial and organizational psychology (pp. 555–575). Washington, DC: American Psychological Association.

Hogan, R. (2007). Personality and the fate of organizations. Hillsdale, NJ: Erlbaum.

Hogan, R., & Hogan, J. (2007). Hogan Personality Inventory manual. Tulsa, OK: Hogan Press.

Hogan, R., & Hogan, J. (2009). Hogan Development Survey manual (2nd ed.). Tulsa, OK: Hogan Press.

Horney, K. (1950). Neurosis and human growth. New York, NY: Norton. House, R. J., & Aditya, R. M. (1997). The social scientific study of

leadership: Quo vadis? Journal of Management, 23, 409 – 473. http://dx .doi.org/10.1177/014920639702300306

House, R. J., & Howell, J. M. (1992). Personality and charismatic leader- ship. The Leadership Quarterly, 3, 81–108. http://dx.doi.org/10.1016/ 1048-9843(92)90028-E

Howell, J. M. (1988). Two faces of charisma: Socialized and personalized leadership in organizations. Charismatic leadership: The elusive factor in organizational effectiveness (pp. 213–236). San Francisco, CA: Jossey-Bass.

Howell, J. M., & Shamir, B. (2005). The role of followers in the charis- matic leadership process: Relationships and their consequences. The Academy of Management Review, 30, 96 –112. http://dx.doi.org/10 .5465/AMR.2005.15281435

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. http://dx.doi.org/10.1037/0021-9010.69.1.85

John, O. P., & Srivastava, S. (1999). The Big-Five trait taxonomy: History,

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

128 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT

measurement, and theoretical perspectives. In L. A. Pervin & O. P. John (Eds.), Handbook of personality: Theory and research (Vol. 2, pp. 102–138). New York, NY: Guilford Press.

Judge, T. A., & Bono, J. E. (2000). Five-factor model of personality and transformational leadership. Journal of Applied Psychology, 85, 751– 765. http://dx.doi.org/10.1037/0021-9010.85.5.751

Judge, T. A., LePine, J. A., & Rich, B. L. (2006). Loving yourself abundantly: Relationship of the narcissistic personality to self- and other perceptions of workplace deviance, leadership, and task and contextual performance. Journal of Applied Psychology, 91, 762–776. http://dx.doi .org/10.1037/0021-9010.91.4.762

Judge, T. A., Piccolo, R. F., & Kosalka, T. (2009). The bright and dark sides of leader traits: A review and theoretical extension of the leader trait paradigm. The Leadership Quarterly, 20, 855– 875. http://dx.doi .org/10.1016/j.leaqua.2009.09.004

Kaiser, R. B., & Hogan, J. (2011). Personality, leader behavior, and overdoing it. Consulting Psychology Journal: Practice and Research, 63, 219 –242. http://dx.doi.org/10.1037/a0026795

Kaiser, R. B., & Hogan, R. (2007). The dark side of discretion: Leader personality and organizational decline. In R. Hooijberg, J. Hunt, J. Antonakis, & K. Boal (Eds.), Being there even when you are not: Leading through strategy, systems, and structures. Monographs in lead- ership and management (Vol. 4, pp. 173–197). London, UK: Elsevier Science. http://dx.doi.org/10.1016/S1479-3571(07)04009-6

Kaiser, R. B., LeBreton, J. M., & Hogan, J. (2015). The dark side of personality and extreme leader behavior. Applied Psychology: An Inter- national Review, 64, 55–92. http://dx.doi.org/10.1111/apps.12024

Kaiser, R. B., Overfield, D. V., & Kaplan, R. E. (2010). Leadership Versatility Index version 3.0 facilitator’s guide. Greensboro, NC: Kaplan DeVries Inc.

Le, H., OH, I. S., Robbins, S. B., Ilies, R., Holland, E., & Westrick, P. (2011). Too much of a good thing: Curvilinear relationships between personality traits and job performance. Journal of Applied Psychology, 96, 113–133. http://dx.doi.org/10.1037/a0021016

Leary, M. R. (2007). Motivational and emotional aspects of the self. Annual Review of Psychology, 58, 317–344. http://dx.doi.org/10.1146/ annurev.psych.58.110405.085658

LeBreton, J. M., Burgess, J. R. D., Kaiser, R. B., Atchley, E. K. P., & James, L. R. (2003). The restriction of variance hypothesis and interrater reliability and agreement: Are ratings from multiple sources really dissimilar? Organizational Research Methods, 6, 80 –128. http://dx.doi .org/10.1177/1094428102239427

LeBreton, J. M., & Senter, J. L. (2008). Answers to 20 questions about interrater reliability and interrater agreement. Organizational Research Methods, 11, 815– 852. http://dx.doi.org/10.1177/1094428106296642

Lord, R. G., Foti, R. J., & DeVader, C. L. (1984). A test of leadership categorization theory: Internal structure, information processing, and leadership perceptions. Organizational Behavior & Human Perfor- mance, 34, 343–378. http://dx.doi.org/10.1016/0030-5073(84)90043-6

Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. (1996). Effectiveness correlates of transformational and transactional leadership: A meta- analytic review of the MLQ literature. The Leadership Quarterly, 7, 385– 425. http://dx.doi.org/10.1016/S1048-9843(96)90027-2

Maccoby, M. (2004). Narcissistic leaders: The incredible pros, the inevi- table cons. Harvard Business Review, 92–102.

