Fear is the path to the dark side: the interplay of leader fear of power loss and leader Machiavellianism on abusive supervision Open Access

Within organizations leaders often hold significant power (Rus et al., 2012), granting them asymmetric control over valued material and social resources. Power offers clear advantages. First, control over resources and the authority to allocate them enables powerful individuals to accumulate financial and material wealth (Guinote, 2017). Since power is also linked to a focus on rewards (Anderson and Brion, 2014) and the pursuit of personal goals (Guinote, 2017), the powerful take deliberate steps to secure present and future resources (Wisse et al., 2019). Second, power provides social benefits, often leading to prestige and status (i.e. admiration and respect in the eyes of others; Anderson and Brion, 2014). It also allows for greater autonomy and provides the possibility and the means to influence others (Sturm and Antonakis, 2015). Third, power brings psychological rewards: it fulfills a fundamental human motive (McClelland, 1985) and individuals have been shown to value power in and of itself (Fehr et al., 2013). Notably, it has been linked to increased optimism (Anderson and Galinsky, 2006), self-esteem (Fast and Chen, 2009) and well-being (Kifer et al., 2013). In sum, having power comes with material, social and psychological benefits (Anderson and Brion, 2014).

In organizations, power can quickly shift with leaders being fired, demoted, sidelined or otherwise stripped of their power, due to internal power struggles (Greer et al., 2017). Beyond internal politics, societal changes can also lead to a redistribution of power. Organizations are increasingly adopting flatter, less hierarchical structures and the rise of Artificial Intelligence is projected to replace large segments of middle management in the not-so-distant future (Govindarajan et al., 2021). Importantly, leaders’ positions are often threatened by rivals amongst their subordinates, seeking to climb the corporate ladder and access the benefits that power provides. The prospect of losing power is thus a reality that most leaders face (Babalola et al., 2023). Whereas the benefits of having power are clear, it is less clear how leaders react when faced with the prospect of losing power, especially to their subordinates.

We draw on COR theory (Halbesleben et al., 2014) to argue that leaders fearing power loss are more likely to engage in abusive supervision. COR theory posits that individuals are loss averse and motivated to “retain, foster and protect those things they centrally value” (Hobfoll et al., 2018, p. 104). The theory depicts individuals as highly sensitive to the potential loss of valuable resources which may lead them to react in defensive, aggressive or irrational ways when these are threatened (Hobfoll et al., 2018). Power should thus have motivational potential: Although intangible, it can be a source of other material, social and psychological resources (Anderson and Brion, 2014) and the prospect of losing power should be psychologically aversive and motivate leaders to avoid the pain of loss. This is corroborated by research linking unstable power positions to higher stress levels, especially for those driven by personal motivational goals of power and dominance (Feenstra et al., 2017). Research has also shown that individuals facing status loss (another valuable intangible asset) are motivated to avoid that pain: To protect their status and avoid its loss, individuals were willing to pay more, work harder and act self-servingly (Pettit et al., 2010). Thus, we posit that leaders fearing power loss are driven to protect and retain their privileged position.

With protecting and retaining power as a salient aim, leaders fearing power loss are likely to actively and persistently pursue this self-focused goal (Situated Theory of Power; Guinote, 2017). Clearly, the behaviors aimed at protecting power can vary, yet the empirical record is surprisingly scant. The only two studies we could identify, found that leaders fearing power loss to subordinates acted more self-servingly (e.g. by taking credit for subordinates’ work; Wisse et al., 2019) and hid knowledge from them (Iqbal et al., 2022). However, according to COR theory (Hobfoll et al., 2018), leaders who fear losing power to their subordinates could also react aggressively when their power is threatened. This could involve using hostile tactics to reassert dominance and control (Bai, 2017) over those who threaten their power (and are likely seen as a source of provocation) while simultaneously signaling to other potential challengers that there will be a price to pay. Research in related domains has shown that those in unstable (vs. stable) power positions tend to distrust others (Mooijman et al., 2019), negatively evaluate subordinates (especially if they are skilled and voice their opinions; Yang et al., 2021) and seek physical proximity to threatening team members (arguably to be able to exert control over them; Mead and Maner, 2012). They may also sabotage skilled employees by withholding information, excluding them from group tasks and physically isolating them from others (Case and Maner, 2014; Maner and Mead, 2010). These findings are also in line with the frustration-aggression hypothesis (Berkowitz, 1989) positing that people react aggressively if they feel blocked in attaining the anticipated gratification associated with an attractive goal. That is, people who feel thwarted in reaching an attractive goal (e.g. maintaining their power) should be more likely to lash out and act aggressively. In sum, we argue that leaders who fear losing power are more likely to reassert and protect this power by engaging in abusive supervision.

