When it comes to severe wrongdoings in the organizational domain, it is not uncommon that leaders who had been praised as visionary and transformational play a crucial role (Bass 2008). Indeed, whereas transformational leadership has often been regarded as a leadership behavior that considers and emphasizes ethical standards (Burns 1978), several scholars have questioned its inherent morality (Bass and Steidlmeier 1999; Conger and Kanungo 1998; Price 2003). Specifically, they have pointed out that transformational leadership behaviors (e.g., vision, inspiration, and role modeling; Bass 1985) do not have to be applied in the interest of the common good. Indeed, these behaviors seem to be equally effective in pursuing immoral purposes and to increase the personal power and status of the leader. As Price (2003) posited, transformational behaviors are “morally neutral” (p. 70) and whether they are used for good or bad depends on the intentions of the leader.

Building on this observation, Bass and Steidlmeier (1999) elaborated on transformational leadership theory and differentiated two types of transformational leaders: (a) authentic transformational leaders who focus on the common good and (b) pseudo-transformational leaders who seek to enhance their status and foster employees’ dependence (see also Bass 1998; Howell and Avolio 1992). Both types of leaders engage in transformational leadership. However, as Bass and Steidlmeier (1999) noted, it is likely that their contradicting (altruistic versus self-focused) goals will also surface in their leadership behaviors.

Whereas the distinction between morally good and bad transformational leadership has contributed to a more balanced perspective on this leadership style, an essential aspect has been largely overlooked: followers’ reactions to these two types of leaders. Specifically, it has remained unclear whether and how altruistic versus self-focused behaviors of transformational leaders affect their followers’ in-role and extra-role efforts. However, this seems to be a crucial question given the profound influence of transformational leaders on followers’ reactions (Bass 2008) and the fact that the success of organizations strongly depends on their employees’ performance (Podsakoff et al. 2009). Therefore, the purpose of this study was to build and examine a model that details the interactive relation between transformational leadership and altruistic as well as self-focused leadership behaviors. In doing so, we draw on attribution theory, which provides an overarching framework to explain followers’ reactions to the perceived intentions of their leader (Dasborough and Ashkanasy 2002). Specifically, given the often unconventional nature of transformational leadership behaviors, we propose that followers will seek for signals that may reveal the intentions of their leader. Accordingly, we expect that the effects of moral and authoritarian behaviors on followers’ in-role and extra-role efforts will be particularly pronounced for transformational leaders compared to non-transformational leaders.

With this focus, this study contributes to the literature in two ways: First, whereas the ethical and unethical sides of transformational leadership have often been discussed, they have received little attention in empirical research. However, the double-edged nature of transformational leadership seems to be a crucial aspect to understand followers’ responses to their leader and, consequently, for the future development of transformational leadership theory. Second, the study extends existing research by providing an integrative model of different leadership behaviors. As Bass and Steidlmeier (1999) noted, it is unlikely that leaders exclusively engage in one type of leadership behavior (e.g., transformational leadership)—the behavioral repertoires of most leaders go beyond one particular style and include additional behaviors. Surprisingly, few studies have addressed this rather intuitive aspect despite several calls for integrative studies of different leadership styles (Casimir 2001; Kerr et al. 1974; Kirkpatrick and Locke 1996). Indeed, to foster a better understanding of the leadership process, it seems to be essential to consider how different leadership behaviors interact in affecting followers’ reactions (De Cremer 2006).

The Double-Edged Nature of Transformational Leadership

The transformational leadership model has had a tremendous influence on the modern understanding of leadership effectiveness. Since its introduction by Burns (1978), leadership researchers have identified several behaviors that are representative for this leadership style. Perhaps the most characteristic one is the articulation of a compelling vision (Bass 1985; Podsakoff et al. 1990). This behavior refers to leaders who voice an appealing picture for the future of the organization and who inspire their followers by identifying attractive opportunities for the group. In doing so, they show a strong sense of purpose and they enthuse followers by using persuasive and image-based rhetoric (Emrich et al. 2001).

