People scoring high on primary psychopathy can be quite successful and thrive in organizations (Mullins-Sweatt et al., 2010) even though they are ruthless toward others, manipulative, callous, and egocentric (Babiak et al., 2010). The influence of so-called ‘corporate psychopaths’ can be damaging, particularly when they are able to reach leadership positions, because, as leaders, they usually have considerable control over one’s own and others’ resources (Galinsky et al., 2015) and have the means to secure and sustain their position (Anderson & Brion, 2014). This allows them to hold the organization in deadlock while they simultaneously hamper subordinate and organizational functioning (LeBreton et al., 2018). For instance, measures of psychopathy in leaders have been linked to reduced future stock returns and profit (Ten Brinke et al., 2018; Wisniewski et al., 2017), self-serving behavior (Barelds et al., 2018), increased workplace bullying and unfair supervision (Boddy, 2011a, 2015a, 2015b), and decreasing employee morale, well-being, and performance (Boddy, 2011a; LeBreton et al., 2018; Mathieu et al., 2014). This research explores whether the organization can successfully enforce certain policies to neutralize the negative impact that primary psychopathic leaders have on others.

Social cognitive theory (Bandura, 1971) stipulates that the expression of people’s traits is ultimately bound by situational affordances (Christiansen & Tett, 2008; see also Schyns et al., 2018). In line with this, we propose that leaders’ expression of primary psychopathic tendencies may be curtailed by enforcing three policies: setting clear organizational rules, sanctioning misconduct, and increasing the transparency of behaviors so that violations can be detected. Such organizational policies set clear boundaries regarding what is permitted and prohibited, make it difficult to conceal destructive behavior, and hinder that leaders ‘get away with it’ when transgressions have taken place. This may neutralize primary psychopathic leaders’ destructive behavior, because it lessens the likelihood of situational affordances in which the organization and their employees can be exploited (Bandura, 1971; Cohen, 2016; Mulder et al., 2015; Witt & Spector, 2012). That is, we argue that because primary psychopaths are not necessarily disinhibited (Patrick et al., 2009), they can operate based on cost–benefit calculus (Stevens et al., 2012) and because they are able to set and follow long-term goals if they think it is to their benefit (Snowden & Gray, 2011), primary psychopathic leaders behavior can be influenced by the organizational context.

All in all, we argue that our understanding of how to intervene and mitigate the destructive influence of primary psychopathic leaders’ may fruitfully include characteristics of the work environment. A more systemic approach, that takes the environment into account, may afford organizations multiple levers to address the negative influence of primary psychopathic leaders and may also lead to more effective and sustainable solutions than interventions solely aimed at the primary psychopathic leaders themselves (Wisse & Rus, 2022). Moreover, practitioners may use information about the extent to which enforcing clear rules, sanctioning misconduct, and transparency of behavior help organizations to deal with corporate psychopathy to their benefit.

The Corporate Psychopath: A Primary Psychopathy Perspective

Various conceptualizations of psychopathy exist, some of them suggesting that it is unitary in nature (e.g., in the context of the short Dark Triad assessment; Jones & Paulhus, 2014), and others arguing for a multi-dimensional structure (Lilienfeld et al., 2014; Miller & Lynam, 2012; Smith & Lilienfeld, 2013). A common distinction is made between primary and secondary psychopathy, with primary psychopathy capturing most of the core of the psychopathy concept (cf. Murphy & Vess, 2003). Individuals with higher levels of primary psychopathy are characterized by their callousness, manipulativeness, glibness, egocentricity, and a lack of empathy and guilt. The primary psychopath can appear charming and likeable while concealing a dark, cold, and affectively empty interior (Cleckley, 1951, 1976). Those who score high on primary psychopathy can fare relatively well in society (Benning et al., 2018; Fowles & Dindo, 2009), perhaps also due to the fact that they are not necessarily impulsive (Levenson, 1993; Patrick et al., 2009) and may, instead, be quite deliberative (Boddy, 2010a).

Individuals with higher levels of secondary psychopathy show impulsive and risky behavior, are stimulation seeking, and lead a parasitic lifestyle with antisocial felonious tendencies. However, their deficits can stem from other conditions, such as anxiety or mood disorders, and they are presumed to be capable of feelings like guilt, empathy, and love (see Lilienfeld et al., 2015). Individuals who score high on secondary psychopathy, or on both primary and secondary psychopathy, tend to end up incarcerated or institutionalized (as a result of aggressive and/or criminal behavior).

Notably, individuals scoring high on primary psychopathy are quite capable of criminal activities too, but they more often resort to white-collar crime such as fraud or embezzlement (see Boddy, 2016). Primary psychopaths have also been named successful psychopaths (Hall & Benning, 2006) or, if they operate in the organizational context, corporate psychopaths (Boddy, 2006; Babiak & Hare, 2006). Recent studies on primary psychopathy in the organizational context have been focusing on gradual differences in the trait rather than on the clinical extremes (Spain et al., 2014).

Primary and secondary psychopathy can be measured by several instruments. Well known is the (revised) PCL two-factor model wherein the first dimension (Factor 1) assesses the core interpersonal and affective features of psychopathy (including grandiosity, lack of guilt, and callousness), and the second dimension (Factor 2) assesses an antisocial and impulsive lifestyle (Hare, 2003). As a self-report psychopathy scale, Levenson’s Self-Report Psychopathy Scale (LSRP; Levenson, et al., 1995) is often used. Here, the primary psychopathy subscale is theoretically aligned with PCL Factor 1 and the secondary psychopathy subscale is aligned with PCL Factor 2 (Lilienfeld et al., 2015; Smith & Lilienfeld, 2013). More current research proposes that the primary psychopathy subscale is comprised of an egocentricity and callousness dimension (see Brinkley et al., 2008; Sellbom, 2011).

It has been argued that those scoring high on primary psychopathy may more often be found in leadership positions—particularly upper management—than in general population samples (e.g., Boddy et al., 2021; Palmen et al., 2021). For instance, Howe et al. (2014), using a sample of 55 corporate finance workers in New York, found that 12.7% of them had a PPI-Fearless Dominance score—a clear marker of primary psychopathy (Lilienfeld et al., 2012) of two standard deviations above the mean, which is significantly higher than found in community samples. Also, Boddy (2011a), evaluating the correlation between workplace bullying and corporate psychopathy, found that of the 346 white-collar employees in the sample, 32.1% reported having a psychopathic manager at one point in their careers.

