The relationship between leader Machiavellianism, leader-member exchange, and emotional exhaustion : what if followers do not recognize manipulative behavior?

(1)

Faculty of Economics and Business Master Thesis 2017/2018 Leadership and Management

The relationship between leader Machiavellianism, Leader-Member Exchange, and emotional exhaustion: What if followers do not recognize manipulative

behavior?

Author: Jessica (E.A.M.) Tol Student number: 10552006

Date of final submission: 22-07-2018

University of Amsterdam – Amsterdam Business School Supervisor: A.H.B. de Hoogh

(2)

Statement of originality

This document is written by student Jessica Tol (10552006), who declares to take full responsibility for the content of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used creating it.

The Faculty of Economics and Business is responsible solely for the supervision of the completion of the work, not for the contents.

(3)

Abstract

Although it is known that Machiavellianism has several negative consequences, research has failed to examine possible explanations for the relationship between leader Machiavellianism and emotional exhaustion. In addition, there is inconsistency with regard to followers’ recognition of leaders’ manipulative behavior. In this cross-sectional, quantitative study, Leader-Member Exchange is examined as a mediator and perceived manipulative behavior as a moderator on the relationship between leader Machiavellianism and emotional exhaustion. The study included 101 dyads of leaders and followers. However, the results show no direct effect between leader Machiavellianism and emotional exhaustion as well as a relationship between leader Machiavellianism and Leader-Member Exchange. Also, no support for the relationship between Leader-member Exchange and emotional exhaustion was found as well as a mediation effect of Leader-Member Exchange on leader Machiavellianism and emotional exhaustion. Finally, the results do not support perceived manipulative behavior as moderator, and therefore, no moderated mediation was found. However, when applying additional analysis, the results support the relationship between perceived manipulative behavior and emotional exhaustion. Since this study concerns important issues, future research may consider different personality traits, Leader-Member Exchange, and follower and leader similarities as possible influences on the relationship between leader Machiavellianism and emotional exhaustion.

Keywords: Leader Machiavellianism, perceived manipulative behavior, Leader-Member exchange, emotional exhaustion

(4)

Nowadays, there is an increasing interest in the effect of the dark side of personality traits, known as the Dark Triad (Andershed, Gustafson, Kerr, & Stattin, 2002). Individuals who score high on these traits tend to behave in toxic ways with respect to interpersonal situations (Rauthmann, 2012). Part of the Dark Triad is Machiavellianism (Paulhus & Williams, 2002). Machiavellianism concerns a strategy of social conduct, which involves manipulation for one’s benefit, and is often against other’s benefit (Wilson, Near & Miller, 1996). Since leadership is an attempt to influence the behavior of followers by means of social interaction (Yukl, 1989), it may be that leader Machiavellianism has severe negative outcomes for followers. As Machiavellians appear to manipulate, lie, and have little concern for the feelings of others (Christie & Geis, 1970), it is expected that leader Machiavellianism is negatively related to followers’ wellbeing, and may lead to emotional exhaustion (Gkorezis, Petridou, & Krouklidou, 2015). Emotional exhaustion is a chronic state of emotional and physical exhaustion caused by excessive pressures from one’s employment and continuous disturbances (Shirom, 1989).

Nevertheless, little research has been done with regard to possible explanations for the effect of leader Machiavellianism on emotional exhaustion (a notable exception is research by Gkorezis et al. (2015), who examined the mediating effect of cynicism on this relationship). Machiavellian leaders tend to be self-interested, exploitative and show little concern for interpersonal relationships with followers and their feelings. This may result in having poor quality relationships with followers (Gkorezis et al., 2015). A low-quality relationship (LMX) is characterized by a lack of trust, and may stir feelings of followers that result in stress and exhaustion (Thomas & Lankau, 2009). Thus, leader Machiavellianism is expected to

(7)

have severe negative consequences for followers and organizations as it may lead to low-quality relationships with followers and in turn to follower exhaustion.

Although previous research has demonstrated a positive relationship between leader Machiavellianism and follower exhaustion (Gkorezis et al., 2015) and a high-quality leader-follower relationship (LMX) has been linked to reduced follower exhaustion (Thomas & Lankau, 2009; Schermuly & Meyer, 2016), so far research has failed to investigate LMX as a mechanism that may (partly) explain the link between leader Machiavellianism and follower exhaustion. In this study, we examine the relationship of leader Machiavellianism with follower exhaustion and additionally investigate the quality of the relationship between the leader and the follower, so-called Leader-Member Exchange, as a possible mediation mechanism in this relationship.

However, there is inconsistency in research concerning followers’ recognition of leaders’ Machiavellian behavior. Namely, Den Hartog and Belschak (2012) argue that even when a Machiavellian leader demonstrates ethical behavior, followers would still recognize this behavior as inauthentic and thus as manipulative. However, Drory and Gluskinos (1980) suggest that it might be hard for followers to identify this inauthentic behavior due to Machiavellian leaders’ manipulative skills (Hawley, 2003; Kwak & Shim, 2017). Since Machiavellians are excellent liars, followers may have problems and could even fail to perceive their leader as manipulative (Kwak & Shim, 2017). So, when followers do not perceive a Machiavellian’s manipulative behavior, this may not be at the expense of the feelings regarding their leader (LMX) that is characterized by contribution, loyalty, respect, and affection (Maslyn & Uhl-Bien, 2000; Zajonc, 1980). Considering this, leader Machiavellianism may have a less

(8)

strong negative effect on the quality of the relationship between leader and follower, when a leader is perceived as showing less Machiavellian behavior.

Therefore, this study contributes to previous literature by not only extending the literature on the relationship between leader Machiavellianism and emotional exhaustion, and examining the mediating role of Leader-Member Exchange as an explanation, but also assessing the moderating role of perceived manipulative behavior of the (Machiavellian) leader in the relationship between leader Machiavellianism and emotional exhaustion.

In order to examine the above, the study is divided into different sections. First of all, a literature review with detailed theoretical background is provided. This is followed by a method section including sample explanation, measures, and statistical procedure. Then, the results of this study are given, which leads to a discussion of the findings. Together with the discussion, limitations and recommendations for future research are presented. Finally, the conclusion of this research is discussed.

(9)

2. Theoretical Framework and hypotheses development

2.1. Leader Machiavellianism

Nowadays, there is increasing interest in socially averse personalities that are related to empathic deficits (Wai & Tiliopoulos, 2012). Deciding the line between ‘normal’ and ’deviant’ personalities has always been a challenging aspect (Allport, 1937). According to Paulhus and Williams (2002), these deviant characters are known as the Dark Triad, and can be divided in Machiavellianism, narcissism, and psychopathy. The Dark Triad is associated with several undesirable personal and societal outcomes that are considered as negative (Bushman & Baumeister, 1998). Namely, some individuals solve problems in prosocial ways (e.g. cooperation, friendships, and reciprocity), whereas others apply more socially averse strategies (e.g. threatening and selfishness) (Jonason & Webster, 2010). Whereas narcissism and psychopathy are known as personality disorders, the concept of Machiavellianism has a different cause, since it was based on beliefs of Niccoli Machiavelli (Furnham, Richards, & Paulhus, 2013).