Another behavior that is strongly associated with transformational leadership is charismatic role modeling (Bass 1985; Kark and van Dijk 2007). In that, transformational leaders do not only influence their followers through convincing communication, they also “walk the talk”. These leaders provide a charismatic model of appropriate and expected behavior and, in doing so, they take personal risks to show their conviction (Podsakoff et al. 1990). In addition, transformational leaders encourage their subordinates to collaborate in their work and to jointly strive toward the same goal (Shamir et al. 1993). They create a sense of identification among subordinates, which supports the acceptance of their vision.

Numerous studies have examined transformational leadership behaviors and have indicated their effectiveness in directing followers’ efforts (Judge and Piccolo 2004). However, one of the vexing aspects of these behaviors is that they seem to be open to pursue morally good and bad goals (Price 2003). For example, the skills of communicating a compelling vision and leading by example can be used to motivate followers toward collectively benefitting ends. However, they can also be applied to foster the power and status of the leader. As Carey (1992) aptly noted, “the gifts of charisma, inspiration, consideration, and intellectual strength [can be readily] abused for the self-interest of the leader” (p. 232).

Building on these observations, diverse aspects of the dark side of transformational leadership have been discussed including undesirable personality traits that may be associated with transformational behaviors (e.g., narcissisms, Machiavellianism, psychopathy, and the need for personalized power; Conger and Kanungo 1998; Khoo and Burch 2008), transformational leaders’ values and ability for moral reasoning (Groves and LaRocca 2011; Turner et al. 2002; also see Du et al. 2012), and leaders’ emotions as cues for their motives (Dasborough and Ashkanasy 2002). Perhaps the most influential account has been provided by Bass and Steidlmeier (1999) who differentiated two types of transformational leaders: (a) authentic transformational leaders, who focus on altruistic goals and the common good and whose behavior “is characterized by high moral and ethical standards” (p. 191) and (b) pseudo-transformational leaders who are primarily concerned with their own status and authority and who “seek power and position even at the expense of their followers’ achievements” (p. 187). As the authors emphasized, a fundamental difference between these two types of leaders lies in their intentions guiding their actions. Whereas authentic transformational leaders focus on altruistic goals that promote the common good, pseudo-transformational leaders are guided by self-centered motives.

Transformational Leadership and Followers’ Attributions

Given that transformational behaviors seem to be equivocal with regard to leaders’ intentions, followers will have to focus on additional cues if they strive to understand their leaders’ motives (Dasborough and Ashkanasy 2002). As attribution theory emphasizes, individuals seek out signals and information that reveal the intentions behind behaviors (Allen and Rush 1998). The theory posits that individuals experience a desire to make sense of other peoples’ motives because understanding why someone acts in a certain way reduces uncertainty and allows predicting future events (Wong and Weiner 1981). This seems to be particularly crucial in the leader–follower relationship given that followers depend on their leaders to get access to desired resources (e.g., the assignment of tasks, bonuses, promotion decisions etc.; Pittman and D’Agostino 1985). Importantly, attributions about another person’s behaviors strongly affect how people evaluate and react to this person (Dasborough and Ashkanasy 2002). As past research shows, reactions to presumably altruistic deeds considerably differ from responses to seemingly self-focused actions (Eastman 1994; Grant et al. 2009). Specifically, these studies suggest that one and the same behavior (e.g., proactivity) can elicit positive or negative reactions depending on the motives that seem to underlie this behavior. Similarly, subordinates’ reactions will hugely differ when they assume that their leader assigned a task to foster the good of the collective or when they feel that they are being used for the personal benefit of the leader (Dienesch and Liden 1986).

Although making attributions about other people’s behaviors is a central aspect of social life, it is a core premise of attribution theory that not all social events trigger attributions (Pyszczynski and Greenberg 1981). Specifically, if a social event meets people’s expectations, individuals generally refrain from causal analysis (Hastie 1984). Under these circumstances a search for further explanations would imply a rather inefficient allocation of cognitive resources. Conversely, unexpected and unconventional behaviors tend to trigger sense-making processes and attributional explanations. These are beneficial to adapt predictions about future behaviors and events (Wong and Weiner 1981).