Whereas most leaders value power, some may value it more than others (Anderson and Brion, 2014), and therefore, may be more likely to abuse their subordinates when faced with the prospect of losing it. COR theory explicitly asserts that individual differences (e.g. values, dispositional tendencies) shape responses towards potential resource losses (Halbesleben et al., 2014). Therefore, it is essential to consider dispositional factors that may moderate the relationship between leader fear of power loss and abusive supervision. Leader Machiavellianism is particularly relevant here because Machiavellians not only highly value power but are also willing to use any means to achieve their goals (Christie and Geis, 1970).

The concept of Machiavellianism derives from Niccolo Machiavelli’s book The Prince (1,513/2003), written as a guide for political rulers on how to seize and retain power by using any means available, even if those involve duplicity, deception or manipulation. As a dispositional tendency, Machiavellianism refers to a relatively stable interpersonal strategy of deceiving and manipulating others to achieve personal gain (Christie and Geis, 1970; Dahling et al., 2009; Jones and Paulhus, 2009). Those scoring high on Machiavellianism are characterized by a particular cluster of attributes that can be summed up by (1) a strong drive to achieve expedient goals (i.e. money, power, status, interpersonal control) and (2) the willingness to use all possible means to achieve these goals (Jones and Paulhus, 2009; Sakalaki et al., 2007). Research indicates that Machiavellians are self-focused and egocentric (Fehr et al., 1992), emotionally detached from their actions, less likely to feel empathy, guilt or remorse (Christie and Geis, 1970; Spain et al., 2014), distrust others, have a cynical view of human nature (Sakalaki et al., 2007), and are willing to manipulate or harm others for personal gain (Dahling et al., 2009).

Leader Machiavellianism could strengthen the relationship between leader fear of power loss and abusive supervision for at least two reasons. First, COR theory (Halbesleben et al., 2014) emphasizes that individual differences shape responses towards potential resource losses. Since leaders high in Machiavellian disproportionately value power, status and control (Dahling et al., 2009; Sakalaki et al., 2007), they should be particularly sensitive to the prospect of losing it and motivated to protect and retain it. This aligns with an evolutionary functional analysis of aggression suggesting that competitive aggression over resources arises from a basic cost-benefit analysis (Brown, 1964). Brown’s (1964) principle of “economic defendability” asserts that individuals will protect resources from potential rivals only if the benefits outweigh the defense-related costs, such as time, energy or the associated risks. Leaders faced with the prospect of power loss likely arrive at different cost-benefit ratios depending on their level of Machiavellianism. Since high-scoring Machiavellians particularly value power, the potential advantages of engaging in abusive supervision (i.e. defending their power) should more readily outweigh any potential drawbacks.