Self-serving Behavior and Abuse of Primary Psychopathy in Leaders

The negative impact that leaders with higher primary psychopathy levels have on the organization and all people in it may be quite severe. Indeed, those who tend to experience shallow affect, little or no remorse, guilt, or empathy, and a grandiose sense of self-worth (Miller et al., 2010) may often behave in a way that does not serve the organizations’ interest. Although in many studies on the effects of leader psychopathy, the multi-dimensionality of the psychopathic personality is not accounted for (making it unclear of any of the dimensions is particularly responsible for the results), the overall picture is that leader psychopathy by and large wields negative effects on organizational functioning. Higher psychopathy levels in employees have been linked to lower job performance, poor team orientation, excessive (financial) risk taking with other people’s money (e.g., mergers and acquisition), and increased counter productive work behavior (Babiak et al., 2010; Jones, 2014; Schilbach et al., 2020; Ten Brinke et al., 2018; Wisniewski et al., 2017). On top of that, corporate psychopaths are more likely to engage in illegal practices and unethical decision making (Stevens et al., 2012), and firm performance is often hampered by top managers scoring higher on psychopathic traits (Bouncken et al., 2020). In fact, Boddy (2011b) even argued that psychopathy levels of key executives may have contributed to the 2008 global financial crisis.

Beyond these negative organizational outcomes, primary psychopathy in leaders also likely negatively affects employees who work with them. Their deficiencies in affective experience (e.g., callousness, lack of empathy and feelings of guilt, egocentricity) may easily trigger behavior that damages the relationship with others. Several studies suggest that this indeed is the case, although—again—many of these studies also do not account for the multi-dimensionality of the psychopathic personality. For instance, employees who report having colleagues scoring high in psychopathy suffer from conflict, bullying, reduced job satisfaction, and increased workload and absenteeism (Boddy & Taplin, 2016; Boddy et al., 2015; Boddy, 2010b, 2015a, 2015b). Employees who have supervisors who score high on psychopathy report lessened personal consideration of those supervisors (Westerlaken & Woods, 2013) and not being allowed to speak up and voice opinions (Boddy, 2017).

Of particular interest to the current study are findings testifying to the link between leader (primary) psychopathy and abusive supervision and self-serving behavior. Abusive supervision, or the “sustained display of hostile verbal and nonverbal behaviors, excluding physical contact” (Tepper, 2000, p. 178), refers to subordinate targeted behaviors such as, lying, rudeness, coercion, humiliation, and public criticism (Tepper, 2007). In line with Wu and LeBreton (2011), these behaviors are often displayed to serve the leader’s self-interests at the expense of subordinates’ needs and the organization’s financial functioning (Tepper et al., 2006). One example is leaders who use their position to obtain benefits at the expense of others (e.g., taking time off while subordinates have to work overtime) or claim credit for subordinates’ work (see Rus et al., 2010). Followers often greatly suffer from the psychological abuse and self-serving behavior of their supervisor, with consequences ranging from increased levels of depression, emotional exhaustion and anxiety, lower job satisfaction and organizational commitment, to insomnia, problem drinking, and reduced satisfaction with life (e.g., Schyns & Schilling, 2013; Tepper et al., 2017). Boddy et al. (2015) found that subordinates of corporate psychopaths (measured by subordinates’ assessment of the resemblance to a leader with primary psychopathic traits) indicated to be subject to abusive tactics by their leaders. This finding corroborates an earlier study by Boddy (2011a) showing that the incidence and frequency of unfair supervision and bullying are more likely to occur in teams headed by corporate psychopaths. Similarly, Mathieu and Babiak (2016)—using the B-scan 360 (Mathieu et al., 2013)—and Lyons et al. (2019)—using the short Dark Triad measuring psychopathy, narcissism, and Machiavellianism (Jones & Paulhus, 2014)—also found that leader psychopathy predicts abusive supervision. Finally, Wisse and Sleebos (2016)—using the Dirty Dozen (Jonason & Webster, 2010)—observed a significant relationship between leader psychopathy and abusive supervision in teams (which was marginally significant when controlling for the other Dark Triad traits). Finally, Barelds et al. (2018) showed that primary psychopathy in leaders was related to their self-serving behavior in two studies. Thus, based on theory as well as on available empirical results, it seems that leaders who score high on primary psychopathy can exert a strong negative influence on their subordinates and may be more likely to engage in abusive supervision and self-serving behavior.

Given the host of negative consequences associated with having employees with higher psychopathy scores in the organization, scholars and practitioners have suggested that organizations should screen for psychopathy in their hiring practices, refrain from giving them power, or fire them (Cohen, 2016; Marshall et al., 2015; Wisse & Rus, 2022). Unfortunately, primary psychopaths perform exceptionally well on job interviews (Babiak et al., 2010), and they can be difficult to remove or demote once they entered the organization. Therefore, instead of focusing potential interventions on the employee, we suggest to focus them on the organization. We propose that organizations may install and enforce clear policies that neutralize the impact of corporate psychopaths once they have entered the organization (Bartels & Pizarro, 2011; Cohen, 2016; Patil, 2015).

The Wholesome Influence of Organizational Context

Social cognitive theory stipulates that people learn about what is permitted and what is not via situational affordances (Bandura, 1971). Central to this theory is that other people’s behavior provides insight into the boundaries of what is allowed. In other words, people scan and interpret their environment to understand which behavior is tolerated or appreciated in the pursuit of personal goals (Crick & Dodge, 1994; Witt & Spector, 2012). Moreover, according to the social information processing theory, individuals often develop attitudes about their surroundings that are in line with normative group behaviors and their typical consequences (Fine et al., 2010). For example, employees who notice that fraudulent behaviors will help you to get ahead in the organization without notice being taken or punishment being exacted will likely conclude that these behaviors are acceptable and may start to adopt these behaviors themselves.

The notion that the environment may affect the extent to which psychopathic traits are reflected in negative behavior has been put forward by several scholars. For instance, Schyns et al. (2018), drawing on trait expression theory (Christiansen & Tett, 2008), argue that “while a trait is unlikely to change, the ways in which and the frequency with which it is expressed can be altered” (p. 242) and they point to the environment as a means to change the manifestation of Dark Triad trait of employees. Cohen (2016) likewise noted that most studies only found a weak relationship between the Dark Triad traits and counterproductive work behavior, and argued that this is perhaps caused by the fact that studies have ignored some important organizational context moderators in this relationship. Likewise, organizations that reward competition may foster perceptions of negative interdependence, a general lack of concern for others’ outcomes, and bullying (Černe et al., 2014), arguably because it teaches employees that such behavior is fitting. Valentine et al. (2018) linked this idea to psychopathy and bullying in a sample of selling professionals, and showed that organizations should be mindful of the behaviors they reward and punish as they may inadvertently strengthen the relationship between psychopathy and bullying. In a related vein, Blickle et al. (2018) found that high levels of ascendency prospects as well as prospects for income increases moderated the relationship between psychopathy (the overall score, and in particular the meanness dimension, that is, tendencies toward excitement-seeking, callousness, cruelty, and predatory aggression) and consideration. Under both environmental conditions, psychopathic managers showed less considerate leadership toward their subordinates. We build on these ideas and propose three policies organizations may enforce to neutralize rather than stimulate the abusive supervision and self-serving behavior of leaders with higher primary psychopathy levels: clear rules and procedures, clear sanctions, and transparency of behavior.