Drawing on this reasoning, it has repeatedly been noted that transformational leadership behaviors are particularly likely to elicit attributional search among subordinates (Bass 2008; Conger and Kanungo 1998; Van Knippenberg and Van Knippenberg 2005). The central reason for this is seen in the often unconventional nature of transformational leadership behaviors (Shamir et al. 1993). For instance, in providing direction for their team, transformational leaders develop an elaborate visionary message instead of focusing on short-term goals (Bass 1985). In doing so, they tend to use metaphoric language and symbols to convey their message to the group whereas non-transformational leaders may largely rely on numbers and facts (Kark and van Dijk 2007). Moreover, transformational leaders often show positive emotions and even enthusiasm in communicating with their employees whereas less charismatic leaders merely use a factual tone in talking to their subordinates (Avolio et al. 2004). In line with these examples, transformational leadership behaviors are often described as extraordinary (Shamir et al. 1993) and, consequently, they are likely to foster employees’ search for signals that may reveal their leaders’ intentions (Berscheid et al. 1976).

In making sense about transformational leaders’ intentions, two kinds of behaviors seem to be particularly informative: moral and authoritarian leadership behaviors. Indeed, following Bass and Steidlmeier’s (1999) analysis, these behaviors seem to be representative for authentic transformational and pseudo-transformational leaders given their focus on altruism and the common good (i.e., moral leadership) and on personalized power and status (i.e., authoritarian leadership), respectively. Following this lead, the following discussion elaborates on the interplay between transformational leadership and these leadership behaviors.

Interactive Effects of Transformational and Moral Leadership

Moral leadership has been defined as an essentially altruistic leadership behavior (Cheng et al. 2000; Pellegrini and Scandura 2009). It comprises behaviors that depict superior personal virtues and selflessness. In that, the primary focus of moral leadership behaviors is the common good and moral leaders use the authority of their position to benefit the collective (Cheng et al. 2004). Moreover, moral leaders tend to actively refrain from using their power for personal gains. For example, they do not use personal relationships, dependencies, and back-door practices. Rather they forgo personal privileges to benefit the group (Cheng et al. 2000).

Such selfless behaviors are regarded as a strong signal for the group-oriented and altruistic intentions of the leader (Van Knippenberg and Van Knippenberg 2005). Indeed, acting in a selfless or even self-sacrificing way is perhaps one of the most direct ways to demonstrate one’s commitment to the collective and its welfare. This is because these behaviors are costly to the individual as they either involve direct negative consequences (e.g., a higher workload) or the abandonment of positive consequences and privileges (e.g., the right to a spacious office). According to attribution theories’ principle of augmentation (Kelley 1973), followers tend to perceive selfless behaviors as a sign for the sincerity of their leaders’ group-oriented intentions. Specifically, these selfless acts contribute to followers’ trust in the leader and their conviction in their leader’s concern for the common good (Van Lange et al. 1997). In line with this view, previous research has found a positive relationship between leaders’ selfless behaviors and followers’ perceptions of leaders’ focus on the collective (Van Knippenberg and Van Knippenberg 2005). Importantly, as the results further revealed, this relation was particularly pronounced if followers had been uncertain about their leaders’ motives.

The extent to which a leader’s actions are perceived as altruistically motivated also tends to have a strong impact on followers’ reactions (De Cremer 2006). Under these circumstances, followers are particularly likely to engage in in-role and extra-role efforts (Van Knippenberg and Van Knippenberg 2005). There are several reasons for this: First, altruistic behaviors tend to create a felt obligation to reciprocate in order to equalize the exchange relationship (Gouldner 1960). Based on the norm for reciprocity, followers are expected to do as is done to them. A failure to do so is likely to harm the relationship with their supervisor (Dienesch and Liden 1986). Second, a leader, in most instances, is the linking pin between the organization and the employees (Levinson 1965). Accordingly, if a leader is perceived to act in a moral and altruistic way, this should positively affect followers’ motivation to engage in in-role and extra-role efforts, which in turn, benefits the organization. Third, it has been argued that engaging in behaviors that are personally costly fosters leaders’ influence on followers (Yorges et al. 1999). These behaviors are perceived as motivated by sincerity and conviction. Consequently, altruistic leadership behaviors should be apt in fostering followers’ efforts for the collective.