Niccoli Machiavelli (1469-1527), a sixteenth century political scientist, developed Machiavellianism as a concept. His most famous work, The Prince, has been widely used by scholars as guidelines as it was based on advice on the way to obtain and maintain power (Wilson et al., 1996). Machiavelli stated that people are self-serving, untrustworthy, vicious, and that one extends its power by exploiting others in a deceitful manner (Wilson et al., 1996).

However, Christie and Geis (1970) were the first psychologists to investigate Machiavellianism. Using The Prince as a foundation, they noticed that high Machiavellian individuals are indeed related to characteristics mentioned by

(10)

Machiavelli. They developed multiple tests to measure individuals’ degree of Machiavellianism in which they referred to ‘’low Machs’’ and ‘’high Machs’’. In their survey, Christie and Geis (1970) translated Machiavelli’s work into three themes. First, high Mach's are considered to have a cynical view toward their surrounding in which it is thought that all other individuals act solely in personal interests. Second, they may voluntary utilize manipulative tactics in order to accomplish desired outcomes. Third, high Machiavellians are willing to neglect ethical requirements if unethical behavior is likely to lead to an advantage over other individuals (Fehr et al., 1992; Dahling kuyumcu, & Librizzi, 2012) Thus, as the latter considers, Mach individuals are not always necessarily behaving unethically.

Thus, according to Christie and Geis (1970), Machiavellianism refers to the dispositional tendency to exhibit manipulative behaviors towards others for one’s own personal gain (Banka & Orlowski, 2012). Given their manipulative behavior, individuals high on Machiavellianism appear to score higher academic grades in school (Burgoon, 1971; Marks & Lindsay, 1966), earn relatively higher incomes (Turner & Martinez, 1977), are selected as leaders (Christie & Geis, 1970), and tend to be more socially appealing than individuals low on Machiavellianism (Cherulnik, Way, Ames, & Hutto, 1981). In addition, it is argued that people high on Machiavellianism tend to have a lack of commitment, are not concerned with morality, are low in interpersonal affect, and they tend to give high priority to acquire money, authority, and competition (Stewart & Stewart, 2006). As Christie and Geis (1970) already found, there is additional evidence for Machiavellianism to have overlap with psychopathy (Fehr et al., 1992; McHoskey, Worzel, & Szyarto, 1998; Cooper, Kline, & May, 1986) as both exist of an individual low emotionality and sensitive to insincere behavior (Fehr et al., 1992).

(11)

The topic has been largely studied in the laboratory. However, existing research argues that Machiavellianism has important consequences for organizational criteria as well (Kessler, Bandelli, Spector, Borman, Nelson, & Penney, 2010). According to Fox and Spector (1999), Machiavellians are more sensitive to engage in counterproductive work behaviors, which are behaviors that are voluntary activities that damage the health of the organization, such as defection, opportunism, and theft. For instance, the association with defection is proven by research by Wilson et al. (1996) in which they state that within a group, Machiavellians hide their true selves, while using their exploitative skills. When their actions become clear, they move to another unsuspecting group. Consequently, this explains why Machiavellians are encouraged to change constantly, which results in higher turnover since they are solely concerned with their own personal advantage (Dahling, Whitaker & Levy, 2009).

Subsequently, high Machiavellians tend to be more willing to steal, even when one is in a trusting relationship. Harrell and Hartnagel (1976) argue that Machiavellians do not make a distinction between trusting and distrustful relationships. Also, they appear to steal larger quantities from the other party and are more likely to hide these thefts than individuals low on Machiavellianism. Furthermore, it is found that high Machiavellians are more likely to take resources, such as knowledge, as they appear to disregard a wide selection of property rights when this is likely to lead to goal attainment (Winter, Stylianou, & Giacalone, 2004). In addition, individuals high on Machiavellianism are more willing to go against the privacy of other individuals. For instance, by accessing personal data without authorization (Dahling et al., 2012).

(12)

Moreover, research argues that people high on Machiavellianism are more likely to engage in so-called influence tactics, such as intimidation and insinuation. Using these tactics, high Machiavellians tend to create the perception of trustworthiness and cooperation as they are suggested to be good liars (Geis & Moon, 1981), which support them in obtaining access to desirable resources (Dingler-Duhon & Brown, 1987). In addition, high Machiavellians are more likely to sabotage in organizations, such as revenue sabotage (e.g., distribute deceptive information regarding revenues), and information sabotage (e.g., circulate upsetting rumors), as they easily justify these acts rather than low Machiavellians (Dahling et al., 2012).

Currently, there is an increasing interest in leaders’ influence on increasing or reducing followers’ well being, and additionally, interest in effect of leader Machiavellianism on followers’ emotions and mood (Bono, Foldes, Vinson, & Muros, 2007; Gooty, Connelly, Griffith, & Gupta, 2010). As Machiavellians are more willing to behave in a deceitful and unethical manner when this leads to personal goal attainment (Christie & Geis, 1970), it may come as no surprise that followers suffer with respect to their wellbeing (Dahling et al., 2012). It is found that follower’s perception of the leader being Machiavellian is negatively associated with follower’s perception of leader credibility, follower motivation, and satisfaction regarding a follower’s job (Teven, McCroskey & Richmond, 2006). A negative correlation has been found between Machiavellianism and empathy toward other people (Barnet & Thompson, 1985), and additionally, high Machiavellians tend to have lower ability to read and feel others emotions that can be seen as having lower emotional intelligence (Simon, Francis, & Lombardo, 1990; Austin, Farrelly, Black, & Moore, 2007). In turn, this lack of empathy and concern may have severe consequences for followers (Cooper & Peterson, 1980). Machiavellian leaders appear to show little interest and

(13)

considerations towards the relationship with their followers as well as their emotions (Drory & Gluskinos, 1980), which may lead to emotional exhaustion (Gkorezis et al., 2015).

2.2. Emotional exhaustion

Emotional exhaustion is defined as chronic feelings of being emotionally drained by demands concerning one’s work (Bakker & Costa, 2014), and is experienced as a loss of concern and feeling, excitement, and trust (Maslach, 1982b). Emotional exhaustion appears to be the first stage leading to burnout (Gaines & Jermier, 1983). Burnout is a particular type of stress that can be seen as a syndrome of emotional exhaustion, depersonalization, and reduced personal accomplishment that can be developed among people that do ‘’people work’’, which develops itself as a cumulative reaction to continuous job stressors (Maslach, 1982; Leiter & Maslach, 2003).

Emotional exhaustion appears to be the key dimension of burnout, and can be seen as the stress dimension and is related to fatigue, anxiety, psychosomatic complaints, and job-related depression (Kahn & Byosiere, 1992). Furthermore, depersonalization represents the interpersonal context dimension of burnout, and refers to negative and insensitive responses to different aspects of an individual’s job, and can be seen as a consequence of emotional exhaustion (Leiter & Maslach, 2003). Burnout’s final dimension is the feeling of reduced personal accomplishment that can be seen as a consequence of burnout’s high negative emotional experience, and refers to an individual’s feeling of incompetence, and a lack of achievement in work (Shirom, 1989; Leiter & Maslach, 2003). Although emotional exhaustion and depersonalization are developed due to work overload and social conflict, the feeling of reduced personal accomplishment emerges from a lack of resources (Leiter &

(14)

Maslach, 2003). However, as mentioned before, research suggests that emotional exhaustion is the most important dimension of burnout, as it shows more consistent effects than the other two dimensions (Kahn & Byosiere, 1992).