Second, while many leaders – even when faced with the prospect of losing power – may hesitate to abuse their subordinates, those high in Machiavellianism are less likely to have such qualms given their emotional detachment, self-centeredness, dominance and predisposition to engage in anti-normative behavior (LeBreton et al., 2018). Thus, they may be especially willing and likely to engage in abusive supervision when their power feels threatened. This aligns with trait activation theory (Tett and Burnett, 2003) which suggests that behavior depends not only on the strength of a trait, but also on trait-relevant situational cues. Accordingly, if the situation (e.g. facing the prospect of losing power) calls for or stimulates a specific behavior (e.g. abusive supervision), the behavior is more likely to manifest for those who already show an innate predisposition towards that behavior (e.g. leaders higher in Machiavellianism). For instance, research has shown that powerful individuals – especially if scoring high on Machiavellianism – were more likely to abuse their subordinates, arguably because power afforded them the freedom to act on their disagreeable tendencies (Wisse and Sleebos, 2016). Similarly, De Hoogh et al. (2021) found that low rule or highly instrumental organizational climates predicted abusive supervision, but only among high-Machiavellian leaders. We argue that the prospect of losing power acts as a trait-relevant situational cue facilitating the expression of Machiavellian anti-normative behavioral tendencies (e.g. abusive supervision). In sum, we propose that leader Machiavellianism strengthens the positive relationship between leader fear of power loss and abusive supervision (see Figure 1).

Study 1 aimed to provide causal evidence for the proposed relationships between our independent variables and abusive supervision. In a sample of U.S.-based business leaders, we manipulated leader fear of power loss, measured leader Machiavellianism, and assessed their effects on leaders’ projected likelihood of engaging in abusive supervision.

Participants in our online business scenario experiment were leaders recruited through Prolific Academic, paid UK £2.50 for their participation. Eligible leaders lived in the United States, were native English speakers, held supervisory roles, and worked at least three days a week. Only those completing all measures were included in the analyses. To address concerns regarding online data quality we excluded participants who failed attention checks ^[1] (N = 10) or requested their data not to be used (N = 29). The final sample included 365 leaders (43.8% female; 55.3% male, 0.8% other; M_age = 33.52, SD = 8.37). Average tenure in the current role was 3.79 years (SD = 3.51), and leaders had an average of 9.58 subordinates (SD = 18.88). Most (81%) held a bachelor’s degree or higher and worked in a diverse set of industries (e.g. business and finance, technology, etc.).

After completing a Machiavellianism assessment, leaders read a workplace scenario in which they imagined supervising a team of seventeen employees. Leaders were randomly assigned to one of two experimental conditions (low fear of power loss: N = 185; high fear of power loss: N = 180; see Supplementary Materials for the full manipulations). They then completed manipulation checks, reported their likelihood of engaging in abusive supervision, and provided demographic information. Respondents received information on the voluntary and confidential nature of the research and provided active informed consent. The study was approved by the university’s ethics committee.

Fear of Power Loss Manipulation. Our fear of power loss manipulation was identical to the Wisse et al. (2019) manipulation. In the high fear of power loss condition, leaders read that they were at risk of losing their power position since their employees were actively striving for it and sabotaging them, whereas in the low fear of power loss condition, they read that they had a good shot of maintaining their power position since their employees were actively rallying behind them and helping them.

Leader Machiavellianism. Leader Machiavellianism was assessed with the short 8-item version of the Mach-IV scale (Christie and Geis, 1970) shown to have good reliability in previous research (De Hoogh et al., 2021). Leaders indicated the extent to which they agreed (1 = fully disagree; 5 = fully agree; α = 0.85) with items such as “It is hard to get ahead without cutting corners here and there”.

Manipulation Check. We used the Wisse et al. (2019) 3-item scale as our manipulation check. Items were adapted to fit the scenario context and were prefaced by “In this situation,”. Leaders indicated the extent to which they agreed (1 = fully disagree; 7 = fully agree; α = 0.94) with items such as “I would fear that my leadership would be undermined by my subordinates”.

Likelihood to Engage in Abusive Supervision. We used Tepper’s (2000) 15-item abusive supervision scale to measure leaders’ likelihood of engaging in abusive supervision (1 = strongly disagree; 5 = strongly agree; α = 0.96). Items were adapted to fit the scenario context and were prefaced by “In this situation,”. A sample item is “I would ridicule some of my subordinates”.