First, clear rules and procedures set boundaries as to what is permitted. Importantly, they must explicate what is allowed as well as what is forbidden, as to reduce wiggle room to maneuver around the rules (Mulder et al., 2015). Clear rules and procedures may deter potential rule-breakers because they signal what is important to the organization and they may foster the expectation that an organization will actually enforce them when they are violated (Hochwarter et al., 2005; Parilla et al., 1988; Weber, 1947). For instance, employees report less observed unethical conduct when the company has formulated formal legal compliance programs (Weaver et al., 1999). Moreover, it has been found that specific rules deter, but general and vague rules foster, unethical behavior (Mulder et al., 2015). This is because clear rules and policies are more difficult to interpret creatively in one’s own favor compared to general and vague rules (Cohen, 2016; Mulder et al., 2015).

Second, clear sanctions show that rule-breaking actually carries consequences. In line with deterrence theory, sanctions increase the perceived probability of being caught and punished, making it more difficult to remain unnoticed and getting away with destructive behaviors (Buckley et al., 1998). Research shows that punishments are particularly effective for those who see breaking the rules as a viable action alternative (Wikström et al., 2011). To put it differently, punishments are less effective in shaping the behavior of those who would not act in a norm violating way in the first place. However, those who do intend to violate rules refrain from that behavior when risking sanctions.

Finally, transparency involves making the behavior of leaders more visible to subordinates and other employees. Managerial positions impart leaders with substantial discretion that can make it easier to conceal behavior (Hogan & Kaiser, 2005). One way to deal with this is to increase “felt accountability” by making behavior transparent. That is, increasing leaders’ perception that they are required to justify and explain their actions and decisions to others (Hochwarter et al., 2014). Indeed, one study demonstrated for instance that accountability (using a manipulation of work-report formatting to elicit accountability perceptions) can reduce overbilling practices of clients (Desai & Kouchaki, 2015). Related research on transparency of behavior shows that formal monitoring and registration of employee performance may help reduce corruption (Grimmelikhuijsen & Welch, 2012; Olken, 2007).

Importantly, we argue that these three policies—clear rules, clear sanctions, and transparency—are particularly effective for those leaders who score high on primary psychopathy. The main intent of these policies is to counteract situations where primary psychopaths feel themselves comfortable—ambiguous poly-interpretable situations. Indeed, it has been argued that those with psychopathic traits may profit from ambiguous situations (Padilla et al., 2007). Having clear policies with consequences diminishes this ambiguity and, therefore, hampers moral justification of unethical behavior and sets a clear benchmark for appropriate behavior (cf. Cohen, 2016). This all needs to be seen in light of common misconceptions that people may have about psychopathy. The archetypal psychopath is portrayed as insensitive to punishments and unstoppably reckless (Patrick, 2018; Walker & Jackson, 2017). Yet, we focus on subclinical manifestations of primary psychopathy—those individuals operate based on cost–benefit calculus (e.g., Baughman et al., 2014; Glenn et al., 2010; Stevens et al., 2012), and thus can be perfectly rational in their economic choices (Yamagishi et al., 2014). That is why we believe these policies can be effective, because making the cost disproportionately higher to the expected pay-off should deter leaders with higher primary psychopathy levels (see also Bartels & Pizarro, 2011). Thus, we argue that when costs loom larger than gains, primary psychopathic leaders may consider refraining from engaging in negative behavior. Also, primary psychopathy is associated with the ability to set agendas, and plan and follow long-term goals (Snowden & Gray, 2011). Thus, primary psychopathic leaders may be equipped with the ability to follow guidelines set by the organization, if they feel it may help them to reach their long-term goals. Finally, these three practices may often work in concert. That is, it is difficult for sanctions to be effective when the rules and procedures are vague and abstract (Cohen, 2016; Mulder et al., 2015). Similarly, no harm may be done according to the rules and procedures when they are vague, and there is nothing to be held accountable for when rule transgressions remain unnoticed due to a lack of visibility of behavior.

Hypothesis

The positive relationship between leader primary psychopathy and negative leader behavior (leader self-serving behavior and abusive supervision) will be weaker to the extent that the organizational context (clarity of rules, sanctionability of misconduct, and transparency of behavior) is stronger.

Overview of the Present Research

We conducted three studies to determine whether organizational context affects leader primary psychopathic traits and their self-serving and abusive behavior. In Study 1, an online experiment with business leaders (N = 91), we measured primary psychopathy, and varied the clarity of rules, sanctionability of misconduct, and transparency of behavior in a compound manipulation. Using an asymmetrical dictator game, we were able to assess leaders’ self-serving behavior at the expense of a fictitious subordinate with a behavioral measure. In Study 2, we collected multi-source survey data (N = 100 unique leader–subordinate dyads) using a questionnaire in which leaders rated their own levels of psychopathy as well as the organizational context (clarity of rules, sanctionability of misconduct, and transparency of behavior), and subordinates rated abusive supervision. In Study 3, we collected team data (N = 86 supervisors with N = 361 subordinates) using the same questionnaires as in Study 2.

Study 1

Method

Participants and Design

Hundred-and-eight leaders from the United States (42.9% women, Mage = 32.35, SDage = 9.52) took part in an online experiment and were randomly assigned to one of two conditions (Organizational context: weak vs. strong).Footnote 1 Seventeen participants who either took less than ten seconds to read the manipulation text or took considerably longer (> 3.5 SD) to continue to the next page were excluded from the analyses (see DeSimone & Harms, 2018), resulting in a final sample of 91 leaders. Leader primary psychopathy was added to the design as a continuous independent variable.

Leaders’ average work experience was 13.63 years (SD = 8.70), on average they supervised 15.45 employees (SD = 53.14), and most of the leaders worked in the field of business and finance (18.7%), human services (16.5%), or technology (15.4%). With regards to leaders’ education, 36.3% had a secondary education degree (high school), 50.5% had a bachelor’s degree, 5.5% had a master’s degree, 4.4% had an MBA degree, and 3.3% had a doctoral degree.