Integrating these insights with our earlier discussion, we expect that signaling group-oriented intentions should be particularly effective if a leader also engages in transformational leadership behaviors. Given that transformational behaviors do not reveal leaders’ intentions, followers will search for additional cues to discern their leaders’ motives (Dasborough and Ashkanasy 2002). Even more importantly, due to the often unconventional nature of their behaviors, transformational leaders attract considerable attention (Bass 2008; Conger and Kanungo 1998). These behaviors promote attribution processes to make sense of these actions and subordinates will seek out additional signals that reveal their leaders’ intentions (Berscheid et al. 1976). Accordingly, the influence of moral leadership behaviors on followers’ reactions should be particularly pronounced for transformational leaders. Lowly transformational leaders, on the other hand, tend to be perceived as rather conventional (Bass 2008). They tend to act in line with followers’ expectations and, therefore, should attract fewer efforts to make sense of their leadership behaviors. As a consequence, followers’ responses to moral leadership behaviors of non-transformational leaders should be weaker than for transformational leaders. We propose:

Hypothesis 1

Transformational leadership and moral leadership interact in predicting followers’ in-role and extra-role performance. The more a leader engages in transformational behaviors, the stronger the positive association between moral leadership and followers’ in-role and extra-role performance.

Interactive Effects of Transformational and Authoritarian Leadership

In contrast to moral leadership that builds on the authority of the leadership position to benefit the collective, authoritarian leaders use the power-asymmetry between leaders and followers to assure personal dominance and to centralize control (Tsui et al. 2004). Specifically, authoritarian leadership has been defined as “leader’s behavior that asserts absolute authority and control over subordinates and [that] demands unquestionable obedience from subordinates” (Cheng et al. 2004, p. 91). In that, authoritarian leadership behaviors emphasize unilateral decision-making through the leader and strive to maintain the distance between the leader and his or her followers (Aryee et al. 2007). These behaviors assure the leader’s authority over the direction of the group and grant him or her full control over desired outcomes and resources. As Bass and Steidlmeier’s (1999) suggested, these behaviors are particularly characteristic of pseudo-transformational leaders. Specifically, the authors noted that, “pseudo-transformational leaders will welcome and expect blind obedience. They will attempt to enhance their personal status by maintaining the personal distance between themselves and their followers (p. 189)”.

Absolute power has often been associated with abuse and leaders’ quest for personal benefits (Sivanathan et al. 2008). In line with this view, authoritarian leadership behavior has been related to leaders’ self-centered motives. Specifically, previous research has linked these behaviors to leaders’ need for personalized power (Aryee et al. 2007), Machiavellianism (Kiazad et al. 2010), psychopathy (Khoo and Burch 2008), and narcissism (Conger and Kanungo 1998). Moreover, several scholars have proposed that authoritarian leadership behaviors signal a strong disregard for the interests and perspective of their subordinates (Chan et al. 2012; De Cremer 2006). For example, authoritarian leaders tend to ignore followers’ suggestions and discount their contribution (Aryee et al. 2007). As Chan et al. (2012) noted, subordinates are likely to perceive these authoritarian behaviors as signs of disrespect, devaluation and, accordingly, as indicators for the self-centeredness of their leaders.

Consistent with this view, previous research found that authoritarian leadership was negatively related to followers’ perceptions of leaders’ selflessness (De Cremer et al. 2004). More importantly, this behavior is likely to undermine followers’ efforts for the collective: First, based on social exchange principles, authoritarian leadership behaviors tend to trigger retaliatory responses (Blau 1964). Followers who perceive their leaders as disrespecting and devaluating their interests will try to resist their influence and tend to withhold beneficial work behaviors (Harris et al. 2007). Second, as leaders represent the organization (Levinson 1965), their actions often directly affect followers’ behavior toward the organization. Consequently, if leaders are perceived as belittling their followers’ interests, this is likely to deteriorate followers’ motivation to engage in in-role and extra-role efforts for the organization (Aquino et al. 1999). Third, self-focused behaviors of a leader also seem to reduce his or her influence on followers (Tepper et al. 2001; Yorges et al. 1999). Under these circumstances, subordinates are more likely to focus on their own interests and their investment in group-oriented activities will be reduced. Accordingly, they should be less likely to engage in in-role and extra-role efforts.