Malmendier, U., & Tate, G. (2005). CEO overconfidence and corporate investment. The Journal of Finance, 60, 2661–2700. http://dx.doi.org/ 10.1111/j.1540-6261.2005.00813.x

McCall, M. W., Jr. (2009). Every strength a weakness and other caveats. In R. B. Kaiser (Ed.), The perils of accentuating the positive. Tulsa, OK: Hogan Press.

McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intraclass correlation coefficients. Psychological Methods, 1, 30 – 46. http://dx.doi.org/10.1037/1082-989X.1.1.30

Mueller, J. S., Goncalo, J., & Kamdar, D. (2011). Recognizing creative leadership: Can creative idea expression negatively relate to perceptions of leadership potential? Journal of Experimental Social Psychology, 47, 494 – 498. http://dx.doi.org/10.1016/j.jesp.2010.11.010

Naragon-Gainey, K., & Watson, D. (2014). Consensually defined facets of personality as prospective predictors of change in depression symptoms. Assessment, 21, 387– 403. http://dx.doi.org/10.1177/1073 191114528030

O’Boyle, E. H., Forsyth, D. R., Banks, G. C., & McDaniel, M. A. (2012). A meta-analysis of the Dark Triad and work behavior: A social exchange perspective. Journal of Applied Psychology, 97, 557–579. http://dx.doi .org/10.1037/a0025679

Oh, I. S., & Berry, C. M. (2009). The five-factor model of personality and managerial performance: Validity gains through the use of 360 degree performance ratings. Journal of Applied Psychology, 94, 1498 –1513. http://dx.doi.org/10.1037/a0017221

Pedhazur, E. J., & Schmelkin, L. P. (1991). Measurement, design, and analysis: An integrated approach. Hillsdale, NJ: Erlbaum.

Pierce, J. R., & Aguinis, H. (2013). The too-much-of-a-good-thing effect in management. Journal of Management, 39, 313–338. http://dx.doi.org/10 .1177/0149206311410060

Pillai, R., & Meindl, J. R. (1998). Context and charisma: A “meso” level examination of the relationship of organic structure, collectiv- ism, and crisis to charismatic leadership. Journal of Management, 24, 643– 671.

Popper, M. (2002). Narcissism and attachment patterns of personalized and socialized charismatic leaders. Journal of Social and Personal Relationships, 19, 797– 809. http://dx.doi.org/10.1177/02654075 02196004

Pozzebon, J. A., Visser, B. A., Ashton, M. C., Lee, K., & Goldberg, L. R. (2010). Psychometric characteristics of a public-domain self-report mea- sure of vocational interests: The Oregon Vocational Interest Scales. Journal of Personality Assessment, 92, 168 –174. http://dx.doi.org/10 .1080/00223890903510431

Quinn, R. E., Spreitzer, G. M., & Hart, S. (1991). Challenging the assump- tions of bipolarity: Interpenetration and managerial effectiveness. In S. Srivatva & R. Fry (Eds.), Executive and organizational continuity (pp. 222–252). San Francisco, CA: Jossey-Bass.

Resick, C. J., Whitman, D. S., Weingarden, S. M., & Hiller, N. J. (2009). The bright-side and the dark-side of CEO personality: Examining core self-evaluations, narcissism, transformational leadership, and strategic influence. Journal of Applied Psychology, 94, 1365–1381. http://dx.doi .org/10.1037/a0016238

Riggio, R. E. (2009). Charisma. In S. J. Lopez (Ed.), Encyclopedia of positive psychology. Oxford, UK: Blackwell.

Sackett, P. R., & Larson, J. R., Jr. (1990). Research strategies and tactics in industrial and organizational psychology. In M. D. Dunnette & L. M. Hough (Eds.), Handbook of industrial and organizational psychology (2nd ed., Vol. 1, pp. 419 – 489). Palo Alto, CA: Consulting Psychologists Press.