Procedure

Leaders were recruited through Amazon’s Mechanical Turk (MTurk) and were allowed participation when they held a supervisory position with a minimum of three direct subordinates. Previous research has demonstrated that MTurk data are at least as reliable as data collected using more traditional methods (Buhrmester et al., 2011), and that results derived from leader samples collected via MTurk are comparable to those derived from leader samples collected in the field (e.g., Van Houwelingen et al., 2017). Participating leaders first read a short description of the study, were informed about confidential treatment of their data, and gave their informed consent. The study consisted of two parts. In the first part, participants filled out some questionnaires measuring their levels of primary psychopathy. In the second part, participants played an asymmetrical dictator game with a fictitious other MTurk worker. Specifically, they were led to believe that they would be randomly paired with another MTurk worker and would be either assigned a leader or subordinate role. In reality, all participants were assigned the leader role. Prior to taking part in the game, participants were presented with the task instructions and our organizational context manipulation. Upon completion of the asymmetrical dictator game, participants answered questions that served as our manipulation check. Finally, participants answered demographic questions, were debriefed, thanked, and paid (including a bonus that equaled the number of cents [up to $0.60] they allotted to themselves during the game).

Asymmetrical Dictator Game

To assess leader self-serving behavior, we relied on the asymmetrical dictator game, in which egoism and selflessness are both plausible courses of action. In a dictator game, money is to be divided between two parties: an allocator and a recipient. The allocator is in charge of dividing the money and the recipient cannot reject the allocator’s offer (e.g., van Dijk & Vermunt, 2000). The game is asymmetrical in the sense that the allocator knows more than the recipient. This realistically reflects business decisions—those in the position to allocate (e.g., leaders) oftentimes have more information at their disposal than those in the subordinate position (Ackert et al., 2011).

Given that all participating leaders were assigned the leader role in the experiment (i.e., the allocator), they were in charge of dividing the money. Participants were informed that (a) they had a total of 60 dollar cents to divide between themselves and their subordinate (i.e., the recipient), (b) their bonus payment would follow a ‘60 minus the number of cents allocated to the subordinate’ formula, (c) their subordinate did not know how many cents were available for division, and (d) their subordinate could not reject the offer they made. The more money participants allotted to themselves, the more the balance tipped toward self-serving behavior at the expense of the alleged subordinate.

Organizational Context Manipulation

A compound manipulation was used to manipulate clarity of norms, sanctionability, and transparency. Specifically, we made clear to participants that, we, as researchers, were responsible for the set-up of the study, and found it important that participants consider the guidelines we set forth for distributing the money between themselves and their subordinate. These guidelines differed depending on the condition.

In the strong organizational context condition, participants were presented with clear norms, told that inappropriate conduct would be sanctioned, and were made aware of the visibility of their actions. Participants read the following guidelines: “In our way of doing things, we always want people to treat each other with respect and dignity. We would like you to do the same and distribute fairly. We will monitor how you distribute the money, and if in our opinion you did not distribute fairly and appropriately, we might intervene and lower your bonus payment.”

In the weak organizational context condition, participants were presented with unclear norms, told that inappropriate conduct would not be sanctioned, and were made aware that their actions remained invisible. Participants read the following guidelines: “In our way of doing things, we expect other people to do what they do best. We would like you to distribute the money how you see fit. Your distribution will remain anonymous, and your decision is final. What you decide to keep for yourself will be your bonus payment.”

Measures

Manipulation Check

The successfulness of the organizational context manipulation was assessed with seven items using a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). A sample item includes “It was made sufficiently clear to me how I should conduct myself appropriately toward others.” An average organizational context score was calculated with a higher score indicating a stronger perceived organizational context (M = 3.77, SD = 1.03, α = .86).

Primary Psychopathy

Leaders’ primary psychopathic traits were assessed with the validated Dutch translation (Barelds et al., 2018) of Levenson’s Self-Report Primary Psychopathy Scale (LSRPA; Levenson et al., 1995). The primary psychopathy scale has 16 items, including items such as “I enjoy manipulating other people’s feelings” and “For me, what is right is whatever I can get away with.” The scale has been validated with prison and nonprison samples and has good convergent reliability with other psychopathy measures (Sellbom, 2011). Participants indicated their agreement with the statements using a 4-point Likert scale (1 = disagree strongly, 4 = agree strongly). Reverse-scored items were recoded and all items were averaged into a single psychopathy score (M = 1.91, SD = 0.51, α = .89).

Leader Self-serving Behavior

The number of cents participants allotted to themselves at the expense of their subordinate in the asymmetrical dictator game comprised our behavioral measure of leaders’ self-serving behavior (M = 38.84, SD = 11.33).

Results

In all consecutive hierarchical regression analyses, we mean-centered leader primary psychopathy and dummy-coded organizational context (− .5 and .5 for low and high organizational context, respectively). In Step 1, we entered the main effects of the predictor variables into the analysis, and, in Step 2, we added the interaction term. All regression results are displayed in Table 1.

Table 1 Summary of regression analyses for Study 1
Full size table

Manipulation Check

A hierarchical linear regression on our organizational context score only revealed a main effect of the organizational context manipulation, b = 1.29, SEb= 0.17, t(87) = 7.40, p < .001, indicating that those in the strong organizational context condition (M = 4.38, SD = 0.65) perceived the context to be stronger than those in the weak organizational context condition (M = 3.10, SD = 0.95; all other ps > .51). Hence, these results show that the manipulation worked as intended.

Leader Self-serving Behavior

A hierarchical linear regression analysis on the number of cents leaders allotted to themselves to the detriment of their alleged subordinate revealed a positive relationship between primary psychopathy and the number of cents allotted to oneself, b = 5.72, SEb = 2.25, t(87) = 2.48, p = .02. We also found a main effect of organizational context, b = -5.74, SEb = 2.32, t(87) =  − 2.57, p < .01, revealing that leaders in the strong organizational context condition awarded less money to themselves (M = 36.25, SD = 9.48) than those in the weak organizational context condition (M = 41.72, SD = 12.59). Furthermore, in line with our hypothesis, we found the predicted primary psychopathy × organizational context interaction, b =  − 10.14, SEb = 4.42, t(87) =  − 2.30, p = .02. As expected, simple slopes analyses indicated that leader primary psychopathy was strongly positively related to self-serving behavior when rules were unclear, behavior unmonitored, and no consequences were attached to rule violation (i.e., lower levels of the moderator), b = 10.06, SEb = 2.94, t(87) = 3.42, p < .001. There was no significant relationship between leader primary psychopathy and self-serving behavior under clear rules, transparency of behavior, and introducing negative consequences of violating the rules (i.e., higher levels of the moderator), b = − 0.08, SEb = 3.30, t(87) = − 0.02, p = .98 (see also Fig. 1) (See Footnote 1).