Drawing on the attributional reasoning developed above, we expect that the detrimental effects of authoritarian leadership should be particularly pronounced for transformational leaders. Given the unconventional nature of their behaviors, subordinates will draw on additional signals to understand their leaders’ intentions (Berscheid et al. 1976). Conversely, the negative influence of authoritarian behaviors should be less pronounced for less transformational leaders as they attract less attributional search (Bass 2008). We propose:

Hypothesis 2

Transformational leadership and authoritarian leadership interact in predicting followers’ in-role and extra-role performance. The more a leader engages in transformational behaviors, the stronger the negative association between authoritarian leadership and followers’ in-role and extra-role performance.

Method

Procedure and Participants

We tested our hypotheses in a sample of 228 individuals representing 114 subordinate–supervisor dyads. Participants were recruited among employees enrolled in a part-time MBA program at a University in Eastern China. Participants completed the subordinate questionnaire in class. Furthermore, they provided contact information of their supervisors. A survey was sent to each supervisor together with a pre-stamped, self-addressed envelope, and a cover letter. The cover letter stated that one of their subordinates, whose name was printed on the letter, had participated in a management study and had identified the addressee as the referent for performance ratings. Supervisors and subordinates were assured that their data would be kept fully confidential and used for scientific purposes only.

Of 145 subordinates who were invited, 129 participated for a response rate of 89.0 %. Of the 129 supervisors, who were contacted, 114 participated for a response rate of 88.4 %. Fifty-four percent of the subordinates were male; their average age equaled 30.12 years (SD = 3.31). They worked in various organizations throughout China representing a wide range of industries with the most frequent being industrial production (17 %), banking and finance (14 %), and informational technology (13 %). Average work experience equaled 4.85 years (SD = 3.00). Sixty-nine percent of supervisors were male and their average age equaled 38.85 years (SD = 7.34). Chi-square and t tests did indicate no significant differences in demographic variables between subordinates whose supervisors responded and those whose supervisors did not respond, alleviating concerns about systematically missing data.

Measures

All variables were measured with established scales developed in previous studies. Subordinates provided ratings of their leaders’ transformational, moral, and authoritarian leadership behaviors. Subordinates’ in-role and extra-role performance was rated by their supervisors. Unless otherwise noted, items were rated on a 7-point scale ranging from 1 = totally disagree to 7 = totally agree. The original language of the scales on transformational leadership, in-role performance, and extra-role performance is English. To ensure translation equivalence, these scales were translated into Chinese and then back-translated into English by two researchers individually both proficient in Chinese and English. The comparisons between the original and the back-translated scales indicated translation equivalence of both versions.

Transformational Leadership

We measured transformational leadership with Podsakoff et al.’s (1990) twelve-item core-transformational leadership scale, which has widely been used in previous studies (e.g., Bettencourt 2004; Jung et al. 2009). Sample items are: “My supervisor inspires others with his/her plans for the future” and “My supervisor leads by ‘doing,’ rather than simply by ‘telling’” (α = .94).

Moral Leadership

We measured moral leadership with the scale by Cheng et al. (2000). This scale was developed in Chinese and has repeatedly been used in prior research (Pellegrini and Scandura 2009; see Cheng et al. 2004 for the English version). Sample items are: “My supervisor uses his/her authority to seek special privileges for himself/herself (reversed)” and “My supervisor does not take advantage of me for personal gain” (α = .90).

Authoritarian Leadership

To measure authoritarian leadership, we used a scale also developed by Cheng et al. (2000). The original language of the scale is Chinese; it has been repeatedly used in previous studies (e.g., Chan et al. 2012; see Cheng et al. 2004 for the English version). Sample items are: “My supervisor asks me to obey his/her instructions completely” and “My supervisor determines all decisions in the organization whether they are important or not” (α = .81).

In-Role Performance

As in prior research (Kacmar et al. 2009; Tepper et al. 2011), supervisors evaluated followers’ performance using five items adopted from Wayne and Liden (1995). In contrast to the other scales of this study, each of the five items was anchored with an individual description corresponding to the question asked (e.g., “strongly agree” or “outstanding”). Sample items are: “This subordinate is superior to other subordinates that I have supervised before” and “Rate the overall level of performance that you observe for this subordinate” (α = .90).