An explanation for the development of emotional exhaustion is the Job Demands-Resources (JD-R) framework by Demerouti, Bakker, Nachreiner, and Schaufeli (2001), in which they divide two different facets of work, namely job demands and job resources (Bakker, Demerouti, de Boer, & Schaufeli, 2003). Job demands refer to facets of work that involve essential effort of an individual, such as cognitive, physical and emotional effort, which are a necessity for a particular job or for the organization in which one is employed (e.g. time pressure, workload, emotional pressures, personal conflicts) (Demerouti et al., 2001; Schaufeli & Bakker, 2004). However, job demands are not necessarily perceived as negative, but when individuals cannot meet the demands that impose high effort, job demands are more likely to lead to job strain (exhaustion) (Schaufeli & Bakker, 2004). Job resources (e.g. feedback, support of colleagues, leader coaching), on the other hand, are facets of work that are able to diminish the effects of job demands, assist the accomplishment of work-related goals, and facilitate possibilities for personal development (Demerouti et al., 2001; Schaufeli & Bakker, 2004). Research shows evidence for the accumulation of high job demands and low job resources as predicting emotional exhaustion, and eventually burnout (Karasek, 1979; Schnall, Landsbergis, & Baker, 1994).

Moreover, when emotional exhaustion leads to burnout, this has severe consequences for employees and people they interact with, such as customers and coworkers, but also for the organization in which they are employed (Maslach & Jackson, 1981). Namely, it can be costly for organizations as it is related to job

(15)

dissatisfaction, turnover, absenteeism, low job performance, and decreasing commitment (Schaufeli & Bakker, 2004; Maslach & Jackson, 1981; Halbesleben & Bowler, 2007). In addition, when a follower leaves as a consequence of burnout, this can cause distress among remaining followers since tasks are disrupted, or among new and inexperienced followers who are filling the vacancies (Maslach, Schaufeli, & Leiter, 2001). Moreover, this can lead to increased costs and even to decreasing trust of clients in the organization, as tasks are not optimally accomplished (Kompier & Cooper, 1999). Besides negative effects for the organization, burnout has severe consequences for individuals as well, such as insomnia, depression, headaches, increased use of drugs and alcohol, and fatigue (Maslach & Jackson, 1981).

Research on emotional exhaustion is important since itself, as it is the key component of burnout, and it has negative implications for organizational functioning as well as the quality of work life (Cordes & Dougherty, 1993). For an employee, emotional exhaustion is often related to colds, headaches, somatic problems, and sleeping disorders (Belcastro & Hays, 1984). In turn, emotional exhaustion is found to correlate with the desire to leave an organization (Jackson, Schwab & Schuler, 1986), certain work attitudes (Leiter & Maslach, 1988), counterproductive work behavior, low organizational commitment, and reduced job performance (Wright & Bonett, 1997; Maslach, Jackson, & Leiter, 1986). Given these negative consequences and the fact that emotional exhaustion appears to be the first stage of burnout, it seems that managing these negative feelings may be in the interest of both parties

Thus, since leadership contains interpersonal interactions between leader and follower (Wu & Hu, 2009) and the fact that Machiavellianism is associated with being selfish, and having little empathy and concern towards followers’ feelings and

(16)

emotions, it is expected that leader Machiavellianism is positively related to followers’ emotional exhaustion. Therefore, the following hypothesis is tested:

H1: Leader Machiavellianism is positively related to follower emotional exhaustion

2.3. The mediating role of Leader-member exchange

In research, there is an increased interest in organizational exchange processes, such as leader-member exchange (Rousseau & Parks, 1993). Leader-member exchange (LMX) theory, consists of four different facts, namely contribution, loyalty, respect, and affect (Dienesch & Liden, 1986). The theory states that differentiating social relationships are developed between leader and followers who directly report to this leader (Liden, Erdogan, Wayne, & Sparrowe, 2006). Namely, with some followers, the leader develops a lower quality exchange relationship (LMX), in which the interaction is strictly contractual (Liden & Graen, 1980). High quality of a leader-member relationship symbolizes high levels of mutual trust, responsibility, and respect (Graen & Uhl-bien, 1995). In addition, when having a high-quality exchange relationship, the leader may provide coaching, support of followers’ social networks, and empowerment, in return for followers’ engagement in organizational citizenship behaviors (OCB) (Settoon, Bennett, & Liden, 1996) and high levels of task performance (Wayne, Shore, & Liden, 1997).

According to role theory, a central theory of LMX, the creation of quality regarding the leader-member relation moves along different stages in which the leader and member test each other (Graen & Scandura, 1987). Under the role-taking phase, explicit role expectations for the member are communicated by the leader, upon which the member reacts. Subsequently, the leader judges this reaction. This cycle

(17)

forms one role episode. Next is the making phase, which occurs once the role-taking phase proceeds in a social exchange relation (Blau, 1964). Ultimately, regardless whether this relation is of high quality, when a certain pattern is developed, the last phase, role routinization, emerges (Graen & Scandura, 1987).

As mentioned before, leaders high on Machiavellianism are characterized by a high degree of manipulative behavior and use others for their own personal gain (Christie & Geis, 1970). Schyns (2015) states that eventually, followers will notice that the leader manipulates. In addition, Machiavellians have little empathy and show little support towards others (Christie & Geis, 1970). In turn, this leads to a low-quality relationship between leader and follower (Schyns, 2015).

H2: Leader Machiavellianism is negatively related to Leader-Member Exchange

Also, research confirmed the belief that work-related social relationships influence follower health (Viswesvaran, Sanchez & Fisher, 1999). Namely, it is shown that followers who have high-quality relationships with their leader appear to acquire better job and organizational-related information, greater job direction, show advanced ability, and receive higher ratings (Gerstner & Day, 1997). These findings are in line with the Job demands-resources model, which proposes that job strain (emotional exhaustion) is a consequence of disruptions in the equilibrium involving demands employees experience and the resources employees can utilize regarding their work (Bakker & Demerouti, 2011). Thus, a high-quality relationship between leader and follower leads to possessing greater resources, which in turn decreases the likelihood of emotional exhaustion (Thomas & Lankau, 2009).

(18)

Besides obtaining greater resources, a high-quality relationship between leader and follower, appears to have other advantages. High LMX is characterized by appropriate communication that is suggested to diminish feelings of ambiguity and uncertainty (Harris & Kacmar, 2006). On the contrary, a low-quality relationship between leader and follower ensures minimal or no support may aggravate emotional exhaustion (Thomas and Lankau, 2009). Finally, it is found that employees compare themselves to others with respect to the division of LMX. Therefore, a low-quality relationship between leader and employee is likely to develop a secondary role stressor, which is perceived as a demand rather than a resource (Erdogan & Liden, 2002).