Fig. 1
figure 1

Leader self-serving behavior as a function of leader primary psychopathy and organizational context in Study 1

Full size image

Study 2

The previous experimental study relied on a compound manipulation, in which we manipulated clear rules and procedures, the sanctionability of misconduct, and the transparency of behavior simultaneously. In Study 2, we differentiate between these three contextual factors to examine which factor exerts the strongest effects in attenuating the link between leader primary psychopathy and abusive supervision. Moreover, Study 1 relied on interaction with a fictitious other person. In Study 2, we aim to improve external validity by relying on a sample of leader–subordinate dyads to examine the interaction between leader psychopathy and organizational context on subordinate’s perceptions of their leader’s abusive supervision in a field setting.

Method

Respondents and Procedure

We collected multi-source data from 100 unique leader–subordinate dyads. A total of 176 dyads of Dutch subordinates and their direct leader were approached (response rate 56.82%). For study participation, we required both leaders and subordinates to work at least 20 h a week, and subordinates were required to have received supervision from their current direct leader for at least 3 months at the time of data collection. Leaders’ mean age was 40.17 (SD = 12.07; 50% women) and subordinates’ mean age was 26.20 (SD = 9.20; 59% women). Leaders’ mean work experience was 18.83 years (SD = 11.01), and on average held their current supervisory position for 7.86 years (SD = 6.73). On average, leaders worked 37.17 h a week (SD = 10.97), and supervised teams consisting of 22.01 subordinates (SD = 20.40). Subordinates’ mean work experience was 7.56 years (SD = 8.28). On average subordinates worked 26.75 h a week (SD = 10.48). The majority of the respondents worked in commercially oriented (service) organizations (e.g., commerce, catering industry, healthcare organizations, education, etc.).

A convenience-sampling method was used to collect the data. Undergraduate students used their personal network and that of acquaintances to approach leaders and their subordinates. Potential respondents were approached via e-mail, through phone calls, or face-to-face contact. We stressed the fact that participation was voluntary and that responses would be treated confidentially. Leaders and subordinates who were interested in participating in the study were asked to fill in the online questionnaire or the paper-and-pencil questionnaire without consulting their colleagues, subordinates, or leader. Respondents gave their informed consent prior to participation. Paper-and-pencil questionnaires could be returned by using an enclosed envelope that was either picked up or returned by mail. Because people often filled out the questionnaires during work hours, we kept the survey short. A coding system was used in order to match subordinates’ answers to those of their leader.

Measures

Psychopathy

Leaders’ primary psychopathic traits were assessed with the same scale as in Study 1.Footnote 2 Reverse-scored items were recoded and all items were averaged into a single primary psychopathy score (see Table 2 for means, standard deviations, and Cronbach’s alpha).

Table 2 Means, standard deviations, reliabilities, and intercorrelations for Study 2
Full size table

Organizational Context

The organizational context comprised three aspects. We used the scales developed by Kaptein (2008) to measure clarity of rules (5 items), sanctionability (5 items), and transparency (4 items). Sample items are “The organization makes it sufficiently clear to me how I should conduct myself appropriately toward others within the organization (clarity),” “In my immediate working environment, employees will be disciplined if they behave unethically (sanctionability),” and “If I do something which is not permitted, someone in the organization will find out about it (transparency).” Leaders indicated their agreement with the statements using a 5-point Likert-type rating scale (1 = strongly disagree, 5 = strongly agree). All 14 items were averaged to form a single organizational context score (M = 3.67, SD = 0.58). We also created scores for each subdimension (for clarity of rules M = 3.92, SD = 0.79; for sanctionability M = 3.31, SD = 0.74; for transparency M = 3.31, SD = 0.89; see also Table 2).

Abusive Supervision

Leaders’ display of negative leader behavior vis-à-vis their subordinates was assessed with 14 items from the Dutch version (Wisse & Sleebos, 2016) of Tepper’s (2000) Abusive Supervision Scale (e.g., “My supervisor ridicules me” and “My supervisor puts me down in front of others”).Footnote 3 Subordinates rated their leaders’ abusive supervision style on a 5-point Likert scale (1 = never, 5 = often), and all items were averaged into one abusive supervision score (see also Table 2).

Results

Preliminary Analyses

Table 2 shows the means, standard deviations, zero-order Pearson correlations, and Cronbach’s alphas for the study variables. We first ran separate confirmatory factor analyses for each measure to see how well each measure performed in terms of model fit. Given that a model with latent factors reflecting all our measured variables required estimating more parameters than there were data points, we established model fit for the dependent variable (abusive supervision) separately from our independent variables. For our dependent-variable model, the 1-factor model resulted in acceptable model fit, χ2(77) = 120.70, p < .001; CFI = .87, TLI = .85, RMSEA < .08, SRMR = .07 (cf. Browne & Cudeck, 1993; Hu & Bentler, 1999). For the independent-variable model, a 4-factor model with latent factors for psychopathy, clarity, sanctionability, and transparency provided best model fit, χ2(344) = 573.43, p < .001; CFI = .79, TLI = .77, RMSEA = .08, SRMR = .11), and fitted significantly better than a 2-factor model with all organizational context items on one latent factor (χ2(349) = 679.25, Δχ2(5) = 105.82, p < .001), or a 5-factor model with a higher order factor for organizational context loading on the separate three context factors (χ2(346) = 581.43, Δχ2(2) = 8.00, p = .02). These findings reveal that psychopathy is theoretically as well as empirically distinct from organizational context. Importantly, these results show that organizational context consists of three separate and distinct aspects (clarity, sanctionability, and transparency).

Hypothesis Testing

We predicted that the organizational context (clear rules and policies, punishments for violating rules, and transparency of behavior) would weaken the positive association between leader primary psychopathy and subordinate-rated abusive supervision. To test this moderation effect, we relied on Hayes (2013; model 1). The overall measure for organizational context did not significantly interact with leader psychopathy to predict subordinate ratings for abusive supervision, although the results of the simple slopes analyses are in line with our predictions (see Table 3). However, findings for clear rules and policies as a moderator were in line with our prediction: leader psychopathy was only positively associated with abusive supervision when rules were unclear (1 SD below the mean), but not when rules were clear in the organization (1 SD above the mean; see Table 3 and Fig. 2). Sanctionability and transparency did not affect the relationship between leader psychopathy and abusive supervision.