Extra-Role Performance

Supervisors also rated followers’ extra-role behavior. They answered the nine-item scale by Janssen (2001). The items focus on innovative work behavior, a central aspect of extra-role effort (Podsakoff et al. 2000). The scale has repeatedly been used (Pieterse et al. 2009). Sample items are: “This subordinate mobilizes support for innovative ideas” and “This employee introduces innovative ideas into the work environment in a systematic way” (α = .95).

Control Variables

In line with the previous research (e.g., Schuh et al. 2012), we controlled for supervisors’ and followers’ age and gender because demographic characteristics can be related to leadership behaviors and their perception (e.g., Bass et al. 1996; Eagly et al. 2003).

Results

Means, standard deviations, and correlations for all variables are displayed in Table 1.

Table 1 Descriptive statistics: means, standard deviations, and zero-order correlations
Full size table

Validity Analyses

To ensure that the concepts of our study were empirically distinct, we conducted a series of confirmatory factor analyses (CFAs). Results showed that our measurement model assuming five factors (transformational leadership, moral leadership, authoritarian leadership, in-role performance, and extra-role performance) fit the data well (χ² = 1057.44, df = 619, CFI = .96, NNFI = .96, RMSEA = .08). We compared this model to three plausible alternative models: (a) a two-factor model combining all leadership variables in one and all supervisor-rated outcome variables in a second factor, (b) a three-factor model combining moral and authoritarian leadership in one factor, in-role and extra-role behavior in a second factor while transformational leadership represented the third factor, and (c) a four-factor model combining moral and authoritarian leadership in one factor while all other variables built individual factors. All three alternative models fit the data significantly worse than the measurement model (two-factor model: χ² = 1758.04, df = 628; CFI = .90, NNFI = .90, RMSEA = .13; Δχ² = 700.60, p < .001; three-factor model: χ² = 1452.90, df = 626, CFI = .92, NNFI = .92, RMSEA = .11; Δχ² = 326.4, p < .001; four-factor model: χ² = 1278.87, df = 623, CFI = .94, NNFI = .94, RMSEA = .10; Δχ² = 221.43, p < .001).

Hypotheses Tests

To test our hypotheses, we conducted hierarchical regression analyses. In the first step, we entered the control variables followed by transformational leadership, moral leadership, and authoritarian leadership. We then added the interaction terms. Following Aiken and West (1991), all variables were standardized prior to analysis and the interaction terms were calculated based on the standardized scores.

Table 2 shows the results of our analyses. Of greatest importance and in line with Hypothesis 1, the interaction between moral leadership and transformational leadership predicted followers’ in-role (β = 18, t = 2.60, p < .05) and extra-role efforts (β = .18, t = 2.02, p < .05). To examine the nature of these interactions, we conducted simple slope analyses (Aiken and West 1991) and estimated the relations between moral leadership and in-role and extra-role performance at low (1 SD below the mean) and high (1 SD above the mean) levels of transformational leadership. In line with Hypothesis 1, the relations between leaders’ moral behaviors and subordinates’ in-role and extra-role efforts were positive and significant for employees who experienced their leaders as highly transformational (β = .45, t = 3.26, p < .01 and β = .36, t = 2.10, p < .05, respectively) and not significant for those followers who perceived their leaders as lowly transformational (β = .08, t = .67, p > .50 and β = .01, t = .05, p > .96, respectively). Figure 1a, b shows the interactions based on the simple slopes.

Table 2 Hypothesis tests: results of hierarchical regression analyses
Full size table
Fig. 1
figure 1

a Interaction between transformational leadership and moral leadership in predicting subordinates’ in-role performance. b Interaction between transformational leadership and moral leadership in predicting subordinates’ extra-role performance

Full size image

Providing support for Hypothesis 2, the interaction between authoritarian leadership and transformational leadership predicting followers’ in-role and extra-role efforts was also significant (β = −.23, t = −3.18, p < .01 and β = −.24, t = −2.60, p < .05, respectively). Furthermore, simple slope tests revealed that the relations between leaders’ authoritarian behaviors and subordinates’ in-role and extra-role efforts were significant and negative for employees who experienced their leaders as highly transformational (β = −.38, t = −3.35, p < .01 and β = −.36, t = −2.48, p < .05, respectively) and not significant for those followers who perceived their leaders as lowly transformational (β = .07, t = .65, p > .51 and β = .11, t = .79, p > .42, respectively). The interactions are shown in Fig. 2a, b.

