The Dark Side of Leadership: A Systematic Literature Review and Meta-Analysis of Destructive Leadership Research

Journal of Business Research xxx (xxxx) xxx
Contents lists available at ScienceDirect
Journal of Business Research

journal homepage: www.elsevier.com/locate/jbusres
The dark side of leadership: A systematic literature review and

meta-analysis of destructive leadership research☆
Jeremy D. Mackey a, *, B. Parker Ellen III b, Charn P. McAllister c, Katherine C. Alexander a
a
Auburn University, Harbert College of Business, Department of Management, 405 W. Magnolia Avenue, Auburn, AL 36849, USA
b
Northeastern University, D’Amore-McKim School of Business, Management and Organizational Development Group, 360 Huntington Avenue, Boston, MA 02115, USA
c
Northern Arizona University, W. A. Franke College of Business, Department of Management, 101 E McConnell Drive, Flagstaff, AZ 86011, USA
A R T I C L E I N F O A B S T R A C T
Keywords: The destructive leadership literature has grown remarkably in recent years. Although the field has generated an
Destructive leadership impressive body of knowledge, we still have an incomplete understanding of destructive leadership. We conduct
Leadership a systematic literature review of destructive leadership research so we can create a solid foundation for
Abusive supervision
knowledge production and theory development within this literature. Further, we draw from 418 empirical
CWB
Meta-analysis
samples of data (k = 418, N = 123,511) to conduct random-effects meta-analyses that estimate the magnitude
and direction of relationships within destructive leadership’s nomological network. Ultimately, our study le
verages and integrates the many insights from the destructive leadership literature to advance knowledge,
facilitate nuanced theory development, generate useful directions for future research, and create evidence-based
recommendations for policy and practice.
1. Introduction for effective organizational functioning.

