Abstract
Drawing on the findings from a multiple case study, we build a process model of toxic leadership and empirically illustrate the toxic leadership process and its outcomes. In doing so, we make two important contributions to current literature. First, we provide a theoretical model of the toxic leadership process that details the intent and outcomes of toxic leadership relative to other dark leadership styles: destructive, abusive, and ineffective. In doing so, we show that, although the intent of toxic leaders is relatively less harmful, the behaviors these leaders engage may be harmful for the organization. More specifically, we show that the primary intent of toxic leaders is to conceal lack of relevant competence and maintain a position of control, at the exclusion of other organizationally relevant objectives. To achieve their intent, toxic leaders engage in upward and downward directed influence attempts that create ambiguity and confusion, thus increasing the toxicity of the context and interfering with others’ ability to perform their work. Second, our findings suggest that followers are more agentic than previously suggested. We show that followers not only choose to unfollow their leaders but also actively work to neutralize the influence of toxic leaders through workarounds and learning. Followers have largely been neglected by the extant leadership literature, and thus, the agency they often display in leadership processes has been insufficiently examined.
Recent events such as abuses of authority (e.g., the forceful removal of a passenger from a United Airlines aircraft), unethical behaviors of those in leadership positions (e.g., unauthorized account openings at Wells Fargo), and hostile work environments (e.g., the diversity crisis at Uber), just to name a few, have demonstrated the pressing need to analyze hostile and unethical environments in organizations. Recent literature suggests that leadership may play an important role in building and maintaining such environments (Padilla, Hogan, & Kaiser, 2007; Pelletier, 2010) via facilitation of counterproductive (Krasikova, Green, & LeBreton, 2013) and unethical work behaviors (Hannah et al., 2013; Kellerman, 2004). The underlying assumption in this line of research is that leaders may (1) have considerable power to leverage organizational resources and influence others in a destructive manner, (2) lead them toward goals that are destructive for the organization (Ashforth, 1994; Conger, 1990; Krasikova et al., 2013; Schyns & Schilling, 2013), or both.
In light of this, researchers have begun to inquire into the dynamics of various dark leadership behaviors such as destructive (Einarsen, Aasland, & Skogstad, 2007; Krasikova et al., 2013), abusive (Tepper, 2000; Tepper, Moss, Lockhart, & Carr, 2007), narcissistic (Grijalva & Harms, 2014; Rosenthal & Pittinsky, 2006); and toxic (Lipman-Blumen, 2005; Pelletier, 2010), among others. The consensus among this research is that leaders who engage in these behaviors (Einarsen et al., 2007; Pelletier, 2010; Tepper, 2000; Thoroughgood, Tate, Sawyer, & Jacobs, 2012) will have a largely negative impact on the organization and its stakeholders (Krasikova et al., 2013; Padilla et al., 2007; Schilling, 2009). However, a degree of vagueness with regard to the nature of dark leadership and the process of leading seems to persist (Grijalva & Harms, 2014; Krasikova et al., 2013; Schilling, 2009), particularly in comparison to studies of different “good” and “effective” leadership styles (Paterson, Luthans, & Milosevic, 2014; Shaw, Erickson, & Harvey, 2011). For example, the following questions seem to remain: Are all dark leaders equally destructive? Do all dark leaders have equal intent to engage in dark leadership behaviors? Do all dark leaders engage in the same dark behaviors, or are some dark behaviors more associated with abusive leaders and less so with toxic leaders? This is problematic because, similar to the variety of “good” leadership styles and their associated host of positive outcomes, dark leadership styles not only matter but may manifest in organizations differently and, therefore, may result in different outcomes.
Therefore, the purpose of this study is to explore toxic leadership, as one type of dark leadership style, and provide insight into the toxic leading process using a multiple case study methodology (Creswell, 2012; Eisenhardt, 1989). Previous studies in the realm of toxic leadership have been largely grouped under dark leadership (Lipman-Blumen, 2005; Padilla et al., 2007), where the terms toxic and destructive are used interchangeably among other dark leadership styles. For example, Padilla et al. (2007) argue that the intersection of susceptible followers and toxic contexts enable destructive leaders to stay in position longer and engage in destructive behaviors. Similarly, Lipman-Blumen (2005) identifies leaders who are intentionally toxic (i.e., engage in destructive influence) and “deliberately harm others or enhance themselves at other’s expense” (p. 29). However, Lipman-Blumen (2005) also suggests that leaders may be unintentionally toxic and cause harm through their recklessness and incompetence. These somewhat conflicting insights render that the nature of toxic leaders is largely unclear.
Building on the previous literature on toxic leadership and the emergent insights from our findings, we define toxic leadership as leadership focused on maintaining position of control via toxic influence attempts (Hornstein, 1996; Pelletier, 2010), whose harmfulness, although relatively unintentional (i.e., relatively low intent to cause harm), “cause[s] serious harm by reckless behavior, as well as by their incompetence” Lipman-Blumen (2005, p. 29). We utilize multiple case study methodology as an appropriate methodology to explore the nature of toxic leadership because it allows rich inductive insight into a phenomenon that is relatively ambiguous (Creswell, 2012), creates space for process theorizing needed in the leadership field (Denis, Langley, & Sergi, 2012; Howell & Shamir, 2005), and provides added opportunities for triangulation and generalizability (Eisenhardt, 1989).
Our study has two important theoretical contributions. First, in explicating the toxic leadership process, we show that the primary intent of toxic leaders is to shield their lack of competence and maintain the position of control via upward influence attempts (i.e., toward superiors), such as ingratiation and selective information sharing, as well as via downward influence attempts (i.e., toward subordinates), such as limiting interaction and micromanagement of followers. These activities are toxic because they create a state of ambiguity where employees (both the subordinates and the superiors) have difficulty evaluating the competence of the toxic leader. As a consequence, these leaders stay in position longer, further increasing the toxicity of the context via political behavior and bullying (Baillien & De Witte, 2009; Griffin, & O’Leary-Kelly, 2004).
Second, our findings contribute to the literature on followership (Carsten, Uhl-Bien, West, Patera, & McGregor, 2010; Uhl-Bien, Riggio, Lowe, & Carsten, 2014). We show that, although followers engage with toxic leaders in a varied manner, they are not passive participants in the toxic leadership process (Padilla et al., 2007). Rather, our findings show that followers display considerable agency in effort to neutralize toxic influence. More specifically, we show that despite feeling powerless at times, followers act to shield their work from the negative impact and neutralize, or at least minimize, the toxic leadership via workarounds and learning. Our overreaching contribution, thus, is a process model of toxic leadership that is grounded in both theory and data.
Theoretical Context
The Nature of Dark Leadership
Dark leadership has received heightened attention in recent years despite the lack of accepted definitions (see Krasikova et al., 2013, for an important exception in the realm of destructive leadership) and a unified theoretical framework (Tepper, 2007). In an early article, Tepper (2000) argued that abusive supervision entails a sustained display of hostile verbal and nonverbal behaviors, excluding physical contact, by superiors toward their subordinates. Following his seminal work, interest in dark leadership (destructive, toxic, abusive, etc.) has grown, albeit with less precise definitions. One reason, perhaps, for slow development of this line of research as well as general focus on abusive supervision rather than leadership, has been reluctance of leadership scholars to recognize dark aspects of leadership and the potential harmfulness of its outcomes (Howell & Avolio, 1992).
However, more recently, there has been a concentrated effort to build theoretical models of dark leadership (Einarsen et al., 2007; Krasikova et al., 2013; Padilla et al., 2007) as well as inquire into specific behaviors (Ferris, Zinko, Brouer, Buckley, & Harvey, 2007; Pelletier, 2010; Thoroughgood et al., 2012) and outcomes (Schyns & Schilling, 2013; Thoroughgood, Hunter, & Sawyer, 2011). For example, Einarsen et al. (2007) introduced a model of destructive leadership that encompasses three specific categories of destructive leading behaviors: tyrannical, derailed, and supportive–disloyal. More recently, Krasikova et al. (2013) defined destructive leadership as volitional behavior by a leader that can harm or intends to harm a leader’s organization and/or followers by (a) encouraging followers to pursue goals that contravene the legitimate interests of the organization and/or (b) employing a leadership style that involves the use of harmful methods of influence with followers, regardless of justifications for such behavior. (p. 1310)
Despite these insights, the ambiguity with regard to toxic leadership and its outcomes relative to other dark leadership styles, such as destructive, seems to persist.