Considering the above-mentioned arguments, it is expected that LMX is negatively related to emotional exhaustion. Therefore, the following hypothesis is tested:

H3: Leader-Member Exchange is negatively related to follower’s emotional exhaustion

The above taken together, leader Machiavellianism may lead to a low-quality relationship between leader and follower, and LMX is likely to lead to emotional exhaustion as the follower experiences reduced resources and high demands. Therefore, it is expected that LMX explains the relation between leader Machiavellianism and emotional exhaustion as a mediating mechanism. The following Hypothesis is tested:

(19)

H4: Leader-Member Exchange mediates the relation positive between leader Machiavellianism and follower’s emotional exhaustion

2.4. The moderating role of follower’s perceived manipulative behavior

As mentioned before, individuals high on Machiavellianism tend to behave in a manipulative and deceitful manner (Exline, Thibaut, Hickey, Gumpert, & Geis, 1970). Nevertheless, there is inconsistency in research about employees’ recognition of leaders’ Machiavellian behavior.

On the one hand, Den Hartog and Belschak (2012) mention in their study that when a leader shows Machiavellian behavior, a follower is still able to recognize these manipulative behaviors as illegitimate and inauthentic. Consequently, although Machiavellian leaders try to behave in an ethical and correct manner, a leader would fail to manipulate others for his/her own benefit and personal initiatives. This inauthenticity is then perceived as a fake expression of emotions (Hochschild, 1987), which results in their behavior being less powerful (Zapf, 2002).

On the contrary, Kwak and Shim (2017) argue that because Machiavellians are excellent liars, followers might fail to detect their manipulative behavior and perceive their behavior as sincere. Their research is supported by Forsyth, Banks, and McDaniel (2012), who argue that Machiavellians have excellent social skills. In turn, by using these social skills, their Machiavellian behaviors may be perceived as positive by their followers. Also, Deluga (2001) states that since high Machiavellians have better abilities to cover their actual behavior, their behavior is often confused with charisma. As a result of concealing their true intentions, Machiavellians may establish great social networks, receive trust and respect of followers, and generate desirable outcomes for customers, which result in higher job performance (Kessler,

(20)

Bandelli, Spector, Borman, Nelson, & Penney, 2010). However, the thought behind these behaviors is of self-interest by the leader by which they subtly turn the situation in their favor by using manipulation (Kwak & Shim, 2017).

The latter view is supported by Thomas and Pondy (1977), who argue that individuals analyze others’ behavior through their own perceptions of others’ intentions. However, in order to understand actual behavior, the underlying intentions of the behavior must first be determined (Fishbein & Ajzen, 1975; Dienesh & Liden, 1986). Individuals have the desire to find clarification concerning others behavior, which in turn leads to the feeling having control with regard to environmental elements (Ferris, Bhawak, Fedor, & Judge, 1995) Applying this to LMX, Ferris et al. (1995) note that personality and social skills of leaders affect their ability to positively control attributions of being authentic and honest. Machiavellians are advanced in hiding their true intentions and have the ability to control others’ attributions (Ferris et al., 1995), which may result in Machiavellianism being often confused with charisma (Deluga, 2001).

Thus, when a Machiavellians’ actual intentions are not recognized by followers, they may not experience the negative consequences of leader Machiavellianism. In turn, this may not lead to a negative impact on the quality of the relationship between leader and follower (Dienesch and Liden, 1986). Building on the mediation model as explained above, it is predicted that the extent to which a follower perceives its leader to behave in a manipulative manner moderates the relation between leader Machiavellianism and LMX. Therefore, the following hypothesis is tested:

(21)

H5: The negative relation between leader Machiavellianism and Leader-Member Exchange is moderated by follower’s perceived manipulative behavior, such that this relation is stronger when the perceived manipulative behavior is high

Taken together, it is expected that:

H6: leader Machiavellianism is related to follower exhaustion via conditional indirect effects, such that the positive effect via the negative relationship with LMX is stronger when perceived manipulative behavior is high

3. Method

3.1. Sample

To gather data, cross-sectional survey design (Appendix A and B) is used. This research consists of data collected by seven students of the Master of Business Administration at the University of Amsterdam. The sample consists of leaders and followers who are required to work a minimum of 32 hours per week. These participants were recruited via personal or work relationships of each of our group members, and therefore, it can be seen as convenience sampling, which was voluntary of nature. Participants gave their answers by means of a Qualitrics survey

(www.qualtrics.com), and since it only involved Dutch participants, the survey was

developed in Dutch. Additionally, anonymity was guaranteed, however, each couple (leader and follower) received a unique code by which they were identified and connected. The participating leader and follower each received a different version of the survey (Appendix A and B) in which they were asked questions regarding

(22)

themselves and the other party. Furthermore, the data collection process was initiated from the beginning of April until mid-May 2018.

Overall, 155 dyads received an invitation to participate in this research. A total of 114 leaders (47,70%=male, 52,30%=female) and 123 followers (43,70%=male, 56,30%=female) actually completed the survey, which leads to a response rate of 76,5%. However, to have usable dyads, both the leader and the follower had to submit the survey. Therefore, after reducing these incomplete dyads, 101 dyads (65,20%) remained and could be used in the analysis process. Of these complete dyads, followers and leaders answered the demographic question regarding their age, resulting in a range from 18 to 61 (M=33,57, SD=12,83) for followers and a range from 21 to 62 (M=47,59, SD=10,63) for leaders. Finally, of the 123 followers, 115 filled in the branch in which he or she is employed. The three largest branches concern education (21,85%), healthcare (14,95%), and Fast Moving Consumers Goods (13,80%).

3.2. Measures

Leader Machiavellianism (∝= 0,83) is measured by using a short ten items version of the Mach-IV scale developed by Christie and Geis (1970). However, item 9 had to be recoded, and item 7 had to be deleted. Questions regarding Machiavellianism are ‘’the best way to handle people is to tell them what they want to hear’’ and ‘’never tell anyone the real reason you did something unless it is useful to do so’’, and are answered on a seven-point Likert scale (1=strongly disagree, 7=strongly agree). Emotional exhaustion (∝= 0,86) is measured by six items of the emotional exhaustion dimension of Maslach’s Burnout Inventory (Maslach & Jackson, 1986). Example questions are ‘’I feel tired when I get up in the morning and know I have got

(23)

another working day ahead of me’’ and ‘’I feel mentally exhausted by my work’’, which are answered on a seven point Likert scale (1=strongly disagree, 7=strongly agree).

Leader-member Exchange (LMX) (∝= 0,90) consists of eleven items and is based on research by Liden and Maslyn (1998). Questions regarding LMX are ‘’my supervisor defends my work actions to a superior, even without complete knowledge of the issue in question’’ and ‘’my supervisor would defend me to others in the organization if I made an honest mistake’’. Furthermore, a seven-point Likert scale (1=strongly disagree, 7=strongly agree) is used.