We conducted confirmatory factor analyses (CFA) to confirm the distinctiveness of our leader-rated measures (i.e. leader Machiavellianism and abusive supervision). Our hypothesized two-factor model was a better fit to the data (χ²(229) = 431.38, p < 0.001, RMSEA = 0.05, CFI = 0.94, TLI = 0.94, SRMR = 0.05) than the one-factor model (χ²(230) = 832.66, p < 0.001, RMSEA = 0.09, CFI = 0.83, TLI = 0.82, SRMR = 0.10; Δχ² = 401.28, Δdf = 1, p < 0.001).

A one-way ANCOVA on the manipulation check, controlling for leader Machiavellianism, revealed that leaders in the high fear of power loss condition (M = 5.80, SD = 1.07) experienced a higher fear of power loss than those in the low fear of power loss condition (M = 2.85, SD = 1.40; F(1, 362) = 518.72, p < 0.001, η²_p = 0.59). We also found a small effect of Machiavellianism, F(1, 362) = 10.07, p < 0.01, η²_p = 0.02: leaders scoring higher (vs lower) on Machiavellianism feared losing power more, which is consistent with research showing that Machiavellians are more focused on maintaining power (Dahling et al., 2009; Li et al., 2024).

We tested our hypotheses using multiple regression analyses, regressing the likelihood of engaging in abusive supervision on leader fear of power loss (−1 = low fear of power loss; 1 = high fear of power loss), leader Machiavellianism, and their interaction (see Supplementary Materials for additional analyses). Leader Machiavellianism was mean-centered before computing the interaction term. The overall model significantly predicted leader likelihood of engaging in abusive supervision, F(3, 361) = 73.39, p < 0.001, R² = 0.38. In support of Hypothesis 1, leader fear of power loss significantly predicted abusive supervision, b = 0.34, SE = 0.04, p < 0.001, 95% CI [0.27, 0.41]. Leaders in the high fear of power loss condition were more likely to abuse their subordinates than leaders in the low fear of power loss condition. This main effect was qualified by our predicted Fear of Power Loss × Machiavellianism interaction (Hypothesis 2), b = 0.14, SE = 0.04, p < 0.01, 95% CI [0.05, 0.22], R²_change = 0.016. As predicted, simple slopes analyses showed that the positive effect of leader fear of power loss on the likelihood of engaging in abusive supervision was stronger when leader Machiavellianism was high (+1 SD; b = 0.46, SE = 0.05, p < 0.001, 95% CI [0.35, 0.56]) than when it was low (−1 SD; b = 0.22, SE = 0.05, p < 0.001, 95% CI [0.12, 0.33]). Additionally, in line with Wisse and Sleebos (2016), we found a positive relationship between leader Machiavellianism and the likelihood of engaging in abusive supervision, b = 0.50, SE = 0.04, p < 0.001, 95% CI [0.41, 0.58].

Study 1 provided causal evidence that leader fear of power loss predicts abusive supervision, and that this relationship is stronger for leaders scoring higher on Machiavellianism. Study 2 aimed to replicate these findings in a multi-source, cross-sectional design with Dutch team leaders and their subordinates. Unlike Study 1, which relied on leaders’ self-reported likelihood of abusive behavior, Study 2 assessed abuse through subordinate perceptions, testing the robustness of our Study 1 findings in a naturalistic setting in a manner less sensitive to social desirability biases.

Data were collected from Dutch leaders and their direct subordinates, resulting in 2,451 completed responses (574 leaders and 1,877 subordinates with an overall response rate of 74%). After excluding incomplete responses, the final sample comprised 546 leaders matched with 1,718 subordinates, with an average of 3.15 subordinates per leader (SD = 1.88). Among leaders, 45.6% were female, 54% male and 0.4% other; M_age = 40.88 years (SD = 11.48); had an average tenure of 5.16 years (SD = 5.79) and 87% had at least a bachelor’s degree. Among subordinates, 55% were female, 44.5% male and 0.5% other; M_age = 35.21 years (SD = 12.88); had an average tenure in the of 3.86 years (SD = 5.25); 74% had at least a bachelor’s degree and worked in a diverse set of industries (e.g. agriculture, manufacturing, education etc.).