The Theoretical Framework: Intent and the Outcomes of Influence
To more clearly understand the boundaries of toxic leadership and its process of influence, we build on the work of Krasikova et al. (2013), Padilla et al. (2007), and Schilling (2009) and theorize that dark leadership styles may be usefully differentiated based on (1) the leader’s intent to create harm and (2) the degree of harm that followers and organizations experience. According to Krasikova et al. (2013), intent is evident in volitional choices leaders make about what goals to pursue and how to achieve them. Building on Krasikova et al.’s (2013) arguments, we further theorize that leaders who are highly competent and experience higher levels of narcissism together with the ability to set destructive goals manifest the strongest intent to create harm. On the other end of the spectrum, leaders with low competence and relatively low aptitude to lead are less likely to be intentional in their actions due to inability or unwillingness to set goals, destructive or constructive, and influence others toward achieving them (Antonakis, Avolio, & Sivasubramaniam, 2003; Toor & Ogunlana, 2009).
We further theorize that leaders with the strong intent to harm others as described above are also likely to engage in actions that will be highly harmful for the followers and organization. They may leverage their competence to manipulate others in fulfilling goals that harm the organization or engage in destructive actions to shield their self-esteem (Grijalva, & Harms, 2014; Kernberg, 1975). However, the intent of the leader may not be to pursue the destructive goals, but the influence tactics they use are destructive in that they create considerable harm for others (James & LeBreton, 2010; Tepper, 2000). For example, frequent emotionally abusive outbursts characteristic of abusive leaders are less intentional (i.e., the intent to induce harm is less calculated), but the harmfulness for the followers is considerable (Tepper, 2000).
This line of theorizing presents a useful framework to understand dark leadership because it delineates different ways in which dark leaders may affect the organization. On one end of the spectrum, the higher the intent to induce harm and the more harmful influence tactics are, the more destructive leadership style is for the organization and employees (Krasikova et al., 2013). On the other end of the spectrum, lack of intent to pursue destructive goals and relative benevolence of influence tactics the more ineffective leadership style is for the organization and employees resulting in less optimal performance (Chan & Drasgow, 2001; Foti & Hauenstein, 2007). For the purpose of parsimony and to more clearly illustrate the nature and boundaries of toxic leadership, rather than engage in extensive theorizing on dark leadership, we endeavor to differentiate among four types of dark leadership styles: destructive, abusive, ineffective, and toxic leadership using the above framework (see Figure 1 1 ). However, we do recognize that there are other dark leadership styles (e.g., autocratic, bullying, tyrannical; Pelletier, 2010) that future research should consider.

Situating toxic leadership.
Destructive leadership
Padilla et al. (2007) postulated several key features of destructive leadership. They hypothesized that it is a construct that yields both positive and negative outcomes for organization and followers, although it is the negative ones that dominate. The positives are embedded in leaders’ often charismatic appeal that enables them to quite effectively motivate action (Grijalva, & Harms, 2014; Krasikova et al., 2013). More specifically, due to the high levels of narcissism (Krasikova et al., 2013), leaders seem charming and engaging on the surface, thus attracting some followers. However, below the surface, these leaders are exploitative and quite ruthless (Grijalva, & Harms, 2014; Kernberg, 1975). Given their ability to attract followers and their propensity for destructive behaviors, the harmfulness of their action is considerable. These leaders are perceived as highly competent and able to influence others to pursue goals that are destructive for the organization (Krasikova et al., 2013). More specifically, Roberts and Robins (2000) suggest that destructive leaders can set self-serving goals that are contrary to organizational objectives and be quite successful in influencing others to achieve them.
However, unlike abusive leaders discussed below, destructive leaders are premediated in their intentions to do harm. In other words, their competence and high levels of narcissism fuel calculative and intentional advancement of destructive goals (Krasikova et al., 2013). Consequently, unlike abusive leaders, these individuals are considerably more manipulative and calculative in their approach to fulfilling their goals and are, thus, more likely to intentionally inflict harm on their path to goal attainment (Hare, 1999; Wu & LeBreton, 2011) resulting in considerably high harmfulness of their actions. The convergence of selfish goals and utter disregard for welfare of other stakeholders, amplified by the existence of the conducive environment, may enable destructive leaders to act in truly vile ways. Thus, the key differentiating factor is the very high intent to harm, either via deliberate destructive influence methods or by successfully influencing others to pursue destructive goals (Krasikova et al., 2013). It is no surprise that focus on causing harm in fact leads to most damaging outcomes for most stakeholders, except the leaders themselves.
Abusive supervision
Tepper (2000) defined abusive supervision as subordinate’s perception of the degree to which supervisors engage in the display of hostile verbal and nonverbal behaviors, not including physical contact. Abusive supervision comprises behaviors that are both intended to inflict harm for personal gain, as well as behaviors that could be considered as a form of indifference, rather than actions undertaken to directly damage organizations and/or its employees. Leaders who display abusive behaviors also tend to have high levels of narcissism but are less calculated in their behavior (Penney & Spector, 2002). More specifically, these individuals have relatively high self-esteem and are sensitive to threats to their self-esteem such as questioning their actions or providing less positive feedback on their performance (Grijalva, & Harms, 2014). When this occurs, these individuals are more likely to experience anger and frustrations, which results in them lashing out on others and engaging in generally abusive behaviors (Kernis & Sun, 1994).
Consequently, the intent to cause harm is somewhat variable, stemming more likely out of leaders’ inability to manage emotions resulting in an inappropriate response (Vazire & Funder, 2006) rather than from a calculated intention to induce harm. Indeed, abusive behaviors are associated with heightened emotions (Anderson & Bushman, 2002) and thus tend to be more spontaneous (less intentional), harder for the leader to manage (Baumeister & Heatherton, 1996), and likely to occur in the presence of witnesses (Grijalva, & Harms, 2014). The lack of awareness of the social impact combined with significant damage they produce places abusive supervision lower in intent than destructive leadership on our conceptual scale but on equal footing regarding the amount of damage caused. Irrespective of the ignorance regarding abusive actions, their effects are profound and can be enormously damaging.
Ineffective leadership style
Ineffective leaders are often described as laissez-faire leaders who, although still occupying the leadership position, tend to avoid making decisions and taking responsibility for the work (Antonakis et al., 2003; Toor & Ogunlana, 2009). These leaders avoid the responsibility of leading that often manifests as a general inaptitude and/or low motivation to lead (Chan & Drasgow, 2001; Foti & Hauenstein, 2007). As a consequence, these leaders do not tend to engage in calculative behaviors or self-promotion. However, Frischer and Larsson (2000) suggest that the harmfulness is evident in their lack of initiative and failure to provide important information to their followers. Skogstad, Einarsen, Torsheim, Aasland, and Hetland (2007) also suggest that these leaders are less likely to provide support for followers who are attacked by others, leading to job dissatisfaction, increased job stress, and overall low efficiency.
Krasikova et al. (2013) argued that although ineffective leadership is indeed harmful as illustrate above and hence categorized as dark here, the key distinction between other dark leadership styles and ineffective leadership is the intent to do harm or, in this case, lack thereof. More specifically, they argue that the volitional nature of harm-doing underlying destructive leadership is a key demarcation line that sets it apart from ineffective leadership (Krasikova et al., 2013, p. 1314). As such, unlike destructive or abusive leaders who are associated with active and highly visible harmful behaviors, these leaders tend to be passive and inflict indirect harm (Skogstad et al., 2007). Nonetheless, despite the low intent to cause harm, the potential harmfulness of their actions (or inactions) is evident (Schyns & Schilling, 2013). Indeed, Einarsen et al. (2007) argue that ineffective leaders violate the objectives of the organizations via inefficiencies and poor performance that weaken the motivation and well-being of employees. To this end, ineffective leadership may be portrayed as having relatively low intent to induce harm and, thus, moderately harmful influence overall.