Table 3 Regression results for Study 2 for abusive supervision
Full size table
Fig. 2
figure 2

Abusive supervision as a function of leader primary psychopathy and clarity of rules in Study 2

Full size image

Study 3

The second study uncovered that only clear rules and procedures deterred the abusive supervision of leaders with higher primary psychopathy scores. This third study seeks to replicate this effect. We collected team data to rule out the possibility that individual variation in abusive supervision confounds the true link between leader primary psychopathy and abusive supervision.

Method

Respondents and Procedure

We approached 170 Dutch leaders and several of each leader’s direct subordinates. Ninety-eight teams participated in our study (58% response rate). After initial screening, we had to exclude 12 teams (in 6 teams only one subordinate completed the questionnaire and in 6 other teams consistent answering patterns resulted in zero variance; see also DeSimone & Harms, 2018), resulting in a final dataset of 86 teams (86 leaders and 361 subordinates). The number of subordinate ratings per leader ranged from 2 to 20 (M = 4.20, SD = 3.12). Thirty-one percent of the leaders were female (Mage = 39.73, SD = 11.77). On average, leaders worked for the organization for 9.58 years (SD = 8.27), had been employed in a supervisory position in their current organization for 6.35 years (SD = 7.23), and supervised their team for 5.82 years (SD = 6.11). Forty-seven percent of the subordinates were female (Mage = 31.18, SD = 11.98). On average, they worked for the organization for 5.48 years (SD = 6.65), of which for 4.36 years (SD = 5.30) in their current function, and for 3.61 years (SD = 4.23) within their current team. A total of 54.26% (SD = 23.70%) of all subordinates within a team completed our questionnaire. Leaders and their direct subordinates mostly worked in commercially oriented service organizations (e.g., shops, banks, cafes, restaurants, schools, healthcare organizations, etc.).

Graduate students recruited leaders by using their work environment, their personal networks, and by visiting local businesses. Leaders were approached via e-mail, phone, or face to face, and they were asked to recruit their subordinates to fill out the subordinate questionnaire. Envelopes with paper-and-pencil questionnaires were distributed in sets to employees and their direct supervisors. We relied on a coding system to match leader–subordinate data. Participants were asked to fill in the paper-and-pencil questionnaires without consulting their colleagues, subordinates, or supervisor, and to return the questionnaires in the enclosed envelope. This envelope was subsequently either picked up or returned by mail. Moreover, we stressed the fact that participation in the study was voluntary and that the data would be treated confidentially.

Measures

Primary Psychopathy

Leaders’ primary psychopathic traits were assessed with the same measure as in Study 2 (M = 1.87, SD = 0.49, α = .86).

Organizational Context

To measure organizational context, we relied on the same measure as in Study 2. All 14 items were averaged to form a single organizational context score (M = 3.67, SD = 0.58, α = .89). Again, we also created scores for each subdimension (for clarity of rules M = 3.94, SD = 0.79, α = .91; for sanctionability M = 3.56, SD = 0.70, α = .79; for transparency M = 3.54, SD = 0.69, α = .77).

Abusive Supervision

Leaders’ abusive behavior was rated by subordinates with the same measure as in Study 2 (M = 1.36, SD = 0.45, α = .90).

Results

Preliminary analyses

To account for nesting in our data (multiple subordinates share the same leader), we used Mplus for multilevel analyses (Goldstein, 2003). We first computed an intercept-only model for the dependent variable (abusive supervision) to test whether a multilevel model was warranted. Findings revealed significant random variation in the intercepts across leaders for abusive supervision (Wald = 1.99, p < .05). We then computed intraclass correlations and found that 15% of the total variability in abusive supervision resided at the leader level, showing that multilevel modeling is appropriate (Hofmann et al., 2000). To justify the aggregation of employees’ ratings of abusive supervision, we calculated the Rwg scores, and the ICC(1) and ICC(2) (Bliese, 1998). For abusive supervision, the ICC(1) was .19, the ICC(2) was .56 (F-ratio = 2.28, p < .0001), and the Rwg = 0.91 (SD = 0.15), which are all satisfactory.

Similar to Study 2, separate multilevel confirmatory factor analyses for each measure were conducted to see how well the items load on their respective latent factor. The dependent-variable model with abusive supervision (excluding item 15; see also Footnote 3) resulted in acceptable model fit, χ2(77) = 173.31, p < .001; CFI = .88, TLI = .86, RMSEA = .06, SRMRwithin = .06. For the independent and moderator variables model, a 4-factor model with primary psychopathy as one latent factor, and three separate factors for each organizational context factor (clarity, sanctionability, and transparency) fitted the data better, χ2(344) = 494.63, p < .001; CFI = .84, TLI = .82, RMSEA = .04, SRMRbetween = .09) than a 2-factor model with all organizational context items loading on one factor (χ2(349) = 630.03, Δχ2(60) = 311.84, p < .001). These analyses reaffirm that clarity of rules, sanctionability, and transparency are distinct constructs that can be empirically distinguished from each other, and that primary psychopathy is not only theoretically but also empirically distinct from organizational context factors. Notably, for RMSEA and SRMR, values less than .05 represent good fit, values of .05 to .08 represent moderate fit, and values of .08 to .10 represent adequate fit (Browne & Cudeck, 1993). For CFI and TLI, values of above .90 indicate good fit (Hu & Bentler, 1999). Our scores for CFI and TLI are slightly below that.

Hypothesis Testing

All results are presented in Table 4. We performed multilevel regressions with abusive supervision as the dependent variable. Prior to analyses, the independent variable and the moderators were grand-mean centered. We first performed a regression with leader primary psychopathy, the overall organizational context, and their interaction as predictors. This analysis revealed the same consistent pattern of results as in Study 2. Again, the interaction effect for the organizational context variable was not significant, although the simple slopes are in the hypothesized direction. Further in line with our predictions and Study 2, there was a significant psychopathy × clarity of rules and policies interaction effect with simple slopes indicating that psychopathy was only positively linked to abusive supervision when clarity of rules and policies were weak (1 SD below the mean), but not when they were strong (1 SD above the mean; see Table 4). There was no significant interaction between leader primary psychopathy and sanctionability, or between leader primary psychopathy and transparency.