Data were collected as part of a larger study on leadership via convenience sampling. Graduate students recruited respondents via their workplaces, personal networks, and networks of acquaintances, contacting them by email, telephone, or face-to-face. To qualify, respondents needed to: (1) have paid jobs; (2) work in teams with one identifiable leader; and (3) the team had at least three members. Leaders and subordinates received separate electronic questionnaire links, hosted on Qualtrics, tailored for their respective roles. Leader-subordinate responses were matched using unique numerical identifiers. Participants were informed about the voluntary and confidential nature of the research and provided active informed consent. The study complied with the university’s ethical committee guidelines.

Leader Fear of Power Loss. Leader fear of power loss was measured with the same 3-item scale (Wisse et al., 2019) used in Study 1 (1 = fully disagree; 7 = fully agree; α = 0.76), the only difference being that “In this situation, I would” was replaced by “I sometimes”.

Leader Machiavellianism. Leader Machiavellianism was assessed with the same short 8-item version of the Mach-IV scale (Christie and Geis, 1970) used in Study 1 (1 = fully disagree; 5 = fully agree; α = 0.80).

Perceived Abusive Supervision. Subordinates’ perceptions of leaders’ abusive supervision were measured with the 15-item abusive supervision scale (Tepper, 2000; 1 = fully disagree; 5 = fully agree; α = 0.95).

Table 1 reports the means, standard deviations, and correlations between the study variables. Due to the nested structure of our data, we conducted a series of multilevel confirmatory factor analyses (MLCFA) in RStudio to assess the construct distinctiveness of our Level-2 (leader fear of power loss, leader Machiavellianism) and Level-1 (perceived abusive supervision) variables. The results indicate that our hypothesized three-factor model provided the best fit to the data, χ²(296) = 726.66, p < 0.001, RMSEA = 0.04, CFI = 0.96, TLI = 0.95, SRMR = 0.03 relative to a two-factor model (combining leader fear of power loss and leader Machiavellianism into one factor), χ²(298) = 1,270.70, p < 0.001, RMSEA = 0.06, CFI = 0.92, TLI = 0.91, SRMR = 0.04, Δχ²(2) = 544.04 or a single-factor model (χ²(299) = 2,782.30, p < 0.001, RMSEA = 0.11, CFI = 0.77, TLI = 0.75, SRMR = 0.12, Δχ²(3) = 2055.64).

Table 2 reports the results of our multilevel regressions analyses (see Supplementary Materials for additional analyses) conducted in Jamovi. Leader fear of power loss and leader Machiavellianism were grand-mean centered and entered as Level-2 predictors, with subordinate perceptions of abusive supervision modelled as a Level-1 outcome. Consistent with Study 1, leader fear of power loss positively predicted subordinate perceptions of abusive supervision (γ = 0.03, SE = 0.01, p < 0.01, 95% CI [0.01, 0.06]). In line with Study 1, this main effect was qualified by our predicted interaction between leader fear of power loss and leader Machiavellianism (γ = 0.05, SE = 0.02, p < 0.01, 95% CI [0.01, 0.09]). Simple slope analyses showed that leader fear of power loss was significantly positively related to subordinate perceptions of abusive supervision when leader Machiavellianism was high (+1 SD; estimate = 0.05, SE = 0.02, p <. 01, 95% CI [0.02, 0.07]) but not when it was low (−1 SD; estimate = −0.01, SE = 0.02, ns, 95% CI [−0.04, 02]). Similar to Study 1, leader Machiavellianism was positively related to subordinate perceptions of abusive supervision (γ = 0.09, SE = 0.03, p < 0.01, 95% CI [0.03, 0.15]).