Toxic Leadership
Although previous research has, to a degree, treated toxic and destructive leadership interchangeably, important distinctions seem to exist. Pelletier (2010) evaluated behaviors associated with toxic, destructive, abusive, tyrannical, bullying, and laissez-faire leadership to determine areas of overlap and to build clarity. In doing so, Pelletier (2010) recognized that although overlap does seem to be prevalent, differences do emerge. For example, differentiating behaviors that toxic leaders engage in include efforts to separate people and limit productive interactions, such as placing obstacles to their work (i.e., harmful influence tactics), rather than intentionally harm or lead others toward destructive goals as destructive leaders may do (Krasikova et al., 2013). Further examples might include leaders going behind others’ backs to achieve their goals, intentionally withholding important information, criticizing follower’s work in demeaning manner, and making subjective assessment about their abilities and their work (Pelletier, 2010).
Consequently, toxic leaders seem to interfere with other’s ability to perform work (similar to the way poison may interfere with individual’s ability to function, therefore being considered toxic) rather than successfully lead followers toward destructive goals as destructive leaders do or emotionally and physically abuse them as abusive leaders do. Furthermore, unlike ineffective leaders, toxic leaders are concerned with the position of control and act to protect that position. Consequently, despite the lack of strong intent to induce harm, the harmfulness of toxic leaders is higher relative to ineffective leaders but somewhat less harmful compared with destructive and abusive leaders whose intent and harmfulness of influence are quite considerable.
Lipman-Blumen (2005) identifies leaders who are intentionally toxic and deliberately harm others (i.e., have high intent to cause harm) and unintentionally toxic and harm others through their recklessness and incompetence (i.e., have low intent to cause harm but relatively harmful influence tactics). This suggests that depending on the type of toxicity, these leaders may be more or less destructive. Based on past theorizing, the level of destructiveness and intent to cause harm may be viewed as a function of relevant competence coupled with narcissism of the leader that fuels deliberate engagement in harmful activities (Grijalva & Harms, 2014; Krasikova et al., 2013). This may act as an important differentiating factor between destructive toxic leaders (what we refer to as destructive leaders to enhance clarity) and toxic leaders (a concern in this study). More specifically, dark leaders with higher competence and higher levels of narcissism are likely to be intentionally destructive and deliberately set goals and engage in influence attempts that harm others (i.e., destructive leaders). Conversely, leaders with lower levels of competence, and perhaps lower narcissism levels, may inflict less harm and correspondingly be less destructive because of the lack of explicit intent and ability to do so. Padilla et al. (2007) implicitly suggested this by arguing that to determine the toxicity, one must consider consequences rather than the immediate effects of leader’s actions. Despite this, the nature and boundaries of toxic leadership relative to other dark leadership styles, namely destructive, remain unclear.
The Role of Followers
Although important in any leadership process (Graen & Uhl-Bien, 1995; Uhl-Bien, 2006), followers tend to receive only a cursory attention in the extant literature. General agreement seems to be that although followers tend to enable dark leaders, they are also the ones who suffer majority of the negative consequences (Shaw et al., 2011; Skogstad et al., 2007). For example, research has found that toxic leaders are more likely to successfully exert influence when surrounded by susceptible followers (Padilla et al., 2007) who lack security and have strong desire to be accepted by their groups (Kellerman, 2004; Lipman-Blumen, 2005).
Followers play an important role in the persistence of toxic leaders as they give them credibility and resources needed to lead. However, further analytical reasoning provides a more refined understanding of followers—with a large difference appearing from those who follow dark leaders due to fear or low tolerance toward instability and those who actively manipulate the situation for personal gain (Thoroughgood et al., 2012). To this end, Padilla et al. (2007) differentiate between two types of followers: conformers and colluders. Conformers comply with toxic leaders out of fear, whereas colluders are involved with the toxic leader to gain favorable outcomes through that association.
Other research has focused on followers as powerless participants who suffer most of the negative consequences. For example, Shaw et al. (2011) argue that most followers of dark leaders experience social stress similar to posttraumatic stress syndrome that considerably hinders their performance. Similarly, Pelletier (2010) distinguishes toxic leaders from bad leaders by emphasizing the physical and psychological harm followers experience in their interactions with leaders. However, other studies seem to suggest that followers may play a more important role in toxic leadership than simply be unwilling recipients of it. For example, Pelletier (2012) found that although leader–member exchange out-group followers perceived the leader to be toxic to a greater extent compared with in-group followers, they also indicated greater intent to challenge the leader. More recently, Fraher (2016) found that followers may counteract toxic environments by questioning suspicious dynamics, challenging taken-for-granted assumptions, and confronting others’ behaviors. Despite this preliminary insight, the role of followers in toxic leadership process requires additional inquiry. In this study, we explore how followers experience and interact with toxic leaders and examine the consequences.
Research Methods
Study Design
The research question for this study emerged during a separate study in which one of the authors examined organizational change failure. At that time, the third author recorded participants’ descriptions of an individual in leadership position regarded as key impediments to positive change. This leader, according to the participants, had insufficient competence and often used bureaucratic mechanisms to suppress positive change that would likely illuminate this lack of competence. Participants referred to this individual as toxic because they seem to be slowly poisoning the organization through their incompetence and interference with others’ work performance. As these descriptions became more prominent, the author began to consider more broadly the nature of toxic leadership and its influence on the organization. In doing so, the author engaged with dark leadership literature in general and toxic leadership literature in particular for insight. Brief review of the literature indicated that these descriptions are indeed relevant to toxic leadership literature as discussed in several studies, but they also illuminated some important omissions (Hornstein, 1996; Lipman-Blumen, 2005; Pelletier, 2010). Building on this, the author embarked on a new study driven by the following preliminary research question: What is the nature of toxic leadership and toxic influence process?
To explore the new research question, we utilized multiple case study methodology (Creswell, 2012; Eisenhardt, 1989). Case study methodology focuses on exploring a phenomenon within a bounded setting (Eisenhardt, 1989). Unlike a single case study that focuses on the uniqueness of a particular case (i.e., understanding a particular incident within a particular context), the objective of the multiple case study is to explore an issue (e.g., toxic leadership) across theoretically related cases. Furthermore, multiple case study methodology enables researchers to expand on the insights from the single case and “clarify whether an emergent finding is simply idiosyncratic to a single case or consistently replicated by several cases” (Eisenhardt & Graebner, 2007, p. 27). In doing so, the multiple case study design enabled us to gain richer insight into toxic leadership, draw parallels with existing research, and create an opportunity for theory building.
Case Selection Procedures
Based on the recommendation from Eisenhardt (1989), we selected cases that had theoretical relevance and used embedded design (Creswell, 2012) where multiple units are embedded within the particular bounded context. Starting with Luthans, Peterson, and Ibrayeva’s (1998) argument that dark leadership may be more prevalent in cultures with high levels of uncertainty avoidance, collectivistic behavioral patterns, and high power distance, we focused on contexts with set characteristics—companies located in a former communist European country. Furthermore, previous research has suggested that environments with high levels of instability and change, as well as lack of checks and balances, may create contexts for emergence of toxic leadership (Pelletier 2012; Thoroughgood et al., 2012). Correspondingly, we focused on five distinct yet theoretically related cases (cases exhibiting ambiguity and change) within the above-mentioned bounded context (national culture): agriculture, banking, education, health care, and information technology (IT), as will be elaborated upon below. Table 1 details the information about each context and corresponding data collection steps.
Data Collection.
Number of toxic leaders in addition to the number in the first stage.
We selected the agriculture sector because most companies in this sector have recently undergone a privatization process and are thus still experiencing ambiguity, instability, and change. The complexities in this context are further amplified by uncertain regulatory context concerning customs and protection tariffs.
We selected educational sector due to the reforms universities underwent for them to be more aligned with the European counterparts. This induced ambiguity with regard to what priorities are and how those priorities should be achieved.
We selected the health care sector for a similar reason. More specifically, health care is still to a large extent governmentally owned in this country and plagued with corruption, inefficiencies, and poorly defined checks and balances and work boundaries. Furthermore, over the past decade, a larger number of private health care institutions emerged resulting in lack of understanding how the care among these institutions should be provided.
We selected the IT sector because it is a relatively new sector in this environment with insufficiently developed procedures and expectations. To this end, due to instability and ambiguity firms in these sectors experience, they provide appropriate settings to explore our research question.