However, destructive leadership research lacks a solid foundation
“If you only knew the power of the dark side.” because prior findings remain disjointed. The multitude of destructive
leadership styles and theoretical foundations applied within this litera
~ Darth Vader ture have generated confusion about the current state of knowledge in
Society’s fascination with Darth Vader and other notorious super the field. This lack of clarity has resulted in the need for more parsi
villains has generated tremendous scholarly and practical interest in monious theoretical frameworks and a cohesive empirical foundation.
understanding destructive leaders who use “the power of the dark side” Schyns and Schilling’s (2013) highly influential meta-analysis provides
to influence followers (Krasikova, Green, & LeBreton, 2013; Martinko, the current empirical foundation for the destructive leadership litera
Harvey, Brees, & Mackey, 2013; Schyns & Schilling, 2013; Tepper, ture. Their study has already generated a tremendous impact in this
2007; Tepper, Simon, & Park, 2017). Despite all of this interest, we still literature (e.g., over 750 citations in the seven years since its publica
do not truly understand destructive leadership, which is a broad tion, according to Google Scholar). However, their study is missing
construct that captures styles of leadership comprised of behaviors nearly a decade worth of knowledge generation (i.e., studies available
embedded within leadership influence processes that harm followers since their study searches ended in September 2010) and only examined
and/or organizations (Krasikova et al., 2013). However, prior business outcomes of destructive leadership. Mackey, McAllister, Maher, and
research has demonstrated that destructive leadership is an enduring Wang (2019) updated some of Schyns and Schilling’s meta-analytic
problem for organizations due to its adverse and expensive effects on findings, but only for performance-related variables. Thus, the empir
followers’ task performance, absenteeism, turnover, and legal actions ical foundation for destructive leadership research has continued to
(Tepper, Duffy, Henle, & Lambert, 2006). Altogether, prior business focus on outcomes of destructive leadership.
research consistently demonstrates that destructive leadership is costly We remedy this problem by conducting a systematic literature re
and adversely affects important workplace outcomes that are essential view (Snyder, 2019) that enables us to assess antecedents and outcomes
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
☆
* Corresponding author at: Auburn University, Harbert College of Business, Department of Management, Lowder Business Building, 405 W. Magnolia Avenue,
Auburn, AL 36849, USA.
E-mail addresses: jmackey@auburn.edu (J.D. Mackey), p.ellen@northeastern.edu (B. Parker Ellen), katie.alexander@auburn.edu (K.C. Alexander).
https://doi.org/10.1016/j.jbusres.2020.10.037
Received 25 November 2019; Received in revised form 9 October 2020; Accepted 11 October 2020
0148-2963/Published by Elsevier Inc.
Please cite this article as: Jeremy D. Mackey, Journal of Business Research, https://doi.org/10.1016/j.jbusres.2020.10.037
J.D. Mackey et al. Journal of Business Research xxx (xxxx) xxx
of destructive leadership so we can generate a “firm foundation for 2020). Martin Shkreli is another example of a destructive leader. Shkreli
advancing knowledge and facilitating theory development” (Snyder, is the former CEO of Retrophin and former CEO of Turing Pharmaceu
2019, p. 333). Our findings expand our knowledge of the nomological ticals. He received international acclaim as the “Pharma Bro” after
networks of destructive leadership and provide more stable meta- Turing obtained the license for the drug Daraprim and raised its price
analytic estimates than prior research so we can generate a strong, per pill for consumers from $13.50 to $750.00. He is currently serving
cohesive foundation from which to leverage the last 20 years of research seven years in prison for engaging in securities fraud and conspiring to
in this area so we can meaningfully advance it forward. commit securities fraud (Long & Hays, 2018).
The purpose of this paper is to conduct a systematic literature review Initial conceptualizations of destructive leadership described it as
of destructive leadership research that leverages meta-analytic tech “the systematic and repeated behaviour by a leader, supervisor or
niques and the enormous body of empirical research available to manager that violates the legitimate interest of the organisation by
improve our understanding of the relationships within destructive undermining and/or sabotaging the organisation’s goals, tasks, re
leadership’s nomological network. We use random-effects meta-ana sources, and effectiveness and/or the motivation, well-being or job
lyses (Hunter & Schmidt, 2004) to evaluate relationships within satisfaction of subordinates” (Einarsen, Aasland, & Skogstad, 2007, p.
destructive leadership’s nomological network that have been examined 208). Thought leaders in this area have argued that destructive leaders
so we can address our central research question “with a power that no tend to control, deceive, dominate, intimidate, manipulate, and threaten
single study has” in order to “uncover areas in which more research is their followers, as well as defraud and steal from their organizations
needed, which is a critical component of creating theoretical frame (Krasikova et al., 2013). However, recent definitions of destructive
works and building conceptual models” (Snyder, 2019, p. 333). We leadership have broadened its conceptualization by emphasizing that it
accomplish this goal by following best practice recommendations for is comprised of negative behaviors with the intent or potential to harm
conducting empirical forms of systematic literature review (i.e., meta- followers and/or organizations that leaders embed into their influence
analysis; Siddaway, Wood, & Hedges, 2018; Snyder, 2019) so we can processes. In their influential review of the destructive leadership
produce findings that help the field move forward in a manner that in literature, Krasikova et al. formally defined destructive leadership as:
forms theory building, business research, policy, and practice.
“volitional behavior by a leader that can harm or intends to harm a
We make an empirical contribution by conducting a meta-analysis of
leader’s organization and/or followers by (a) encouraging followers to
the destructive leadership literature that has unprecedented breadth for
pursue goals that contravene the legitimate interests of the organization
this literature. Our massive scope (k = 418, N = 123,511) enables us to
and/or (b) employing a leadership style that involves the use of harmful
provide a much more detailed and accurate assessment of the nomo
methods of influence with followers, regardless of justifications for such
logical network of destructive leadership than previously possible. Our
behavior” (page 1310, italics original).
systematic examination of the literature and the use of rigorous meta-
analytic techniques enable us to make an empirical contribution that
illuminates novel and nuanced insight into the magnitude of relation
ships within the nomological network of destructive leadership. Further, 2.2. Destructive leadership styles
our meta-analytic findings make a theoretical contribution by serving as
building blocks for nuanced theory construction and extension (Hunter Krasikova et al.’s (2013) definition alludes to the leadership styles
& Schmidt, 2004; Snyder, 2019) within this literature as it moves for that dominate most empirical examinations of destructive leadership.
ward. Finally, we make practical contributions by producing estimates Extant empirical research almost exclusively examines destructive
that inform evidence-based practice. Overall, our contributions leverage leadership as a leader-centric leadership style, which is consistent with
the massive scope of our meta-analysis to conduct a systematic literature how the majority of the broader literature examines leadership (Ashford
review of the destructive leadership literature so we can build the solid & Sitkin, 2019). Thus, we examine destructive leader styles in our study
empirical foundation necessary to advance knowledge. The foundation because examining followers’ perceptions of destructive leadership
we provide can facilitate nuanced theory development, generate styles serves as the current empirical foundation of our knowledge. The
actionable directions for future research, and create evidence-based deliberately broad nature of the destructive leadership construct has led
recommendations for policy and practice from the many insights to the examination of many specific styles of destructive leadership.
available within this literature. Ultimately, our study is necessary for However, abusive supervision (i.e., “subordinates’ perceptions of the
this literature to reach its unrealized potential. extent to which supervisors engage in the sustained display of hostile
Below, we provide an overview of destructive leadership, destructive verbal and nonverbal behaviors, excluding physical contact”; Tepper, 2000,
leadership styles, and destructive leadership’s nomological network. p. 178, italics original) has been especially influential in the literature.
Then, we describe the meta-analytic method we used to explore our Reviews (e.g., Martinko et al., 2013; Tepper, 2007; Tepper et al., 2017)
research question about the magnitude of relationships within destruc and meta-analyses (e.g., Mackey, Frieder, Brees, & Martinko, 2017;
tive leadership’s nomological network. Then, we describe the results of Zhang, Liu, Xu, Yang, & Bednall, 2019) of abusive supervision report
our meta-analysis. Finally, we describe the implications of our findings, that it has adverse effects that are consistent with the effects of the
as well as identify our study’s contributions to theory, research, practice, broader destructive leadership phenomenon.
and future research. Finally, we conclude by summarizing why our key However, construct proliferation (Shaffer, DeGeest, & Li, 2016) is
takeaways so we can highlight why our study is important. rampant in this literature and there are ongoing debates about how
confounding (Martinko, Harvey, & Mackey, 2014) the various destruc
2. Theoretical foundations tive leadership styles has limited their conceptual and empirical
distinctiveness (Hershcovis, 2011; Tepper & Henle, 2011). For example,
2.1. The conceptualization of destructive leadership destructive leadership styles include aversive leadership (i.e., leading
through intimidation, threats, and punishment; Bligh, Kohles, Pearce,
The presence of destructive leadership is a reality of organizational Justin, & Stovall, 2007), despotic leadership (i.e., leading by using
life, as evidenced by famous destructive leaders involved in modern personal dominance to pursue leaders’ self-interests; De Hoogh & Den
corporate scandals. For example, Elizabeth Holmes, the former chief Hartog, 2008), exploitative leadership (i.e., leading with the main intent
executive officer [CEO] of Theranos, made false claims about her com to further leaders’ self-interests; Schmid, Verdorfer, & Peus, 2019), and
pany’s blood-testing technology, engaged in wire fraud, and conspired leader narcissism (i.e., leaders’ behaviors that are “principally moti
to commit additional wire fraud that would have distributed falsified vated by their own egomaniacal needs and beliefs”; Rosenthal & Pit
blood test results to consumers (Carreyrou, 2018; Hartmans & Leskin, tinsky, 2006, p. 631).
2
Additionally, leader bullying (i.e., leaders targeting negative be Table 1