Follower’s perceived manipulative behavior (PMB)(∝= 0,86), is based on research by Dasborough and Ashkanasy (2004). The PMB scale consists of four items and is answered based on a seven-point Likert scale ranged from 1 (=strongly disagree) to 7 (=strongly agree). An example of a question regarding follower’s perceived manipulative behavior is ‘’my leader manipulates his/her subordinates’’. Control variables consisted of demographic questions, such as followers’ gender and leader tenure. According to Posig and Kickul (2004), there may be differences in gender regarding emotional exhaustion. In addition, leader tenure is controlled for as the intensity of the dyadic relation is likely to influence LMX (Sin, Nahrgang & Morgeson, 2009).

3.3. Statistical procedure

To test the given Hypotheses of this research, some adjustments have been made to the raw data. Firstly, the data set of the leader and the follower had to be merged into one SPSS file, whereby they were connected using a unique code. After that, missing values had to be deleted using discrete missing values (999). Additionally, the data

(24)

was checked on normality and collinearity, which resulted in the exclusion of 28 outliers to guarantee the quality, based on the z-score (>3).

Subsequently, the reliability of the used scales was calculated. Only from the Mach-IV scale, item 9 and 7 had to be recoded, since these items had a negative Crohnbach’s Alpha, using recode into different variables. Additionally, it was calculated whether Crohnbach’s Alpha could be increased when deleting an item of that particular scale. This resulted in one deleted item (7) from the Mach-IV scale, which increased its Crohnbach’s Alpha from 0.797 to 0.829 respectively. After calculating reliability, the mean for each variable was computed. Finally, to test the hypothesis, regression analysis is used by means of Process (Hayes, 2013), since the model contains both moderation (PMB) and mediation (LMX).

4. Results

In the following section, the results are discussed concerning the predicted hypotheses. A distinction has been made between descriptive statistics, linear regression analysis to test the first hypothesis, and regression analysis by Process 2.0 model 4 and 7 to test Hypotheses 2 to 6.

(25)

4.1. Descriptive statistics

Table 1

Means, Standard Deviation, Correlations among variables, and Reliabilities

Note: N=86, *Correlation is significant at the 0,05 level (2-tailed), **Correlation is significant at the 0,01 level (2-tailed). LMX = Leader-Member Exchange, PMB = Perceived Manipulative Behavior

Table 1 shows the means, standard deviations and correlation scores of the variables of this research. Additionally, the Cronbach’s Alpha’s of the main variables are displayed and are considered as highly reliable as they each have a Cronbach’s Alpha > 0.8, which indicates great internal consistency. Furthermore, gender, age, and tenure of both leader and follower are included to explore whether these correlate with the models’ main variables.

Interestingly, it is found that leader Machiavellianism is negatively correlated with age of the follower (r=-0,36; p=0,01), meaning that the younger the follower, the higher the leader scores on Machiavellianism. Subsequently, there is a negative correlation between leader Machiavellianism and leader tenure in months (r=-0,29; p=0,01), which indicates that the longer the leader works at a certain organization, he or she is less likely to be Machiavellian. In addition, LMX is negatively correlated with PMB (r=-0,25; p=0,05), meaning that when a follower perceives its leader to

M SD 1 2 3 4 5 6 7 8 9 10 1. GenderLeader 1,45 0,50 2. Gender Follower 1,56 0,50 0,34** 3. Age Leader 47,59 10,63 0.08 -0,12 4. Age follower 33,57 12,83 -0,08 -0,09 0,08 5. Tenure months Leader 147,95 129,89 -0,12 -0,10 0,24* 0,44** 6. Tenure months Follower 78,73 110,97 -0,05 -0,14 0,08 0,72** 0,60** 7. Leader Mach 2,71 0,83 -0,06 -0,12 -0,14 -0,36** -0,29** -0,33** (0.83) 8. PMB 1,74 0,93 -0,02 -0,14 -0,65 0,16 0,11 0,02 -0,04 (0.89) 9. LMX 5,64 0,60 0,06 0,09 0,07 -0,08 -0,08 -0,08 -0,07 -0,25* (0.90) 10. Emotional Exhaustion 2,82 0,19 -0,15 -0,28** 0.05 0.15 0,18 0,13 0,09 0,29** 0,09 (0,86)

(26)

show Machiavellian behavior, the quality of the relationship between leader and follower is rated lower. Also, emotional exhaustion is negatively correlated with follower gender (r=-0,28; p=0,01), which surprisingly represents that men in this sample are more likely to feel emotionally exhausted. Finally, emotional exhaustion is positively correlated with PMB (r=0,29; p=0,01). This signifies that when a follower perceives its leader to show Machiavellian behavior, emotional exhaustion is more likely to occur.

4.2. Regression analysis

4.2.1. Direct effect

The first Hypothesis predicts that leader Machiavellianism is positively related to follower’s emotional exhaustion. To test this, a linear regression using SPSS, has been conducted of which the results are displayed in Table 2. This analysis has been divided in two steps to test the direct effect of leader Machiavellianism on follower’s emotional exhaustion. In the first step, two control variables were added, namely leader tenure and follower gender. This was done to see whether the observed effect of emotional exhaustion is independent of the effect of leader tenure in months and follower gender. In the second step, leader Machiavellianism was added as an independent variable to investigate its effect on emotional exhaustion.

Although the linear regression shows no significant effect of leader tenure (B=0,00; p>0,05), follower gender appears to be significantly related to emotional exhaustion (B=-0,50; p<0,05). When incorporating leader Machiavellianism, it shows no improvement in the ability to predict emotional exhaustion F(1,91)=3,56; p>0,05. Indeed, leader Machiavellianism is not found to be significantly related to emotional

(27)

exhaustion (B=0,31; p>0,05). Thus, Hypothesis 1 is rejected, since we found no relationship between leader Machiavellianism and emotional exhaustion.

Table 2

Linear regression analysis of the direct effect

R R2 R2 Change B SE β T Model Step 1 0,32 0,10 0,09 Leader tenure 0,00 0,00 0,19 1,91 Follower gender -0,50 0,20 -0,24* -2,44 Step 2 0,32 0,11 0,08 Leader tenure 0,00 0,00 0,19 1,92 Follower gender -4,48 0,21 -0,24* -2,30 Leader Mach 0,31 0,12 0,03 -0,26

** Significant at the level of 0,01 (2-tailed) * Significant at the level of 0,05 (2-tailed)

4.2.1. The mediating effect of Leader-Member Exchange

Next, a mediation analysis is discussed by using simple mediation analysis model 4 of Process by Andrew Hayes (2013) to test Hypotheses 2 to 4. First, it is tested whether leader Machiavellianism is negatively related to LMX, and whether LMX is negatively related to emotional exhaustion. Thereafter, it is examined whether LMX acts as a mediation mechanism on the relationship between leader Machiavellianism and emotional exhaustion. The results of the analysis are displayed in Table 3, in which it is additionally controlled for leader tenure and follower gender.