We selected the banking sector for three reasons. First, with the transition from command to a market economy, the role banks played in the society changed. Second, over the past decade, several foreign banks entered the market, thus fundamentally disrupting established institutional logics. Finally, after rapid growth in early years, the 2008 Great Financial Crisis affected banking in this region heavily. Once the pinnacle of white-collar job stability, the banking sector entered a downsizing cycle and changed the career dynamics in this industry.
Data Collection Procedures
First Stage of Data Collection
Our data collection occurred across two stages as illustrated in Table 1 (see interview protocol and the interview questions in the Supplementary Material available online). In the first stage, we contacted individuals who introduced us to toxic leadership during the previous study of organizational change failure as noted above. These three individuals worked in the agricultural sector and served as our key informants (Creswell, 2012). Key informants are participants who are familiar with the nature of research and who can provide initial insight (in qualitative research, the primacy is placed on participants’ experiences rather than on theory, Eisenhardt, 1989). They were reinterviewed as part of this study on their specific experiences with the toxic leader that they discussed in the previous study. Subsequently, the first (I.M.) and second author (S.M.) conducted formal interviews with participants from two other sectors: education (3) and banking (2) for a total of 8 interviews at this stage of data collection (see Table 1). We began the interviews with general questions about participants’ background and their role to establish rapport (Creswell, 2012). We then inquired into their experiences with the specific toxic leader they identified and how toxic leadership may shape their work practices.
To ensure completeness and increase accuracy, we inquired about specific examples and events where they interacted with toxic leader rather on their general belief of toxic leaders. The concrete examples of the interactions provided us with data that illustrated the specifics of toxic leaders (e.g., their lack of relevant competence) rather than their general likability or lack thereof. In addition, we collected general notes on the interview process as well as wrote memos in the field. Interviews lasted between 45 and 95 minutes, were conducted in the local language, and recorded when possible. In instances where participants refused to be recorded due to the sensitive nature of our topic, the interviewer took copious notes during and immediately after the interview.
Second Stage of Data Collection
After the first stage of data collection was completed, two of the authors engaged in preliminary analysis of the data to build initial insight as recommended by Creswell (2012) and Gioia, Corley, and Hamilton (2013). In qualitative research, data collection and analysis often advance together (Creswell, 2012; Gioia et al., 2013; Langley, 1999). The analysis was conducted by the two authors jointly where we “adhered faithfully to informant terms” (Gioia et al., 2013, p. 20) and generated approximately 40 individual codes. Two authors subsequently engaged in axial coding (Strauss & Corbin, 1998) where we discussed similarities and differences among individual codes. This endeavor gave rise to three general themes: the competence of toxic leadership, toxic influence attempts, and follower responses. This enabled us to revise our approach slightly and focus more clearly on the emergent themes in subsequent rounds of interviews. As a consequence, our interview protocol in the second stage included probes related to leaders’ competence (if the participant mentioned competence, we inquired more into it), how leaders enacted influence (we inquired about specific instances where leaders were enacting toxic influence), as well as how the participants would respond to those influence attempts.
In the second round of data collection, we used snowball sampling where our key informants from the first stage helped us identify other individuals who we could contact for insight into the nature of toxic leadership with the particular focus on the three elements identified above. At this stage, we also decided to expand the number of cases and recruit participants from the health care (5) and IT (5) sectors based on the theoretical rationale discussed above. At this time, we recruited additional participants for the agriculture (1 additional interview), education (2 additional interviews), and banking (3 additional interviews) sectors resulting in a total of 24 formal semistructured interviews across five theoretically connected cases.
Data collection occurred until we reached theoretical saturation, and there were diminishing returns with each new participant (Eisenhardt, 1989). We followed the same interview procedure as above and used revised interview protocol focused on three critical elements: leader competence, toxic influence attempts, and follower responses. This added focus, further strengthening our confidence in the data and interpretations. The interviews across the two stages were taken at the location outside the primary place of employment (due to the sensitive nature of the phenomenon of interest) that was agreeable with both the interviewer and the interviewee and conducted in the local language.
Data Analysis Procedures
In the embedded design (Creswell, 2012), analysis begins with a focus on subunits, allowing the higher perspective to emerge and shift during fieldwork. Because our objective was to remain open to emergent insights (Pratt, 2009), the embedded design provided appropriate tools to collect data. As indicated above, the preliminary data analysis took place in between the two stages to ensure that our data collection efforts were complete and our questions appropriate. Once we reached theoretical saturation in the second stage of data collection (additional interviews provided little new insight), we commenced the formal data analysis process. Given that all authors are fluent in the local language, to ensure trustworthiness of the insights (Bass & Milosevic, 2018), two of the authors completed the analysis using data in the original form. The representative quotes used in the final manuscript were later translated into English language.
The formal analysis process developed across four stages. In the first stage, we carefully read through all the interview transcripts and interview notes to immerse in detail and understand the overreaching story (Agar, 1980). In the second stage, all three authors discussed the emergent insights to understand the general direction as well as to identify any potential remaining gaps in the data. Once we ensured that theoretical saturation was indeed reached, the first author (I.M.) embarked on the third stage where they analyzed each case separately and adopted a categorical aggregation approach to data analysis (Creswell, 2012). Categorical aggregation approach to data analysis entails identifying a collection of instances from the data and allowing higher order themes to emerge. Based on the guidelines provided by Creswell (2012), we focused on identifying three types of codes: expected (codes that affirms previous findings such as leader’s efforts to limit interaction), surprising (codes that either conflict with previous findings or offer novel insight such as incompetence and upward influences attempts), and unusual (codes that are rather unexpected given the current literature such as followers’ agentic responses). The surprising and unusual codes aid in theory building and theory extension (Creswell, 2012).
In the final stage of the data analysis, the first author (I.M.) went back and forth between the cases to build an overarching narrative (Eisenhardt, 1989). At this stage, we also engaged in the theory-building process “via recursive cycling among the case data, emerging theory, and later, extant literature” (Eisenhardt & Graebner, 2007, p. 25). This process helped build relationships between codes and identify aggregate themes that illustrate the nature of toxic leadership. This helped us ensure that our findings were grounded in both the literature and the data, thus enhancing their trustworthiness (the validity). We provide insight into the emergent data structure and corresponding evidence in Table 2.
Data Structure.
We engaged in three validity procedures as suggested by Creswell (2012): member checking, triangulation, and thick description. As indicated above, all interviews were transcribed in their original form and data analysis was completed in the local language by two authors. The translated representative quotes as well as the complete manuscript were discussed with key informants who were fluent in both English and the local language to ensure that themes that we identified made sense and that our overall narrative as well as representative quotes reflected their experiences. We triangulated our findings by interviewing participants across organizational levels and across cases. The consistency of the insights across these differences enhanced our confidence in the data. Finally, we employed thick description in the presentation of our findings. More specifically, in an effort to uncover the nature of toxic leadership, we asked our participants to provide tangible examples of interactions and events involving particular toxic leaders. This in turn enabled us to provide richer, emic insight into the process of toxic leadership as experienced by our participants.
Dynamics of Toxic Leadership Process
To explicate dynamics of toxic leadership process, we present (1) three elements of the process—the intent, the influence, and the follower’s response and (2) interconnections between them (see Figure 2). Our findings indicate that toxic leaders may lack relevant competence in the field they are leading. Because of this, their primary intent is to conceal the lack of relevant competence and maintain the position of control to the exclusion of other organizationally beneficial objectives. Our findings indicate that to do so toxic leaders induce ambiguity that diminishes opportunities for performance evaluation (that may illuminate their lack of relevant competence) via upward and downward influence tactics. Through these tactics, toxic leaders manipulate both their superiors and subordinates, thereby increasing the toxicity (or “poisoning,” in the words of one participant) of the organizational context. This toxicity interferes with followers’ work performance and limits productive interactions. To this end, our findings suggest that although the intent of toxic leaders is less harmful compared with destructive and abusive leaders (i.e., there is no explicit intent to induce harm), the outcomes of their influence are harmful for the followers and the organization.

Process of toxic leading.