Fig. 2
figure 2

a Interaction between transformational leadership and authoritarian leadership in predicting subordinates’ in-role performance. b Interaction between transformational leadership and authoritarian leadership in predicting subordinates’ extra-role performance

Full size image

Discussion

Building on Bass and Steidlmeier’s (1999) differentiation between authentic transformational leadership and pseudo-transformational leadership, we set out to answer the question whether transformational leadership interacts with moral and authoritarian leadership behaviors in predicting followers’ in-role and extra-role efforts. Building on attribution theory, we hypothesized that the positive and negative effects of these leadership behaviors would be particularly pronounced for highly transformational leaders given that this leadership style elicits strong attention and sense-making efforts among followers. In line with our hypotheses, results revealed that moral and authoritarian leadership behaviors, indeed, had a stronger relationship with employees’ in-role and extra-role efforts for transformational leaders. In contrast, these leadership behaviors did not affect followers’ reactions to non-transformational leaders. These findings add to the literature in several ways:

First, whereas ethical aspects of transformational leadership have attracted considerable attention, very little is known about followers’ reactions to moral and immoral conduct of transformational leaders. This study sought to addresses this gap. Our findings suggest that transformational leadership behaviors can intensify the positive effects of altruistic (i.e., moral) leadership behaviors on followers’ in-role and extra-role efforts. Importantly, however, they also enhanced the adverse consequences of self-centered (i.e., authoritarian) leadership behavior with regards to followers’ reactions. In contrast, no such effects were found for non-transformational leaders. These findings are in line with the notion that followers of transformational leaders strive to make sense of their leaders’ transformational behaviors and, accordingly, are sensitive to additional signals that reveal their leaders’ intentions (Berscheid et al. 1976). Moreover, they shed initial light on the differential effects of authentic versus pseudo-transformational leadership. Specifically, these findings indicate that transformational leadership behaviors do not automatically foster followership but also can diminish followers’ compliance. This seems to be an important finding given that it contradicts the commonly held assumption that transformational leadership is quasi inherently effective. In that, it may prove beneficial for the future development of transformational leadership theory as it may help to delineate the contingencies that determine the effectiveness of transformational leaders.

Second, by building and testing an integrative model of different theories, our study provides a fresh perspective on the effects of leadership. Although it appears to be a rather intuitive insight that most leaders’ behaviors do not exclusively fall into one theoretical category (e.g., transformational leadership) and despite several calls for a theoretical integration of different leadership theories (Casimir 2001; De Cremer 2006; Kerr et al. 1974; Kirkpatrick and Locke 1996), very little is known about the joint effects of different leadership styles. Indeed, the majority of leadership research to date has focused on main effects and on contingencies with context and follower variables (Bass 2008). However, a multidimensional approach to leadership, integrating different leadership models, seems to provide beneficial insights for understanding leaders’ effects on followers. Specifically, our findings suggest that the influence of moral and authoritarian leadership behaviors increases with the degree to which leaders are perceived as being transformational. This important insight has largely been masked by the unidimensional approach to leadership applied in most previous studies.

Third, on a conceptual level, this article may also contribute to a better understanding of the nature of authentic versus pseudo-transformational leadership. Whereas this distinction has been frequently cited, previous research has largely refrained from elaborating on the specific behaviors that characterize these two leadership types. Building on Bass and Steidlmeier’s (1999) work and the wider leadership literature, we developed the argument that it is merely the interplay of transformational leadership behaviors with moral and authoritarian leadership behaviors that lies at the heart of authentic and pseudo-transformational leadership, respectively. This notion rests on and integrates three important insights: (a) the observation that transformational leadership behaviors are morally neutral (Price 2003), (b) the notion that most leaders’ behavioral repertoires exceed one particular leadership style and comprise additional behaviors (Casimir 2001; De Cremer 2006), and (c) the conceptualization of authentic transformational and pseudo-transformational leadership as serving the interests of the collective versus the power and status of the leader (Bass and Steidlmeier 1999). Given that identifying the behaviors of authentic and pseudo-transformational leadership is essential to further understand these two leadership types, we hope that our reasoning will stimulate and contribute to future discussions of the morality of transformational leaders.