Hypothesis 2 predicts a negative relationship between leader Machiavellianism and LMX. As can be seen in Table 3, this relationship is insignificant (B=0,04; p>0,05). When considering control variables, it can be seen that leader tenure in months (B=0,00; p>0,05) as well as follower gender (B=-0,01; p>0,05) has no effect on LMX. In addition, the model does is not found to be

(28)

significant (F(3,91)=0,32, p>0,05). Thus, it cannot be proven that leader Machiavellianism is related to the quality of the relationship between leader and follower (LMX), meaning that Hypothesis 2 is rejected.

Subsequently, it is predicted that LMX is negatively related to emotional exhaustion. As displayed in Table 3, an insignificant relationship has been found between LMX and emotional exhaustion (B=-0,03; p>0,05). When considering control variables, it appears that follower gender (B=-0,48; p<0,05) affects emotional exhaustion. Although, it is found that the model explains 11% of the variance in emotional exhaustion (F(4,90)=2,65; p<0,05), Hypothesis 3 is rejected since it cannot be proven whether a leader who is Machiavellian leads to a low-quality relationship between leader and follower.

Table 3

Regression analysis simple mediation (model 4)

Finally, Hypothesis 4 predicted LMX to act as a mediating mechanism on the relationship between leader Machiavellianism and emotional exhaustion. However, as can be seen in Table 4, a bias-corrected bootstrap 95% confidence interval for the

Consequent

LMX (M) Emotional exhaustion (Y)

Antecedent Coeff. SE p Coeff. SE p

Leader

Machiavallianism (X) 0,04 0,08 0,60 0,03 0,12 0,79

LMX(M) - - - -0,03 0,16 0,86

constant 5,56 0,36 0,00 3,52 1,05 0,00

Leader tenure in months 0,00 0,00 0,47 0,00 0,00 0,06

Follower gender -0,01 0,14 0,96 -0,48 0,21 0,02 R2 = 0,01 R2 = 0,11 F (3,91) = 0,32, p>0,05 F (4,90) = 2,65 , p<0,05 Note: N=95

(29)

indirect effect (B=0,00) appears to contain zero (LLCI =-0,05; ULCI=0,02). Therefore, it cannot be proven that LMX is an explanation for the relationship between leader Machiavellianism and emotional exhaustion, which results in the rejection of Hypothesis 4.

Table 4

Confidence interval: direct, indirect, and total effect (model 4)

Effect SE p LLCI ULCI

Direct effect 0,03 0,12 0,79 -0,20 0,27

Total effect 0,03 0,12 0,79 -0,20 0,26

Boot SE Boot LLCI Boot ULCI

Indirect effect 0,00 0,02 -0,05 0,02

4.2.1 The moderating effect of follower’s perceived manipulative behavior

Besides the mediating effect as examined above, it is expected that PMB has a moderating effect on the relation between leader Machiavellianism and follower’s emotional exhaustion, so that this relation is stronger when PMB is high. As can be seen in Table 5, no significant evidence of moderation has been found (B=0,05; p>0,05), meaning that it cannot be proven whether the relationship between leader Machiavellianism and LMX is affected by different levels of the perception of a Machiavellian leader’s behavior. Therefore, Hypothesis 5 is rejected. Also, considering control variables, no significant effect is found for leader tenure (B=0,00; p>0,05) and follower gender (B=-0,05; p>0,05).

(30)

Table 5

Regression analysis moderated mediation (model 7)

Consequent

LMX (M) Emotional exhaustion (Y)

Leader Machiavallianism (X) 0,02 0,08 0,76 0,03 0,12 0,79 LMX(M) - - - -0,02 0,16 0,86 PMB (W) -0,17 0,07 0,015 - - - Leader Mach x PMB (XW) 0,05 0,08 0,58 - - - constant 5,72 0,24 0,00 3,61 0,98 0,00

Leader tenure in months 0,00 0,00 0,62 0,00 0,00 0,06

Follower gender -0,05 0,14 0,72 -0,48 0,21 0,02 R2 = 0,08 R2 = 0,11 F (5,89) = 1,60, p>0,05 F (4,90) = 2,65 , p<0,05 Note: N=95

LMX = Leader-Member Exchange, PMB = Perceived Manipulative Behavior

Finally, the moderated mediation as formalized in Hypothesis 6, predicts that leader Machiavellianism is related to follower exhaustion, via conditional indirect effects, such that the positive effect via the negative relationship with LMX is stronger when perceived manipulative behavior is high. As can be seen in Table 6, a closer inspection of the findings with regard to a 95% confidence interval, supports the insignificance of the relationship as explored at different levels of PMB appear to be insignificant. Therefore, hypothesis 6 is rejected.

Table 6

Bootstrapping results for the test of conditional indirect effects at specific values of the moderator (PMB) Mediator PMB Conditional indirect effect SE 95% CI LLCI ULCI LMX -0,80 0,00 0,02 -0,04 0,04 0,00 0,00 0,01 -0,04 0,02 0,95 0,00 0,02 -0,07 0,03

Note: Results are based on 5000 bootstrap samples, CI=confidence interval

*p<0,05, **p<0,01

(31)

4.3 Additional analysis

Additionally, it might be interesting to investigate the same model, but with LMX as a moderating effect, and PMB as a mediating effect. This thinking supports research by Den Hartog and Belschak (2012) suggesting that followers are able to recognize Machiavellian behavior that is showed by their leader. This in turn leads to negative outcomes, such as emotional exhaustion (Dahling et al., 2012). In addition, LMX may act as a moderator on the relationship between leader Machiavellianism and PMB. Namely, when a Machiavellian leader senses an advantage in having a high-quality relationship with a follower, LMX is likely to be high. However, they will cut this relationship when the follower is no longer needed (Schyns, 2015). Therefore, the following hypothesis is tested:

H7: leader Machiavellianism is positively related to follower’s perceived manipulative behavior

H8: follower’s Perceived Manipulative Behavior is positively related to follower’s emotional exhaustion

H9: The relation between leader Machiavellianism and follower’s emotional exhaustion is mediated by follower’s Perceived Manipulative Behavior

H10: The positive relation between leader Machiavellianism and follower’s Perceived Manipulative Behavior is moderated by Leader-Member Exchange, so that this relation is weaker when Leader-Member Exchange is high

(32)

H11: Leader Machiavellianism is related to follower exhaustion via conditional indirect effects, such that positive relationship with Perceived Manipulative Behavior is weaker when Leader-Member Exchange is high.

Figure 2: Research model 2

4.3.1. The mediating effect of Perceived Manipulative Behavior

To examine the mediation, model 4 of Process by Andrew Hayes (2013) is used to test Hypotheses 7 to 9. First, it is examined whether leader Machiavellianism is positively related to PMB, and whether PMB is positively related to emotional exhaustion. Then, it is tested whether PMB acts as a mediation mechanism on the relationship between leader Machiavellianism and emotional exhaustion. The results of the regression analysis are shown in Table 7, in which it is additionally controlled for leader tenure and follower gender.

Hypothesis 7 predicts a positive relationship between leader Machiavellianism and PMB. As shown by Table 7, this relationship is not significant (B=-0,07; p>0,05), meaning it cannot be proven that when a leader is Machiavellian, their manipulative behavior is recognized by followers. Consequently, Hypothesis 7 is rejected. When

(33)

considering control variables, it can be seen that both leader tenure in months (B=0,00; p>0,05) and follower gender (B=-0,22; p>0,05) do not appear to be significantly related to PMB.