Our findings also illustrate, surprisingly, that followers of toxic leaders are more agentic than previously suggested in their effort to neutralize the toxicity of their leaders. We show that the toxicity of the context contributes to the feeling of powerlessness and demotivation of followers, as they face considerable difficulties in performing their duties at the desired levels and achieving their work objectives. However, our findings also illustrate that these individuals are agentic and create workarounds to neutralize when possible, or at least minimize, the leaders’ toxicity, thus defeating the toxic leader. In the subsequent paragraphs, we present our findings visually depicted in Figure 2.
The Intent of Toxic Leaders: Concealing Incompetence and Protecting a Position of Control
Toxic leaders either lack relevant competence for the position they are in (i.e., lawyer leading an IT team) or have insufficient competence (relative to others) to be in a position of control (i.e., medical doctor who has insufficient medical competence). This lack of relevant competence manifests in poor understanding of the work processes and inability to meaningfully contribute to the work processes. Despite the lack of relevant competence, these individuals do reach leadership positions largely due to the inadequate performance management systems and ambiguity of the context. Given this, unlike destructive leaders who have considerable intent to do harm or lead others toward destructive goals (Krasikova et al., 2013) or abusive leaders who abuse others due to their emotional imbalance (Tepper, 2000), the primary intent of toxic leaders is to conceal the lack of competence and protect their position of control.
Concealing the Lack of Relevant Competence
The lack of relevant competence manifests in two ways: micromanagement of followers and drastic shifts in decision making. The micromanagement manifests via focus on minute details in everything the followers do. For example, they request to be involved in all decision making that is taking place in their realm and inspect each individual piece of work or communication to be presented to others. However, in inspecting the work, they often focus not on the essence of work but on issues such as spacing, sentence structure, or formatting choices—minor elements of the work that necessarily slow down the progress. In doing so, they are in a position to demonstrate their superiority, albeit on irrelevant tasks, to others, thus concealing the general lack of work-related competence: The micro management can manifest usually on the most banal things from controlling emails . . . is the text of the email appropriate? And here we talk about people who have the experience—they have sent numerous emails before. And all the way to checking whether presentation of information is correct and wanting to be a point of contact for everything. And this last thing is probably the most problematic. You always have to ask this person what the next step is.
In addition to micromanaging their followers, toxic leaders engage in somewhat haphazard decision making permeated with sudden shifts and changes—“full left and full right” as one participant described. More specifically, because these leaders tend to lack relevant competence, they also tend to lack conviction in their actions and their decisions. This is particularly evident with toxic leaders who are generally competent but lack competence in the field they are charged with (e.g., lawyer leading the IT team). As such, they display considerable initiative to make decisions and establish authority. However, once those decisions are made, the lack of relevant competence makes them doubt the decisions, particularly when challenged by others. When this occurs, they often reverse their initial decision creating confusion and ambiguity with regard to their competence levels: The situation is bit better with those who have some competence [the toxic leader in question was considered generally competent—he was a lawyer leading IT department]. But in some ways they are more dangerous because they tend to make drastic decisions fast without discussing them with others who are perhaps more qualified. I call it “full left and full right” [i.e., dramatic changes in decisions].
Protecting the Position of Control
Furthermore, in an effort to protect the position of control, these leaders act, as our participants described, to “exclude those who have high levels of competence” and who could potentially endanger their positions by pointing out a lack of competence. By limiting information available to others or maneuvering so that others are left out of the key conversations, leaders create a boundary space permeated with ambiguity and half thoughts that neutralize others’ ability to directly confront them and challenge their competence. For example, toxic leaders actively work to exclude others who could question their decisions and their legitimacy: “[They] neutralize people who do not agree with them. They either remove them from the project or intentionally limit information they receive so they cannot perform adequately on the project. And sometimes they simply fire them.” A different participant described this as a “syndrome of an empty table”: I think that toxic leaders create what I see as a syndrome of empty table. They create barriers and remove high quality people from the table [from the interaction] or include them too late by not sharing information. I understand why they do it . . . it enables them to stay in the position. But it is a problem for the collective, for the hospital.
As a consequence, the harmfulness of their intent is primarily implicit and considerably lesser compared with destructive or abusive leaders. More specifically, unlike destructive leaders whose intent is to harm others, toxic leaders’ intent is to conceal their lack of competence and protect their own position (see Figure 2). However, the lack of explicit intent to induce harm does not imply that the influence of these leaders is harmless. More specifically, our findings suggest that as a consequence of their intent, toxic leaders engage in toxic influence attempts to build the “right” context that complements their leadership (Osborn, Uhl-Bien, & Milosevic, 2014). In doing so, they interfere with other’s work performance, thus creating harm for the organization as presented below.
Toxicity in the Toxic Leadership Process: Building the “Right” Context Through Influence
Toxic leaders influence others not to reach goals (constructive or destructive; Krasikova et al., 2013; Yukl, 2008), form relationships (Graen & Uhl-Bien, 1995), or enable positive performance (Judge & Piccolo, 2004). Rather, they engage in influence attempts to build the “right” context that strengthens their leadership—context permeated with ambiguity and confusion. They do so because such contexts allow them to isolate, at least partially, from others’ evaluation, critiques, or disproval. Toxic leaders build right contexts via upward influence, in which they work to shape evaluations of higher level managers and obtain their approval, and downward influence, in which they work to interfere with employee task performance. Together these influence attempts create toxic contexts that enable leaders to fulfill their intent—conceal their incompetence and maintain a position of control.
Upward Influence Attempts
Toxic leaders engage in several self-promoting behaviors to enact upward influence (Table 2), thereby managing perceptions of individuals they report to. For example, they actively work to present results in the most “rosy manner” even when those results do not reflect reality, are achieved using unacceptable means (ethical misconduct), or are in reality far from the level they could have been. For example, one of the participants in agricultural section described a situation where the toxic leader engaged in upward influence by presenting mediocre project results as optimal results, thus actively manipulating perceptions of his supervisors: Well, for example the project went relatively well . . . it met the expectations in a way —[but] everything is presented as perfect even though that project could have generated much better results if more competent people were involved and proper analysis conducted.
Toxic leaders also enact upward influence through friendly relationships with their supervisors. More specifically, toxic leaders frequently praise their supervisors’ ideas and support their initiatives, regardless of their actual potential for the organization. When those initiatives fail, toxic leaders present anecdotal evidence of partial success, thus manipulating their supervisors’ perceptions of their competence (i.e., making supervisors believe they are more competent than they are) as well as concealing their own incompetence. When this occurs, supervisors are distracted with anecdotal evidence of partial success, thus failing to consider the activities of toxic leaders. Toxic leaders, on the other hand, use this as an opportunity to get closer to their supervisor and further distort their perceptions, thus amplifying their toxic influence. These circular ingratiating interactions between toxic leaders and their supervisors give rise to “semi-friendly relationships that feed the power of toxic leaders,” as one of our participants explained. They feed the power by further obscuring opportunities for objective evaluations (due to anecdotal evidence and reluctance of supervisors to doubt their own competence) and legitimizing toxic leaders’ position of control.
Downward Influence Attempts
At the same time, however, these leaders engage in a spectrum of downward influences to maintain the control by limiting interaction and autonomy of employees. One of the most dominant downward influence tactics includes efforts to control subordinates using ostensibly legitimate, as well as clearly illegitimate, means. The legitimate methods include rules and regulations that although often created for legitimate purposes are misused in an effort to control the flow of information and interfere with employee work. Toxic leaders do so to impose their authority and show others that the position they occupy and the current rules afford them the authority to do what is needed (i.e., I am enacting the rules and thus doing good for the organization). For example, one of the participants from the health care sector discussed the new schedule for morning patient visits: So she created these nebulous ways to conduct work in the cardiology department such as morning patient visits. These are absolutely necessary but not in the number she requires. This amount of visits makes it impossible for the program to start at 7 in the morning. So instead of the program starting at 7 a.m. and running till 7 p.m., the program starts at noon [leading to inability to see all patients]. And when I say there are a lot of patients who are waiting—she just says “ok those are the rules.” And that extreme approach to rules endangers the essence of medical work [work with patients].
In other words, although morning visits are a critical and legitimate element of their work, the way the visits are organized and enacted limits the autonomy of medical doctors to approach their work in the most appropriate and useful manner.