Limitations and Future Research

In discussing important limitations of this study, we also want to point out potential directions for future research. The first limitation is the cross-sectional nature of our data. However, although this design, which the present research shares with most studies in the organizational domain, does not allow for a test of causality, we believe that there are robust arguments for the assumption of leadership behavior affecting followers’ reactions. Indeed, although we acknowledge that at times influence in organizations may weave its influence from the bottom to the top, it appears plausible to assume that the predominant direction of influence follows the direction of formal power. This view parallels theoretical accounts (e.g., upper echelon theory; Hambrick and Mason 1984) and has been supported in empirical studies (Jung and Avolio 2000; Kirkpatrick and Locke 1996). Nevertheless, we agree that concerns related to causality should not be lightheartedly discarded and certainly be addressed in future studies.

The second concern may be related to the fact that our research was conducted in Mainland China. Accordingly, it can be questioned whether Western leadership models also apply in other cultures. However, from a theoretical perspective it has been posited that the transformational leadership model transcends cultural borders (Bass 2008) and, in line with this reasoning, previous research has supported the validity of the transformational leadership model across cultures (Chen et al. 2007; Walumbwa et al. 2005). These findings may reflect the fact that in the process of opening to global markets many Chinese organizations have adopted Western management concepts (Lin et al. 2001). Nonetheless, as our study is the first to demonstrate the interactive effects of transformational leadership with moral and authoritarian leadership behaviors, we encourage future research to test these interactions in additional cultural contexts.

Another promising avenue for future research seems to lie in testing whether our theoretical model also applies to additional important employee outcomes. Whereas for an initial test of a model other-rated in-role and extra-role performance are often regarded as the gold standard, the explanatory power of the interactions among transformational, moral, and authoritarian leadership may go beyond these dependent variables. For instance, it appears to be worthwhile to examine whether this model also generalizes to indicators of employees’ ethical conduct (e.g., whistle blowing, organizational deviance) and employee health (e.g., exhaustion, absenteeism).

Future research may also extend the present model by considering leaders’ personality traits. Particularly, the dark side of leaders’ personality has recently emerged as a promising field of research linking leaders’ traits to unfavorable employee outcomes (Gudmundsson and Southey 2011; see also Boddy 2011). Most relevant in this context are the traits of narcissism, Machiavellianism, and psychopathy, termed as the “dark triad of personality”, which relates to callous-manipulative behavior in interpersonal relationships (Jones and Paulhus 2010; Paulhus and Williams 2002; Rauthmann 2012). By integrating these traits with the present model, pseudo-transformational leadership may emerge as an important mediator that links the dark side of leaders’ personality and undesirable employee outcomes.

Finally, future research may also follow the lead of this study and examine the interactive effects of transformational leadership with additional leadership behaviors. In view of the present findings, we believe that examining the contingencies of different leadership behaviors can significantly advance the understanding of leadership effectiveness. We hope that this neglected field of study will attract considerable attention in future research.

Practical Implications and Conclusion

Our findings also bear valuable insights for practice. Specifically, the interaction of transformational leadership with moral and authoritarian leadership suggests that particular caution is required with regards to leadership selection, training, and compensation. Whereas previous research has suggested that organizations should promote transformational leadership (e.g., Liao and Chuang 2007), our findings add yet another consideration: It appears to be equally important to implement practices targeted at enhancing moral leadership behaviors while reducing authoritarian conduct.

Indeed, the pattern of our results suggests that focusing on transformational leadership while ignoring moral and authoritarian aspects may not only undermine efforts of promoting effective leadership but also result in negative consequences. When paired with low moral leadership or high authoritarian leadership subordinates’ showed least in-role and extra-role performance. Conversely, underlining the double-edged nature of this leadership style, transformational leaders were also able to elicit the highest level of followers’ in-role and extra-role efforts—when they were perceived as moral or non-authoritarian behaviors. In light of these findings it appears crucial that organizations, including private and public organizations but also universities and business schools, complement existing efforts in fostering transformational leadership with measures focusing on high morality and low authoritarianism. For instance, training programs discussing the importance of ethics, increasing awareness for moral conduct, and introducing ethical role modeling may prove to be particularly effective (Mayer et al. 2009). Focusing on this and related approaches, organizations are likely to develop and sustain the virtues of (authentic) transformational leadership.