Sankowsky, D. (1995). The charismatic leader as narcissist: Understanding the abuse of power. Organizational Dynamics, 23, 57–71. http://dx.doi .org/10.1016/0090-2616(95)90017-9

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. http://dx.doi.org/10.1287/orsc.4.4.577

Shamir, B., & Howell, J. M. (1999). Organizational and contextual influ- ences on the emergence and effectiveness of charismatic leadership. The Leadership Quarterly, 10, 257–283. http://dx.doi.org/10.1016/S1048- 9843(99)00014-4

Tellegen, A. (1985). Structures of mood and personality and their relevance to assessing anxiety, with emphasis on self-reports. In A. H. Tuma &

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

129CHARISMATIC PERSONALITY

J. D. Maser (Eds.), Anxiety and anxiety disorders (pp. 681–706). Hills- dale, NJ: Erlbaum.

Tsui, A. S. (1984). A role-set analysis of managerial reputation. Organi- zational Behavior & Human Performance, 34, 64 –96. http://dx.doi.org/ 10.1016/0030-5073(84)90037-0

VanBroekhoven, J. (2011). Playing the Trump card. Retrieved February 5, 2017 from http://info.hoganassessments.com/blog/bid/109230/Playing- the-Trump-Card

Waldman, D. A., Ramirez, G. G., House, R. J., & Puranam, P. (2001). Does leadership matter? CEO leadership attributes and profitability under condi- tions of perceived environmental uncertainty. Academy of Management Journal, 44, 134 –143. http://dx.doi.org/10.2307/3069341

Wanous, J. P., & Hudy, M. J. (2001). Single-item reliability: A replication and extension. Organizational Research Methods, 4, 361–375. http://dx .doi.org/10.1177/109442810144003

Yukl, G. A. (2006). Leadership in organizations (6th ed.). Englewood Cliffs, NJ: Prentice Hall.

Zaccaro, S. J. (2012). Individual differences and leadership: Contributions to a third tipping point. The Leadership Quarterly, 23, 718 –728. http:// dx.doi.org/10.1016/j.leaqua.2012.05.001

Zibarras, L. D., Port, R. L., & Woods, S. A. (2008). Innovation and the “dark side” of personality: Dysfunctional traits and their relation to self-reported innovative characteristics. The Journal of Creative Behavior, 42, 201–215. http://dx.doi.org/10.1002/j.2162-6057.2008.tb01295.x

Appendix

Justification for Aggregating Observer Ratings

Aggregated observer ratings—including ratings of subordinates, peers, and superiors—were used for overall effectiveness (i.e., in Study 2 and Study 3) and for the leader behaviors (i.e., in Study 3). To provide additional justification for this aggregation method, the rwg(j) interrater agreement coefficient (James et al., 1984) and the one-way random effects intraclass correlation coefficient (ICC; McGraw & Wong, 1996) were computed within superior, peer, and subordinate groups, as well as across these three sources (LeBreton et al., 2003). In the computation of rwg(j) for the overall effectiveness rating, the moderately skewed random response null distribution was used to control for a moderate skew because most ratings were between 6 and 10 on the 1 to 10 scale. To account for central tendency bias, a triangular null distribution was used

in the computation of interrater agreement for the leader-behavior scales (see LeBreton & Senter, 2008). Additionally, intraclass corre- lations (ICC[1]) were computed to evaluate the reliability of an individual rater and ICC(k) to estimate the reliability of the average rating across k raters—where k � 2 for superiors (i.e., the most common number of multiple raters in the superior group); k � 4 for peers; k � 5 (Study 2) and k � 4 (Study 3) for subordinates (i.e., the modal number of raters in these groups); and k � 3 for the aggregate rating across the three sources (i.e., the grand mean of the three rater group means). The results in the Table indicate that, in both studies, the level of similarity across superior, peer, and sub- ordinate ratings is sufficiently high to support aggregation (LeBreton & Senter, 2008).

Inter-Rater Reliability (ICC) and Inter-Rater Agreement (RWg(J)) for Leader Behavior Scales and Observer-Rated Leader-Effectiveness Rating in Study 2 and Study 3

Scale

Superiors Peers Subordinates Aggregated across observer

sources

ICC(1) ICC(k) rwg(j) ICC(1) ICC(k) rwg(j) ICC(1) ICC(k) rwg(j) ICC(1) ICC(k) rwg(j)

Study 2 Leader effectiveness .47 .64 .85 .38 .71 .81 .32 .70 .83 .25 .51 .95

Study 3 Forceful .56 .72 .98 .32 .66 .89 .29 .62 .92 .57 .80 .99 Enabling .22 .37 .98 .29 .62 .93 .25 .57 .94 .64 .84 .99 Strategic .49 .66 .98 .28 .61 .96 .22 .54 .95 .51 .76 .98 Operational .09 .17 .98 .25 .57 .93 .17 .45 .94 .38 .64 .99 Leader effectiveness .21 .35 .87 .29 .62 .73 .24 .55 .71 .57 .80 .93

Note. ICC(k) was based on k � 2 for superior ratings, k � 4 for peer ratings, k � 5 (Study 2) and k � 4 (Study 3) for subordinate ratings, and k � 3 for ratings aggregated across the 3 sources. The rwg(j) values represent the Mean rwg(j) statistic computed across all focal managers (N � 201 for superiors, 306 for peers and subordinates in Study 2; N � 21 for superiors, 287 for peers and subordinates in Study 3).

Received May 18, 2016 Revision received April 5, 2017

Accepted April 5, 2017 �

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

130 VERGAUWE, WILLE, HOFMANS, KAISER, AND DE FRUYT