However, our findings indicate that these controlling behaviors are more often illegitimate, stemming from fear, in which toxic leaders directly interfere with individual ability to perform work. These behaviors range from petty comments on individual work, thereby minimizing their contributions, to trying to exert control over employees’ personal lives. For example, a participant described the powerlessness she felt as a consequence of toxic leader’s continuous questions aimed at her personal life: It involved frequent questions such as: who did you sit with? About what did you talk about? These questions related to activities in both personal and professional sphere—the distinction that is often difficult to maintain. There is no professional boundary between life and work. You were seen with this person on your break. What did you talk about?
By persistently crossing the boundary between professional and personal, toxic leaders interfere with individuals’ ability to perform their work and slowly increase the toxicity of the context where individuals are not motivated to excel in their work or interact with the leader, as will be discussed later. In doing so, toxic leaders create context that eases upward and downward influence tactics—providing anecdotal evidence of their own excellence where others are failing to those they report to and turning other employees into either conformers or colluders (Padilla et al., 2007). Our participants described it as the “poisoning of the context” similar to the way the toxins poison their victims, whereby leaders slowly increase the toxicity of their environments, thus making any positive action difficult as discussed below.
Frustration and Powerlessness of Followers
The initial outcome of toxic influence attempts are obstacles that prevent followers from preforming their work tasks. These obstacles entail unnecessary activities (induced by imposed bureaucracy or overemphasis on insignificant details) as well as dealing with requests that are outside of the work process, and/or crossing the boundary between personal and professional. These obstacles create a state of frustration and powerlessness in followers (see Figure 2). For example, one of the participants in the health care sector (nurse) described how her supervisor expected her to deal with custodial issues—whether custodians are smiling or not. The participant experiences frustration because expectations such as these are not just demeaning, but more important, they are preventing her from dealing with more pressing issues. A medical doctor similarly described his frustration with dealing with unnecessary obstacles to his work and exclusion from the decision making: Because of some conflict that she believes happened . . . that I am not aware of, she decided to add an incompetent person to my team and take away one outstanding resident. In other words, she is disturbing the way my department works . . . and is making all these decisions outside any discussion with me.
Indeed, the harmfulness of toxic leadership is perhaps most evident in its outcomes: prolonged engagement with toxic leaders creates a state of powerlessness and frustration among followers where they begin contemplating giving up and leaving the organization. However, because finding a different job in this context is difficult, many decide to stay in toxic environment with little belief in the ability to do anything or contribute to organizational goals. This powerfulness manifests by unwillingness to engage beyond minimum, on one hand, but also a personal struggle that this is not the right thing to do: At one point you just let it go. But I do not think that is the right thing to do. Unfortunately because it is hard to find another job—it all depends on the connections you make and the influence you can yield. So you decide not to get upset anymore about these issues and hurt your health—you let it go and just do the bare minimum.
As individuals lessen engagement with work and decrease performance, powerlessness fuels gossip and unprofessional activities, thus creating the toxic context that in turn further enhances the feeling of powerlessness: One of the outcomes is that when people are demotivated and powerless they will begin to gossip. They will talk behind the leader’s back. They will talk behind everyone’s back in order to remove the ambiguity or feel better. And that fuels this bad working atmosphere. It does not matter that we are talking about people who are professionals. They are demotivated and every day someone new is ready to quit. They are poisoned by the negative atmosphere.
The Emergence of Toxic Context
As the powerlessness of followers increase, so does the toxicity of the context. According to our participants, this process is akin to slow acting toxin that slowly corrodes its surrounding. Barriers to work performance together with powerlessness and frustration isolate employees from one another, making them believe that they are powerless. Toxic leaders further strengthen the toxic contexts by allowing and celebrating only a small number of solid performers who tend to confirm to the authority. In doing so, these leaders ensure that results, albeit not optimal ones, are achieved on one hand and prevent potential challenge to their authority on the other. As one of our participants explained, When you do not have the critical mass of smart people who can drive positive change, then those leaders who are already at the positions of the authority take over and the people who have some value begin to retreat from interaction.
This in turn creates a negative, or poisonous, context that enables toxic leaders to persist: The more capacity the leader has to influence the context, build the network of conformers, and eliminates procedures that constrain them, the more likely that the transparency of work will decline, thus enabling toxic leaders to stay in power.
Defeating Toxic Leaders: The Agency in Followers’ Actions
When faced with the persistent obstacles toxic leaders create, many decide to ultimately leave the team, department, or organization despite lack of other job opportunities. However, surprisingly, many participants discussed the activities and workarounds that enable them to emerge out of powerlessness, such as circumventing the toxic leader, removing themselves from the proximal sphere of influence, and protecting their work from immediate harm. This is precisely because much of the influence toxic leaders engage in does not intend to cripple the organization. This leaves the space for others to minimize the negative outcomes. For example, participants discussed that toxic leaders are particularly problematic because through their influence attempts they endanger the process of work. At the same time, however, individuals work to protect the process of work. As one participant described, “Our process of work is such that work has to be in continuity. We are like one circle, one chain. And if we cannot hear and respect different opinions, that chain will break and new problems will be created.” In other words, many of our participants felt torn between feeling of powerlessness and the commitment to doing the work well because they considered the work itself important.
Protecting the Work
Consequently, although some participants continued feeling powerless and contemplated leaving the organization, many discussed activities they were engaging in to shield the work from the toxic influence. These activities involved building relationships with those around toxic leaders to gain understanding and support as well as learning and acquisition of resources that would help them support their arguments and stand up to toxic leaders. For example, one of the participants described how he worked to develop separate relationships with individuals in the headquarters and open a new line of communication where he could show his work directly rather than work with the toxic leader. In other words, he worked to circumvent the toxic leadership influence: We had functional responsibility for the headquarters. We ultimately were accountable to them. I developed a direct line with COO. And I built it as a counter balance to protect me from his [toxic leader’s] pressure.
Our participants also discussed proactively pursuing opportunities for learning and development that would help them to stand up to toxic leaders and minimize their toxic impact. The learning equipped them with resources they need to complete the work validate their arguments. Consequently, and surprisingly, many of the individuals engaged in these positive activities in response to toxic leaders and in an effort to protect their work. As one of the participants explained, she firmly relied on her understanding of the law and used established facts to stand up to the toxic leader and protect the process of work: I referred to the inspection of work that we cannot do it the way the manager wanted. I say based on article this and this we cannot pursue this course of action. There are 6 different laws that say we cannot do this. I do not know whether the desire for control and desire to be seen as the smartest person . . . it was obvious that I created a problem—I illuminated the incompetence. But, on the other hand, I kept true to the tenents of the law and continued to emphasize what needs to be done and how it has to be done correctly.
Discussion
Dark leadership styles seem to be prevalent in contemporary organizations, resulting in a host of negative outcomes such as poor individual performance (Harris, Kacmar, & Zivnuska, 2007; Tepper, 2000), workplace deviance (Krasikova et al., 2013; Tepper et al., 2009), and unethical work behaviors (Hannah et al., 2013; Kellerman, 2004). Despite considerable negative outcomes, studies on the dark side of leadership have been sparse relative to the studies on positive leadership styles such as transformational, charismatic, and authentic (Howell & Avolio, 1992). Furthermore, apart from the work on abusive supervision (Tepper, 2000) and destructive leadership style (Krasikova et al., 2013), there seems to be considerable conceptual overlap between different dark leadership styles, resulting in some ambiguity.
The purpose of this study was to explore one particular dark leadership style—toxic leadership—and its process of influence using multiple case study methodology. The findings contribute to the literature in three ways. First, we illustrate the toxic influence process and more clearly identify intent and the behaviors of these leaders as distinct from other dark leadership styles. Second, in taking a relational perspective, this study provides insight into the proactive roles that followers may assume to minimize the toxic impact. Finally, building on the findings and extant literature, we offer a framework of dark leadership styles that situates toxic leadership within, as well as differentiates it from, destructive, abusive, and ineffective leadership.