In two studies, we found that leader fear of power loss is positively related to the likelihood of engaging in abusive supervision (Hypothesis 1) and that this effect is particularly strong for leaders scoring higher on Machiavellianism (Hypothesis 2). In line with earlier work (Wisse and Sleebos, 2016), leader Machiavellianism was positively related to abusive supervision in both studies.

Our work contributes to the leadership and abusive supervision literature in several ways. First, it underscores the value of examining the effects of leader fear of power loss. In today’s organizational landscapes, power is not a given: Recent trends towards flatter organizational structures coupled with age-old internal power struggles make the prospect of losing power a reality that most leaders face (Babalola et al., 2023; Greer et al., 2017). Despite having made significant strides in understanding how power is gained and how it affects those who have it, our insight into how leaders react when faced with the prospect of losing power – especially to their subordinates – remains underexplored (Wisse et al., 2019). Our finding that leader fear of power loss can result in abusive supervision (especially for those scoring higher on Machiavellianism), extends earlier limited research linking it to self-serving behavior (Wisse et al., 2019) and knowledge hiding (Iqbal et al., 2022) and enriches our understanding of the broader range of possible destructive behaviors that may be triggered by leader fear of power loss.

Second, this research increases our understanding of when leader fear of power loss may result in destructive behaviors. We show that fear of power loss does not invariably translate into abusive supervision, but rather hinges on dispositional tendencies, namely Machiavellianism. This aligns with COR theory, which asserts that individual differences shape responses towards potential resource losses (Halbesleben et al., 2014), and trait activation theory (Tett and Burnett, 2003), which suggests that behavior is contingent on both trait strength and trait-relevant situational cues. Our findings imply that leader fear of power loss is more likely to translate into abusive behaviors for those who highly value power and have a predisposition towards abuse. Future research could examine other dispositional tendencies that moderate this relationship, such as need for power (McClelland, 1985), bottom-line mentality (Greenbaum et al., 2012) or humility (Morris et al., 2005). Second, by investigating a dispositional moderator, our work also builds on Wisse et al. (2019), who identified competitive climate as a situational moderator. Future studies might explore additional organizational factors, such as aggressive norms (Mawritz et al., 2014) or climates of objectification (Zhang and Wisse, 2024), which may normalize abuse—unlike ethical climates that may inhibit it (Martin and Cullen, 2006).

Third, by examining how leaders’ perceptions of subordinate behavior (as power-threatening) relate to subordinate perceptions of abuse, our research aligns with calls for a more dynamic perspective on destructive leadership. Abusive supervision is increasingly seen as a process shaped over time through leader–subordinate interactions within specific contexts (Fischer et al., 2021). Yet, research rarely takes a dynamic perspective considering the interaction of leader, subordinate, and contextual factors (Breevaart et al., 2022). By incorporating leaders’ fear of power loss—triggered by subordinate behavior—and leader Machiavellianism, our study explores this interplay, offering a more nuanced view of how abuse emerges. Future research could build on this by employing longitudinal, multi-source designs (e.g. at least 6 months) to track how these processes evolve over time and explore potential discrepancies in leaders’ and subordinates’ perceptions of this same behavior. Additionally, future research could examine how engaging in abuse affects leaders over time by considering relational dynamics (e.g. the quality of leader-subordinate relationships; subordinates’ reactions to the abuse; the timeframe under consideration).

One clear strength of our research is its multi-study, multi-method approach, combining a scenario experiment and a multi-source field study across diverse samples (U.S. business leaders; Dutch team leaders and subordinates) and varied operationalizations of key concepts (i.e. manipulation vs. measurement of leader fear of power loss; self-reported vs. subordinate-reported abusive supervision). Although each paradigm has certain drawbacks, replication over different paradigms, populations and operationalizations should warrant some confidence in the validity and generalizability of our findings. Study 1 provides causal evidence but is limited by its hypothetical nature and reliance on self-reports. Study 2, although mute in matters of causality, employs a multi-source design – linking leader-reported fear of power loss and Machiavellianism with subordinate-reported perceptions of abusive supervision – reducing common source method variance.