Table 4 Regression results for Study 3 for abusive supervision
Full size table

General Discussion

We predicted that three organizational context factors—clear rules, appropriate punishments for those who violate the rules, and transparency of behavior—would weaken the positive relationship between leader primary psychopathy and negative leader behavior (leader self-serving behavior and abusive supervision). The first study (Study 1) showed that a manipulation of all three contextual factors inhibited the self-serving behavior of leaders with higher primary psychopathy scores. Our dyads data (Study 2) and team data (Study 3) showed that only clear rules deterred leaders’ abusive supervisory behaviors. Thus, we consistently showed that clear rules and procedures deterred the negative behavior of leaders with higher primary psychopathy scores, but that punishments of transgressions and transparency of behavior did not exert similar consistent effects.

Theoretical Implications

As described by the holistic perspective on destructive leadership (Thoroughgood et al., 2018), negative leadership processes are rarely the result of individual leaders alone. Instead, leader characteristics, such as leader primary psychopathy, are assumed to only translate into negative outcomes if followers assist or are unable to resist destructive leader behaviors, and if the organizational environment is conducive (i.e., enables or supports these behaviors). This paper focuses on what the organization can do to stop from being conducive by highlighting possible organizational-level interventions aimed at tackling destructive influences of leaders scoring high on primary psychopathy. As such, the results of our study offer support to the holistic perspective on destructive leadership (Thoroughgood et al., 2018) by showing that the organization can formulate clear rules and procedures in order to mitigate the negative behaviors of leaders with higher levels of primary psychopathy who have already entered the organization.

Combining insights from social cognitive theory (Bandura, 1971) and social information processing theory (Fine et al., 2010) with primary psychopaths’ cost–benefits perspective (Bartels & Pizarro, 2011; Patil, 2015), we predicted and showed that the behaviors of leaders with higher levels of primary psychopathy may be positively affected by creating organizational contextual cues. We consistently showed that clear rules and procedures weaken the positive relationship between leader primary psychopathy and abusive supervision. One theoretical explanation for this result is that primary psychopaths simply do not know how to behave in that they lack the capacity to judge an act to be morally wrong (they suffer from “moral blindness”). As such, they need clear rules and regulations to inform them about appropriate behavior (i.e., behavior that—over time—may help them to achieve their self-interested goals, like being promoted or obtaining a pay raise). Indeed, a similar observation was made in the context of psychopathic offenders by Fine and Kennett (2004) who argued that while psychopaths do not appear to have the ability to assess the morality of behavior by themselves, they certainly appear to know what acts are prohibited by society or the law (and therefore know that their transgressions are legally wrong). In support of this argument, it has been argued that people who are antisocial often have impairments in brain areas that subserve moral decision making and that people with primary psychopathy have the most neuro-moral impairment of all (Raine, 2019). As such, our results may point to the notion that individuals high on primary psychopathy are amoral (because of their lack of empathy they do not know what is good or bad): Letting them know what is proper conduct may therefore make a difference, particularly if it helps them to further their self-interest.

Also important in this respect is the notion that primary psychopathy is not strongly associated with impulsive or disinhibited behavior (Drislane & Patrick, 2017). Indeed, those scoring high primary psychopathy can be quite deliberative in their actions and can set agendas, and plan and follow long-term goals (Snowden & Gray, 2011). Our findings likewise suggest that primary psychopathic leaders may be equipped with the ability to follow rules and procedures set by the organization. However, it may be important for them to have the feeling that following those organizational rules if they feel it may help them to reach the goals that they find important. At least, such is suggested by research showing that primary psychopaths’ sensitivity to rules related to rewards and punishment is high to the extent that these rules coincide with their dominant focus (Wallace & Newman, 2008). That is, primary psychopaths respond to rules as long as doing so fits with what is important to them at that point in time. This suggests that organizations may want to make certain that following rules and procedures also has positive consequences.

Interestingly, sanctionability of misconduct and transparency of behavior did not consistently exert similar effects. We think this may be explained by the idea that clear rules and procedures are a prerequisite for sanctionability and transparency of behavior to be effective. When rules and procedures are not clearly defined, threats of punishments for undesirable behavior may not be effective because when rules can be stretched it is less clear when misconduct is taking place (cf. Mulder et al., 2015). Likewise, when behavior is transparent and easily observable by other colleagues it may not deter leaders’ behavior when they do not perceive that they have violated the general rules in the first place. Taken together, our findings may be explained by considering that sanctionability of misconduct and transparency of behavior are only effective to the extent that rules are so clearly defined that leaders with higher primary psychopathy scores know when they have or have not violated those rules. Indeed, Cohen (2016) similarly argued that clear rules are a requisite for effective punishments as the existence of the rule announces that violation of the rule has consequences.

Finally, our research may also have some theoretical implications for the conceptualization of psychopathy. In our approach we have made use of the distinction between primary and secondary psychopathy. However, recently the triarchic model of psychopathy (Patrick & Drislane, 2015) is often used. This model suggests that psychopathy comprises three distinct constructs: (1) disinhibition, that is, tendencies toward impulsiveness and a substantial difficulty regulating one's affect urges; (2) boldness, that is, high dominance, the ability to remain calm in stressful situations, and venturesomeness; (3) meanness, or the tendency toward excitement-seeking, cruelty, detachment from others, and predatory aggression. It has been found that boldness overlaps mainly with primary psychopathy (or the interpersonal and affective facets in the PCL-R), that disinhibition overlaps with secondary psychopathy (the antisocial and impulsive lifestyle facets), and that meanness has commonalities with both factors (see Sleep et al., 2019 for a meta-analyses). It would be interesting if future research would assess the moderating effect of clarity of rules, sanctionability of misconduct, and transparency of behavior on the dimensions of the triarchic model to investigate to what extent similar effects for the boldness and meanness facets can be found.

Practical Implications

The main focus of this research is to provide organizations with effective tools to deter the negative behavior of leaders with primary psychopathic traits, specifically their self-serving and abusive behavior. We provide three concrete policies that organizations can use to their advantage to mitigate the negative behavior of leaders with primary psychopathic traits. Most importantly, clear rules and procedures should be formulated. That is, to effectively exploit the cost–benefit orientation of those with higher levels of primary psychopathy, there should not be room to interpret the rules and procedures creatively. Preferably, these rules describe which behaviors are allowed and which behaviors are prohibited; the rules should be clear and not open to interpretation; and the rules should be clearly communicated and stressed (see Treviño et al., 2003). The provision of clear rules is a measure that should have positive effects on the behavior of other employees too, making it a policy with few potential disadvantages. Second, application of appropriate punishments for the violation of rules and increasing the transparency of behavior might be relevant to the extent that clear rules and procedures are in place.