The Toxic Leadership Process: The Intent and Outcomes of Influence
Dark leaders are particularly detrimental for contemporary organizations because of the influence process they engage in (Krasikova et al., 2013; Padilla et al., 2007). As a consequence, there has been an increase in interest in understanding the nature of dark leadership as well as factors that contribute to the dark leadership emergence. However, there are still remaining questions with regard to how dark leaders influence others and whether different dark leadership styles may engage in the same or similar influence process. In other words, do abusive leaders lead others in the same way that toxic leaders do? Our findings indicate that the primary intent of toxic leaders is to conceal their lack of competence and maintain their position of control. To do so, these leaders engage in a host of upward and downward influence attempts to induce ambiguity that interferes with other’s work and minimizes opportunities for evaluation.
To this end, our findings show that, although the explicit intent to induce harm is lacking, thus differentiating it from destructive leaders characterized by considerable intent, the harmfulness of toxic leaders is considerable. More specifically, these leaders attempt to direct others’ attention away from work objectives and toward anecdotal evidence of success (in the context of upward influence attempts) or toward unnecessary work and artificially created obstacles (in the context of downward influence attempts) in order to position themselves as the sole problem solver. Downward influence attempts often include misapplication of rules as well as illegitimate attempts to exclude individuals from interactions, thus creating silos in the organization that interfere with the work process. This may result in less optimal decision task performance and long-term employee frustration and feeling of powerlessness. In doing so, toxic leaders build the “right” context that strengthens their leadership.
Upward Influence as an Impression Management Tactic
Individuals engage in impression management to create a new positive image or maintain the current one (Bolino, Kacmar, Turnley, & Gilstrap, 2008; Bolino, Long, & Turnley, 2016; Bozeman & Kacmar, 1997). According to the literature, these actions are most evident during performance appraisal events when subordinates engage in impression management (most often flattering the supervisors) to elicit positive evaluations (e.g., Barsness, Diekmann, & Seidel, 2005; Ellis, West, Ryan, & DeShon, 2002; Wayne & Ferris, 1990). Impression management is also present in leadership research, albeit less extensively. For example, Gardner and Avolio (1998) theorize that leaders use impression management tactics to create and maintain their identities as charismatic leaders. Sosik, Avolio, and Jung (2002) build on the work of Gardner and Avolio (1998) and found that although prosocial impression management tactics were related to charismatic leadership, self-serving impression management tactics did not.
Our findings contribute to this line of research in two ways. First, we affirm the extant theorizing by empirically illustrating several impressions management tactics toxic leaders engage in, such as ingratiation and self-promotion to enact upward influence (Jones & Pittman, 1982) and manage perceptions of individuals they report to. Second, we extend current insight by illustrating that these leaders also employ other tactics to amplify their upward influence and prolong their tenure in the position of control. For example, our findings suggest that toxic leaders present anecdotal evidence of partial success, thus manipulating their supervisors’ perceptions of their competence (i.e., making supervisors believe they are more competent than they are). When this occurs, supervisors are distracted with anecdotal evidence of partial success, thus failing to consider other activities of toxic leaders. Toxic leaders, on the other hand, use this as an opportunity to get closer to their supervisors and further distort their perceptions, thus amplifying the toxic influence. In other words, toxic leaders carefully use different tactics. We hope these findings will inspire future research in the realm of leader’s impression management tactics.
Theoretical Extension: The Agency in Followership
Followers have long held an important, yet somewhat invisible, role in leadership research (Baker, 2007; Bligh, 2011; Uhl-Bien et al., 2014). Part of the reason for this is that leadership has been predominately tied to the characteristics and behaviors of leaders rather than on the relationship between leaders and followers (Fairhurst & Uhl-Bien, 2012; Uhl-Bien et al., 2014). Yet understanding followers and their readiness to be influenced by leader, as well as their contribution to the leadership process, are critical in understanding effectiveness of a leader. Indeed, this is particularly relevant in the realm of dark leaders where followers are often portrayed as enabling these leaders to stay in power (Padilla et al., 2007) due to their lack of security and their strong desire to be accepted by others (Kellerman, 2004; Lipman-Blumen, 2005). In doing so, on one hand, these followers give credibility to the leader and provide resources they need to continue to lead regardless of how destructive that leadership is. On the other hand, these followers are usually the recipients of the destructive behavior and tend to experience considerable harm (Pelletier, 2010; Figure 1).
Our findings contribute to this line of research in two ways. First, we affirm extant findings by showing that certain followers do enable toxic leaders to stay in power. Namely, the findings show that higher level managers are often the target of the toxic leaders where toxic leaders actively engage in upward influence to shape the evaluations of those above them. As a consequence, these higher level managers are in essence followers of toxic leaders who provide much needed resources, as extant literature suggest, to toxic leaders and enable their stay in power longer. Our findings also illustrate that employees may feel powerlessness as a result of their interactions with toxic leaders, resulting in decreased performance, avoidance behaviors (avoiding situations that would require interaction with the particular leader), and failure to stand up to that leader, thus implicitly enabling the leader to stay in control (see Figure 1). Although we failed to find evidence of outright colluding, perhaps due to the perceived lack of leaders’ competence, we did discover that some followers strengthen these leaders through their powerlessness to respond.
However, our findings also extend current understanding by illustrating that, despite feelings of powerlessness, many followers engage in proactive activities to minimize the impact of toxic leadership on their work. The role of followers has traditionally been absent from leadership research, driven by the assumptions that these individuals are passive recipients in the leadership process (Uhl-Bien et al., 2014). Our findings challenge this assumption by illustrating that followers, who feel committed to their work and their careers, endeavor to shield their work despite the toxic leader influence attempts. More specifically, we show that followers display considerable agency in not just deciding to “unfollow” these leaders, as suggested before (Carsten et al., 2010), but in working to minimize their toxic influence. This work manifests in proactive relationship building with others around the toxic leader to circumvent toxic leaders as well as to engage in personal learning and development in order to acquire resources needed to stand up to toxic leaders (Figure 1). We theorize that this somewhat surprising finding may have emerged because toxic leadership is less harmful relative to destructive and abusive leadership as discussed below. That is, due to lack of explicit intent to harm others manifested by lack of relevant competence and deliberate goal setting, toxic leaders may create space for followers to challenge the status quo and drive positive change. We hope future research will examine this more and inquire into follower proactivity in difficult contexts.
Situating Toxic Leadership: The Boundaries and Connections
Building on our model of toxic leadership process and extant literature, we endeavor to extend theory by situating toxic leadership relative to other dark leadership styles—namely, destructive, abusive, and ineffective. Previous literature suggests that dark leaders may differ based on the intent to influence as well as degree of harmfulness that their influence creates (Kellerman, 2004; Krasikova et al., 2013; Schilling, 2009). However, apart from theorizing provided by Krasikova et al. (2013), there is limited insight into the nature of different types of dark leadership styles (e.g., do all dark leaders have the same intent to induce harm?), as well as boundaries of influence of each style (e.g., do abusive and destructive leaders influence followers in the same manner?). Although there is some rationale to allow a degree of overlap in order to build a more inclusive theory (Pelletier, 2010), it seems prudent to more explicitly discuss different dark leadership styles, recognize the degree of overlap, and analyze different outcomes.
In this study, we relied on a framework put forth by Krasikova et al. (2013) that differentiates dark leadership styles based on leader’s intent to harm and resulting harmfulness of their actions. This is a useful framework for several reasons. First, it provides key factors based on which different dark leadership styles may be differentiated (intent vs. outcome). Second, it is both parsimonious (in focusing on two factors) and overreaching (in encompassing leadership as a process). Finally, it both differentiates between the dark leadership styles systematically and provides space for overlap suggested by the literature and evident in practice. Using this framework, we discuss the typology of dark leadership that ranges from ineffective leadership (low intent and low to moderate harm) to destructive leadership (high intent and high harm) and situate toxic leadership within (see Figure 1).
Our findings suggest that the intent of toxic leaders is self-directed (i.e., conceal lack of relevant competence and maintain the position) and outcomes are relatively less harmful. This is in contrast to destructive leaders who have the high intent to induce harm that manifests in leader’s competence, narcissism, and volitional encouragement of followers to pursue destructive goals (Krasikova et al., 2013) or to abusive leaders who, although less intentional and calculated compared with destructive leaders, are equally narcissistic and engage in hostile emotional outbursts, thus inflicting significant harm on others (see Figure 1). Furthermore, unlike ineffective leaders who are relatively passive with low aptitude for leadership, toxic leaders are more direct and intentional in their behavior. More specifically, our findings suggest that toxic leaders are inward oriented and have little concern for others (i.e., relatively low intent to induce harm) but still intentional in their behavior in terms of achieving their own objectives—concealing the lack of competence and maintaining the position of control—thus exhibiting higher intent compared with ineffective leaders (see Figure 1).