Subsequently, it is predicted by Hypothesis 8 that PMB is positively related to emotional exhaustion, meaning that followers would recognize manipulative behavior, which in turns leads to emotional exhaustion. As can be seen in Table 7, a statistically significant relationship has been found between PMB and emotional exhaustion (B=0,22; p<0,05). In order words, perceived manipulative behavior significantly affects emotional exhaustion. Also, when considering control variables, it can be seen that follower gender significantly affects emotional exhaustion (B=-0,48; p<0,05). In addition, it is found that the model explains 11% of the variance in emotional exhaustion (F=(4,90)=3,56; p<0,05)). Taken together, Hypothesis 8 is supported. This means that followers, who perceive their leader to be manipulative, are more likely to develop emotional exhaustion.

Table 7

Regression analysis simple mediation (model 4)

Consequent

PMB (M) Emotional exhaustion (Y)

Leader Machiavallianism (X) -0,07 0,11 0,52 0,05 0,12 0,69 PMB (M) - - - 0,22 0,11 0,04 constant 2,26 0,53 0,00 2,87 0,59 0,00 Leader tenure in months 0,00 0,00 0,47 0,00 0,00 0,06 Follower gender -0,22 0,2 0,28 -0,48 0,21 0,02 R2 = 0,02 R2 = 0,11 F (3,91) = 0,76, p>0,05 F (4,90) = 3,56 , p<0,05 N=95

(34)

Table 8

Confidence interval: direct, indirect, and total effect (model 4)

Effect SE p LLCI ULCI

Direct effect 0,05 0,12 0,69 -0,18 0,28

Total effect 0,03 0,12 0,79 -0,20 0,26

Boot SE Boot LLCI

Boot ULCI

Indirect effect -0,02 0,03 -0,11 0,02

Finally, Hypothesis 9 predicts PMB to act as a mediating mechanism on the relationship between leader Machiavellianism and emotional exhaustion. However, as can be seen in Table 8, a bias-corrected bootstrap 95% confidence interval for the indirect effect (B=-0,02) tends to contain zero (LLCI=-0,11; ULCI=0,02). This means that it cannot be proven that the relationship between leader Machiavellianism and emotional exhaustion is indirectly affected by PMB. Therefore, Hypothesis 9 is rejected.

4.3.2. The moderating role of Leader-Member Exchange

Next, using model 7 of Process by Andrew Hayes (2013), it is tested whether LMX moderates the relationship between leader Machiavellianism and PMB, as formalized by Hypothesis 10. As can be seen in Table 9, no significant interaction has been found (B=0,08; p>0,05). This means that it cannot be proven whether different quality-levels of the relationship between leader and follower affect whether the follower perceives the behavior of the Machiavellian leader to be manipulative. As a result, Hypothesis 9 is rejected. In addition, when considering control variables, it can be seen that both

(35)

leader tenure in months (B=0,00; p>0,05) and follower gender (B=-0,20; p>0,05) do not significantly affect PMB.

Table 9

Regression analysis moderated mediation (model 7)

Consequent

PMB (M) Emotional exhaustion (Y)

Leader Machiavallianism (X) -0,06 0,11 0,57 0,05 0,12 0,40 PMB(M) - - - 0,22 0,11 0,04 LMX (W) -0,37 0,48 0,22 - - - Leader Mach x LMX (XW) 0,08 0,18 0,45 - - - constant 2,10 0,36 0,00 2,99 0,43 0,00 Leader tenure in months 0,00 0,00 0,62 0,00 0,00 0,08 Follower gender -0,23 0,20 0,25 -0,44 0,21 0,04 R2 = 0,09 R2 = 0,14 F (5,89) = 1,85, p>0,05 F (4,90) = 3,80 , p<0,05 N=95

Finally, the moderated mediation as formalized in Hypothesis 11, predicts that leader Machiavellianism is related to follower exhaustion, such that the positive relationship with PMB is weaker when LMX is high. As can be seen in Table 10, a closer inspection of the findings with regard to a 95% confidence interval, support the insignificant relationship as explored at different levels of LMX appear to be insignificant. Therefore, Hypothesis 11 is rejected.

(36)

Table 10

Bootstrapping results for the test of conditional indirect effects at specific values of the moderator (PMB)

Mediator PMB Conditional indirect effect SE 95% CI

LLCI ULCI

LMX -0,80 0,00 0,02 -0,04 0,04

0,00 0,00 0,01 -0,04 0,02

0,95 0,00 0,02 -0,07 0,03

Note: Results are based on 5000 bootstrap samples, CI=confidence interval

*p<0,05, **p<0,01

5. Discussion

The aim of this research was to elaborate on current research by investigating whether leader Machiavellianism is positively related to follower’s emotional exhaustion. In addition, it was examined whether the quality of the relationship between leader and follower (LMX) is indirectly related to the relationship between leader Machiavellianism and emotional exhaustion as a mediation mechanism, as research has failed to examine LMX as a possible explanation of this relationship. Since Machiavellians are primarily concerned with themselves, personal goal attainment by manipulating others, and show little consideration and empathy, this is likely to lead to low-quality relationships between leader and follower. In addition, social relationships are important concerning followers’ health. Therefore, LMX was tested as a possible explanation of the relationship between leader Machiavellianism and emotional exhaustion.

Also, there is inconsistency in research on whether followers recognize Manipulative behavior since Den Hartog and Belschak (2012) argue that followers have the ability to recognize Machiavellians inauthentic behavior. On the contrary, Kwak and Shim (2017) state that followers may have difficulties in recognizing

(37)

leaders’ manipulative behavior as Machiavellians are superior liars. Therefore, the possible moderation of PMB on the relationship between leader Machiavellianism and LMX was incorporated. Namely, it was expected that this relationship would be stronger when higher levels of PMB were present. Although the overall results do not support the conceptual research model, additional analysis supports the positive relationship between perceived manipulative behavior and emotional exhaustion. This means that when a follower recognizes leader’s manipulative behavior, the follower is more likely to score high on emotional exhaustion.

First, the results do not confirm the positive direct effect of leader Mach with follower’s emotional exhaustion, as formalized by Hypothesis 1. This can be the result of the small sample size in this research (101 dyads). Research has confirmed that leader Machiavellianism has negative consequences for follower emotional exhaustion, which can even lead to burnout (Gkorezis et al., 2015), although there may be other variables that need to be taken into consideration.

For instance, follower’s personality traits may moderate the relationship between leader Machiavellianism and follower emotional exhaustion as people differ in the way they respond to leaders due to their characteristics (Kahill, 1988). For example, an individual who is high on extraversion may be less sensitive towards leader Machiavellianism and in turn, to the development of emotional exhaustion as extraverts demonstrate positive emotions, such as enthusiasm and optimism, than individuals low on extraversion. Also, since people high on neuroticism are related to strong emotional reactions to stressful circumstances (Van Heck, 1997), people who are emotionally stable may act in a more calmly and problem solving manner (McCrae & Costa, 1986). Also, high emotional stability is related with the ability to deal with stress, increased self-esteem, and optimism (Johnson & Ostendorf, 1993),

(38)

which may influence their reactions regarding the leader. Therefore, the degree of neuroticism/emotional stability may influence the relationship between leader Machiavellianism and emotional exhaustion.