Strengths and Limitations

Like most studies, the present set of studies is not without its limitations and the findings should be interpreted in light of these limitations. We opted for a multiple-study multiple-method approach, so that the shortcomings of specific study designs would be offset by the strengths of other designs (cf. Scandura & Williams, 2000). For instance, in Study 1 we tested the interactive effects of leader primary psychopathy and organizational context using a behavioral measure. However, a drawback of this first study was the relatively artificial nature of the design, and that we relied on a compound manipulation for organizational context—meaning that we could not disentangle individual effects of clarity of rules, sanctionability, and transparency. Study 2 and 3 addressed these limitations. In Study 2 (a survey of dyads) and 3 (a survey of teams), we gathered data from leaders and subordinates in organizations, and tested the interactive effects of leader primary psychopathy and each organizational context variable separately. In addition, by relying on field studies, external validity concerns were alleviated. Notably, the reliance on multi-source data in both Study 2 and 3 can be considered a strength, as common source bias (Podsakoff et al., 2003) cannot explain our findings. Notably, we did not perform power analyses for our three studies. The sample sizes for all three studies were modest, creating the probability that our results are somewhat deflated (high Type 2 error rates) or that estimates are less accurate (Schönbrodt & Perugini, 2013). Also, while we did omit some participants in Study 1 based on response latency to improve data quality, we did not use attention checks in Study 2 and Study 3. However, researchers are not unequivocal in their recommendations with regard to their use. Indeed, some have argued that they cause respondents to give lower-quality responses later in the survey and may induce Hawthorne effects, and that excluding participant based on their answers to these checks does not alter the results in a meaningful way (Clifford & Jerit, 2015; Gummer et al., 2021). Yet, despite these potential limitations, we believe that the combination of these three studies, replicating the results using different designs, boosts confidence in our findings.

Across all three studies, we relied on a self-report measure of leader primary psychopathy. Some have argued that self-report psychopathy measures may be affected by psychopathic individuals’ deliberate attempt to underreport their psychopathic features (cf. O'Boyle et al., 2012). Yet, this appears to be a misconception as several studies have shown that self-report psychopathy measures are reliable (Levenson et al., 1995) and have shown to be negatively related to indicators of social desirability and positive impression management (Lilienfeld & Fowler, 2006; Verschuere et al., 2014) (See Footnote 1). Still, it would be interesting to see if our results would be replicated if observational measures or subordinate ratings of primary psychopathy would be used.

Our measurement of leader primary psychopathy poses another potential limitation. That is, we only assessed primary psychopathy and not secondary psychopathy (Levenson et al., 1995). Although this seems to make sense from the notion that primary psychopathy does not hinder individuals from functioning reasonably well in society (contrary to secondary psychopathy) and therefore seems to be the more relevant dimension for studies conducted in the work place context, future studies might, however, also examine secondary psychopathy. It would be interesting to see if clarity of rules would indeed have less effect on the behavior of those scoring (also) high on this dimension (which may be the case due to increased impulsivity). Moreover, future research may also consider using other scales. For instance, one of the most commonly used measures of psychopathy is the Self-Report Psychopathy scale (Hare, 1991). We opted against using that scale because of its length (the current version of the scale, the SRP-III, has 64 self-report items) which would have likely suppressed the response rate of our leader sample. However, for reasons of comparison, future research may try to replicate our findings using the SRP.

A final potential practical limitation of our research is related to whether our model applies to primary psychopathic leaders functioning at the top level (presidents, chief executive officers, managing directors, business owners, etc.). Due to their position power, these leaders might be untouchable, able to fudge or change rules and procedures if they are not to their liking, or hide misbehavior from view (see Boddy, 2015b, 2016). Moreover, those with high power may be less likely to pay attention to rules and procedures and may be less likely to behave in accordance to them. Indeed, it has been argued that power makes people less responsive to situational information (e.g., rules and procedures) than low-power individuals because power increases sensitivity to internal states (Galinsky et al., 2008). Thus, as power increases the concordance between their internal beliefs, states, and traits on the one hand and behavior on the other, the neutralizing effect of clear rules and procedures might be dampened by the extent to which primary psychopathic leaders have position power (Wisse & Sleebos, 2016).

Directions for Future Research

One important research question that requires further attention is what employees do in response to being exposed to the self-serving or abusive behaviors of leaders with higher primary psychopathy scores. One insight may be derived from the trickle-down model of abusive supervision, which showed that employees exposed to abusive supervision start behaving to abuse themselves (e.g., Mawritz et al., 2012). A key insight from this literature is that employees typically do not retaliate against the powerful because they fear the consequences of doing so. This fear may be particularly strong if they are led by leaders scoring high on psychopathy (cf. Boddy & Taplin, 2016; Boddy et al., 2015; Boddy, 2015a, 2015b). Instead, employees may displace their discontent with how they are treated by their leaders on others, such as their fellow employees, creating a hostile culture. Alternatively, they may engage in increased acts of counterproductive work behavior (Aquino et al., 1999).

Moreover, it would be valuable if future research would focus on the potential interactive effects of clear rules, sanctionability of misconduct, and transparency of behavior. Our experimental study employed a compound manipulation in which we focused on the combination of all three together, and we did not manipulate them separately. Also, our field samples were too small to test if certain combinations of organizational context variables may be particularly suited to influence the leader primary psychopathy–negative behavior relationship, which might be the case. The collection of large datasets may be helpful in this regard.

In terms of outcomes, it would also be interesting to see whether a field study that includes and intervention that incorporates the three contextual tools would indeed lead to improvements in terms of leader behavior (self-serving behavior, abusive supervision), organization performance (net profit), and employee perceptions and behavior (satisfaction, counter productive work behavior). Indeed, attempts at treating psychopathy by trying to change the person are not very successful, but changing the risk factors can lead to reductions in displays of psychopathy (cf. Polaschek & Skeem, 2018). Notably, this is in line with trait expression theory (Christiansen & Tett, 2008) that postulates that while a trait (like psychopathy) is unlikely to change, the way in which and the frequency by which it is expressed can be altered by the social context (see Schyns et al., 2018). Similarly, an intervention of organizational context would not likely change primary psychopathy ratings per se, but it would—so we argue—diminish the displays of behaviors linked to primary psychopathy, such as abusive supervision and self-serving behavior.

Conclusion

In sum, the present research provides insight into what organizations can do to tackle the negative behaviors of leaders with higher primary psychopathy scores once they are present in the organization. Particularly the formulation and communication of clear rules and procedures seem to deter such leaders from displaying self-serving behavior and abusive supervision. Fortunately, we showed that there is hope for organizations that want to protect themselves against primary psychopathic leaders’ destructive influence.