Our findings further suggest that toxic leaders engage in influence attempts that differ from those of other dark leaders. More specifically, we show that toxic leaders engage in downward and upward influence attempts that create toxic contexts. Downward influence attempts entail creation of obstacles that interfere with the followers’ ability to perform their tasks. Upward influence attempts entail tactics that obstruct supervisors’ ability to evaluate leader’s competence. These influence attempts create a toxic context that negatively affects the organization and followers—they slowly poison the context through rising frustration and feelings of powerlessness. Consequently, these influence tactics are more harmful relative to ineffective leaders whose lack of initiative and general inaptitude to lead result in inefficiencies and poor performance that weaken the motivation and well-being of employees (Einarsen et al., 2007; Schyns & Schilling, 2013; Skogstad et al., 2007). However, they are less harmful relative to destructive or abusive leaders whose negative effect on followers and organizations is considerable and long term.
In addition to intent and outcomes of dark leaders, we point out another important difference that concerns the “negativity” of characteristics these leaders have. Previous literature has suggested that both destructive and abusive leaders tend to score high on the narcissism scale (Grijalva & Harms, 2014; Penney & Spector, 2002; Rosenthal & Pittinsky, 2006; Shaw et al., 2011). For example, narcissistic individuals tend to be less aware of their performance levels and, thus, may engage in abusive behaviors (i.e., “lash out”) toward those presenting them with information that is not consistent with their views of their performance (Grijalva & Harms, 2014; Robins & John, 1997). In addition, Krasikova et al. (2013) argued that narcissism, alongside psychopathy and Machiavellianism, is what fuels destructive leaders. More specifically, due to higher levels of narcissism, these individuals tend to engage in destructive acts and pursue destructive goals.
The degree to which toxic leaders may be narcissistic is left somewhat unclear by our findings. Building on the Grijalva and Harms (2014) argument that narcissism is a risk factor for the emergence of destructive leadership, as well as extant research that associated narcissism with a host of counterproductive work behaviors (Braun, Aydin, Frey, & Peus, 2015; Roberts & Robins, 2000), it is reasonable to expect that toxic leaders may have narcissistic characteristics as well. 2 Indeed, Hogan, Raskin, and Fazzini (1990) argued that narcissists have an innate desire for status and power that guides their behaviors, and Roberts and Robins (2000) found that narcissists tend to pursue goals that will advance their careers over other objectives. Therefore, the intent of toxic leaders to maintain the position of control may be fueled by their narcissism. However, other key aspects often associated with narcissism seem to be lacking, indicating, perhaps, that toxic leaders may have lower levels of narcissism relative to abusive and destructive leaders, or perhaps a different type of narcissism. We hope that these theoretical insights will invite further dialogue on the process of influence of dark leaders as well as their potential characteristics such as narcissism.
Limitations and Future Directions
Despite several important contributions to the extant literature on dark leadership, this study is not without limitations. First, we chose the research setting that was theoretically relevant. More specifically, Luthans et al. (1998) suggested that broader environments characterized by high levels of instability and high power distance are more conducive to the emergence of toxic leadership. Similarly, Padilla et al. (2007) suggested that toxic leadership strengthens in conducive environments and with followers who are responsive to toxic influence attempt. Although this provided a unique and theoretically relevant insight into our central phenomenon (Creswell, 2012; Siggelkow, 2007), it has also limited the generalizability of our findings. It is likely that contexts with both proper institutional environments as well as strong organizational systems will weaken toxic leadership. Karakitapoğlu-Aygün and Gumusluoglu (2013) imply that human resources managers should find and use effective procedures in the selection process to identify and screen out potentially negative and nontransformational leaders. These findings also imply that a heavy emphasis on ethical behavior and judgment (Treviño, Brown, & Hartman, 2003) in management training and education is essential. We hope future research will inquire into dynamics of toxic leadership across different contexts.
Second, the nature of qualitative research requires researchers to make several trade-offs. Tracy (2010) discusses trade-offs qualitative researchers have to embrace in their effort to provide valid findings on one hand and preserve the dynamism of qualitative research on the other. Atherton and Elsmore (2007, p. 72) discuss the trade-offs in the realm of theoretical precision by discussing how messiness of the data creates messiness in theoretical framing. To this end, Karpiak (2006, p. 93) states that qualitative researchers should embrace “the unpredictable and evolving nature of the research.” We as well were torn between increasing theoretical precision while remaining true to the experiences of our participants and presenting the complete story (Sandberg & Tsoukas, 2011; Suddaby, Hardy, & Huy, 2011). This trade-off is perhaps most evident in our presentation of our findings where we endeavored to present our participant’s narratives in a as theoretically precise manner as it was possible. We invite future research to employ other methods to continue advancing dark leadership literature and strengthen theoretical precision.
Finally, in our focus on participants’ experiences as suggested above, we primarily relied on dark leadership literature to theoretically situate our findings. Doing so enabled us to infuse additional clarity into dark leadership literature and identify the usefulness of the theoretical framework we used. However, it also prevented us from fully considering and incorporating several theories that may be relevant to our findings. For example, one potentially important aspect of dark leadership is impression management. That is, we show that toxic leaders engage in several impression management behaviors to maintain the position of control. This is likely relevant for other dark leadership styles as they work to manage perceptions of multiple stakeholders. For example, Grijalva and Harms (2014) discuss leadership style of Steve Jobs who was considered by most as a visionary charismatic leader who changed the world of technology. The same leader was described by his follower as an arrogant and abusive manager who was often very cruel to others. Given this, it seems prudent to further explore the role impression management plays in leadership, particularly in the realm of dark leadership.
In addition, the motives of dark leaders will likely differ. In other words, why do some dark leaders have high intent to harm while other less so? This is likely a function of their dark characteristics such as level of Machiavellianism, narcissism, and psychopathy as suggested by Grijalva and Harms (2014) and Krasikova et al. (2013). However, it may also be a function of the underlying motives (Higgins, 1998). Although some dark leaders will primarily be driven by advancements and accomplishments (e.g., narcissistic destructive leaders) and thus engage in behaviors to achieve those, others may be driven by security and safety and non-losses (e.g., toxic or ineffective leaders). Furthermore, although our findings indicate that competence, or lack thereof, is an important element in the intent to induce harm, we did focus only on the task competence and not on other types of competence (leadership competence, political savvy, etc.). We hope our findings will trigger additional exploration of these dynamics and consideration of other relevant literatures.
Conclusion
Toxic leadership, as a dark leadership style, has considerable influence on individual and organizational performance. Yet the nature of toxic leadership, as well as its process influence, has been only partially understood. Building on the extant research on dark leadership as well as findings from a multiple case study, this study contributes to current theorizing in three ways. First, toxic leadership may inflict less harm compared with destructive or abusive leadership due to the lack of explicit intent to harm others. Second, the toxicity of this leadership style is in influence process through which these leaders interfere with others ability to perform the work tasks, both their superiors (by interfering with their ability to evaluate leader’s performance) and subordinates (by imposing unnecessary rules and limiting productive interactions). Finally, our findings illustrate that followers of toxic leaders may exhibit considerable agency in not just refusing to follow toxic leaders but also in actively working to minimize their harmful impact. In focusing on the toxic leading process, our study thus illuminates the nuances of the toxic leadership, and more important, how followers may overcome negative outcomes and neutralize toxic leadership influence via workarounds and learning. To this end, the findings provide a step forward in the study of the dark leadership and illuminate several important avenues future research should consider.
Supplemental Material
Interview_Protocol – Supplemental material for Defeating the Toxic Boss: The Nature of Toxic Leadership and the Role of Followers
Supplemental material, Interview_Protocol for Defeating the Toxic Boss: The Nature of Toxic Leadership and the Role of Followers by Ivana Milosevic, Stefan Maric and Dragan Loncar in Journal of Leadership & Organizational Studies
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
Notes
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