We used a composite measure of Machiavellianism. While this aggregate approach is common (e.g. De Hoogh et al., 2021), several authors have argued that Machiavellianism comprises distinct facets – such as amoral manipulation, desire for status, distrust, and control – and have developed measures to assess them (Collison et al., 2018; Dahling et al., 2009). Although these efforts have had mixed success (Miller et al., 2015), future research could assess whether specific facets drive the interaction with fear of power loss on abusive supervision. Moreover, future work might examine whether the strength of this relationship depends on the source of Machiavellianism ratings (e.g. self-vs. employee reports).

We operationalized leader fear of power loss, by examining leaders’ fear of losing power to their subordinates. This choice was central to our research, as we aimed to understand whether leaders feeling threatened in their power by their subordinates would turn to abuse. However, leaders may fear losing power for different reasons, such as feeling threatened by competitive peers or organizational changes signaling demotions or layoffs. Leaders might also fear losing potential future power due to evaluations negatively impacting their career trajectory. This raises the question of whether different sources of threat to leader power similarly predict their tendency to engage in abusive supervision. Based on the displaced aggression hypothesis (Miller et al., 2003), external threats to leader power, regardless of source, would positively predict subordinate abuse, since these are safe targets. However, this assumption would need to be empirically tested.

Our research has implications for workplace policies and managerial practice. First, since leaders’ fear of losing power may inadvertently contribute to subordinate abuse, it is important to safeguard leaders’ sense of safety in their positions. Organizations could increase trust between leaders and subordinates through task and outcome interdependencies (De Jong et al., 2016) and reduce uncertainty with fair, transparent procedures (e.g. promotion, demotion decisions). Tools like visible task management systems or collaborative platforms, could ease leaders’ fears by making contributions and decision-making processes more visible. Such transparency may, however, also expose poor leader performance, potentially heightening power insecurity. Thus, organizations should monitor the effects of such interventions closely.

Second, when selecting/promoting individuals into leadership positions, organizations should consider Machiavellianism levels. Indeed, they should not only select individuals into leadership roles by assessing positive traits but also select them out of leadership roles by assessing dark traits (Wisse and Rus, 2022), bearing in mind that there are optimal and dysfunctional levels of both positive and negative traits (LeBreton et al., 2018). Since Machiavellianism has been linked to increased faking, conducting background checks (Wisse and Rus, 2022) and employing assessment tools less vulnerable to faking (e.g. ethical dilemmas; LeBreton et al., 2018) could be beneficial.

Third, because leaders high in Machiavellian are especially responsive to contextual cues (Rauthmann and Will, 2011), organizations should adopt a systemic perspective that includes both informal and formal contextual factors to limit Machiavellian expression. Informally, organizational climates can serve as effective constraints by signaling valued behaviors. Promoting ethical (Martin and Cullen, 2006) and rule-based climates (De Hoogh et al., 2021; Laurijssen et al., 2024), along with cultures that encourage employee voice, appears particularly promising. Formally, organizations should review performance management systems, especially incentive structures based on all-or-nothing rewards, which may foster anti-normative behavior and hyper-competitive “winner-takes-all” climates (Park et al., 2022). Further, Machiavellian behaviors can be curbed through mechanisms that promote transparency, accountability (Rus et al., 2012), checks and balances (e.g. compliance departments), and visible sanctions for rule violations (Wisse and Rus, 2022).

In this research, we drew on COR theory and trait activation theory to examine how the interplay between leader fear of power loss and leader Machiavellianism engenders abusive supervision. We found that leader fear of power loss was especially likely to translate into abusive supervision to the extent that leaders scored higher on Machiavellianism. With this work, we hope to have opened an avenue for further exploring the potential effects of leader fear of power loss.

The supplementary material for this article can be found online.