In addition, organizational climate may be a possible moderator as well on the relationship between leader Machiavellianism and emotional exhaustion. As Verbeke, Ouwerkerk and Peelen (1996) argue, an organization that is characterized by high levels of internal communication and behavior-focused control regulations, may discourage individuals to behave Machiavellian. In turn, this may lead to Machiavellians resign from this organization.

Second, as was not expected, no relationship has been found between leader Machiavellianism and LMX. This means that it cannot be proven that the quality of relationship between leader and follower suffer from the leader being high on Machiavellianism. A possible explanation is that the frequency of working at home may influence the relation between leader Machiavellianism and emotional exhaustion. Due to technological innovations, there are more options for followers to work from home. In turn, working from home increases feelings of autonomy and decreases the contact between leader and follower (Shamir & Salomon, 1985). As a result, the follower is less frequently confronted with its leader. In line with research by Gajendran and Joshi (2012), the frequency of contact may act as a moderator on the relationship between leader Machiavellianism and LMX.

Third, the relation between LMX on follower’s emotional exhaustion was not supported by the results, meaning it cannot be proven whether a high quality relation between leader and follower diminishes the likelihood of feeling emotionally drained. Halbesleben & Buckley (2004) acknowledge, besides the effects of demands and resources, the importance of other variables, such as received support of colleagues.

(39)

Therefore, when this support is present, it may moderate the effect of a low-quality relationship between leader and follower and emotional exhaustion. Thus, the levels of received support of colleagues could be a possible moderator on the relationship between LMX and emotional exhaustion.

Taken the above together, no mediation effect has been found of LMX on the relationship between leader Machiavellianism and emotional exhaustion. This means that it cannot be proven whether LMX is an explanation of this relationship. This may be the result of the fact that there is another explanation for the relationship between leader Machiavellianism and emotional exhaustion.

Finally, it was investigated whether perceived manipulative behavior moderates the relationship between leader Machiavellianism and emotional exhaustion, so that when one strongly recognizes a Machiavellians manipulative behavior, this relation would be stronger. However, this cannot be proven by the results. A possible explanation may be that followers perceive themselves as similar to leaders. Followers may not perceive their leader to be manipulative, but rather as successful (Weiss, 1977). In addition, Schyns (2015) argues that when leader and follower are both high on Machiavellianism, two things can happen. First, leader and follower may try to manipulate each other, leading to a low-quality relationship, or second, leader and follower conclude to cooperate to manipulative others in order to achieve common goals (Schyns, 2015). Therefore, it may be that the degree of perceived similarity moderates the relationship between leader Machiavellianism and Leader-Member Exchange.

(40)

5.1. Limitations

As the results do not support the research model, and since there is a need to continue to examine these issues, some facets of this study can be improved and extended. The important limitations are discussed in this sub-section.

First of all, since this study is characterized by cross-sectional design, causality cannot be proven. Therefore, future research should carefully investigate these relationships in a longitudinal manner. Furthermore, participants were recruited using personal and work relationships of each of the seven students who worked on the master thesis course of the Amsterdam Business School. Thus, participants were acquired based on simplicity and availability, resulting in a less generalizable sample (Kam, Wilkin & Zechmeister, 2007). In order to increase generalizability, future research should randomly select participants.

Also, although anonymity of the survey was guaranteed, participants received a code, which they had to use in order to link their answers to the other party, participants may have been careful with their responses concerning their leader. Future research may act on this issue by implementing a different approach in which no code needs to be entered.

Subsequently, the survey measured multiple variables for seven different thesis studies. Since the duration of completing the follower survey was estimated at more than 15 minutes, this may have been a disadvantage for the reliability of the findings. As some participants completed the survey in a short time, this could indicate that questions were not carefully read. This especially applies to the variables that were at the end of the survey.

Furthermore, this research only consisted of 101 useful and complete dyads, which is a disadvantage regarding the power of the research. Therefore, caution is

(41)

required with regard to generalizability and power of the results. In order to diminish this issue, future research should approach more individuals to participate in their research in order to create power with respect to the results.

5.2. Future research directions

The above mentioned limitations suggest future improvements regarding research design and method to, for instance, increase generalizability and research power. In this sub-section possible future directions are discussed.

First of all, although this research did not confirm the relationship between leader Machiavellianism and emotional exhaustion, previous research by Gkorezis et al. (2015) does confirm this relationship. However, little research has been done regarding the moderating effects of followers’ personality traits on the relationship between leader Machiavellianism and emotional exhaustion. Since individuals differ with regard to their response due to their personality characteristics, it is likely that individuals differ in their reaction with respect to leader Machiavellianism. For instance and as mentioned before, individuals high on emotional stability may respond differently on leader Machiavellianism than individuals high on neuroticism, with regard to emotional exhaustion. It may therefore be interesting for future research to investigate the effect of personality characteristics on the relationship between leader Machiavellianism and emotional exhaustion.

In addition, it may be interesting to incorporate a situation in which the follower and leader are both high on Machiavellianism (perceived similarity). Felfe and Schyns (2004) argue that when followers perceive their leader to be similar, they are likely to perceive their leader to be successful. Thus, the degree of perceived similarity with respect to Machiavellianism may influence the relationship between

(42)

leader Machiavellianism and the quality of the relationship between leader and follower. In turn, perceived similarity between leader and follower with regard to Machiavellianism may have consequences for other followers as well. Namely, as mentioned before, with both follower and leader are high on Machiavellianism this may lead to the collaboration of jointly manipulating others, or to a low-quality relationship when leader and follower try to manipulative each other (Schyns, 2015). Therefore, the degree of perceived similarity between a leader and follower may influence the relationship between leader Machiavellianism and emotional exhaustion of other individuals.

Finally, future research may examine LMX as a moderator, rather than a mediator on the relationship between leader Machiavellianism and emotional exhaustion. As mentioned before, Machiavellians are characterized by a high degree of manipulative behavior, and are low on consideration and empathy (Christie & Geis, 1970). Therefore, leaders high on Machiavellianism may do their best to have a high-quality relationship with a follower when the leader perceives this as beneficial with regard to goal attainment. However, when the follower is not needed anymore, the leader may cut off the relationship with the follower, resulting in a low-quality relationship. Therefore, it is possible for LMX to have a moderating role on the relationship between leader Machiavellianism and emotional exhaustion, which may be interesting to investigate using future research.

5.3 Practical implications

Although this study found support for one Hypothesis by additional analysis, this may have practical implications for organizations. Namely, it is found that when followers perceive their leader to show manipulative behavior, they are more likely to develop emotional exhaustion. Since emotional exhaustion has severe consequences for not

The relationship between leader Machiavellianism, leader-member exchange, and emotional exhaustion : what if followers do not recognize manipulative behavior?

Contents