haviors toward followers who have difficulty defending themselves; Definitions for Various Styles of Destructive Leadership.
Einarsen & Skogstad, 1996), leader exclusion (i.e., leaders denying fol Construct Definition
lowers acceptance or consideration; Scott, 2007), leader incivility (i.e.,
Destructive Leadership “Volitional behavior by a leader that can harm or
leaders displaying a lack of regard for followers; Andersson & Pearson, intends to harm a leader’s organization and/or
1999), and leader undermining (i.e., leaders hindering followers’ followers by (a) encouraging followers to pursue goals
interpersonal relationships and work-related success; Duffy, Ganster, & that contravene the legitimate interests of the
Pagon, 2002) are also considered destructive leadership styles. Table 1 organization and/or (b) employing a leadership style
that involves the use of harmful methods of influence
provides complete definitions of these and several other styles of with followers, regardless of justifications for such
destructive leadership that have been identified in prior research. behavior” (Krasikova, Green, & Lebreton, 2013, p.
All of the aforementioned destructive leadership styles emphasize 1310)
harmful methods leaders can use to influence their followers. It is Abusive Supervision “Subordinates’ perceptions of the extent to which
supervisors engage in the sustained display of hostile
important to note the nuances of each destructive leadership style within
verbal and nonverbal behaviors, excluding physical
the broader destructive leadership framework because they each have contact” (Tepper, 2000, p. 178)
implications for theory development within this literature. For example, Aversive Leadership “Leadership behaviors that emphasize the use of
the theoretical mechanisms that explain why insincere leadership (i.e., threats, intimidation, and punishment” (Bligh, Kohles,
leaders who use deceitful influence tactics; Schilling, 2009), corrupt Pearce, Justin, & Stovall, 2007, p. 530)
Corrupt Leadership “The leader and at least some followers lie, cheat, or
leadership (i.e., leaders who lie, cheat, and/or steal; Kellerman, 2004), steal to a degree that exceeds the norm, they put self-
petty tyranny (i.e., leaders who lord their power over followers; Ash interest ahead of the public interest” (Kellerman,
forth, 1997), and evil leadership (i.e., leading by committing atrocities; 2004, p. 44)
Kellerman, 2004) all impact followers’ outcomes likely differ. Further, Derailed Leadership “Leaders may display anti-subordinate behaviours like
bullying, humiliation, manipulation, deception or
there are numerous styles of destructive leadership that broadly capture
harassment, while simultaneously performing anti-
negative leader behaviors (e.g., derailed leadership and negative lead organisational behaviours like absenteeism, shirking,
ership), leaders who primarily focus on their self-interests (e.g., fraud, or theft” (Einarsen, Aasland, & Skogstad, 2007,
personalized charismatic leadership and pseudo-transformational lead p. 212-213)
ership), and leaders who are generally considered poisonous to those Despotic Leadership Leadership that is “self-aggrandizing and exploitative
of others” because it “is based on personal dominance
around them (i.e., toxic leadership). Some styles of destructive leader
and authoritarian behavior that serves the self-interest
ship even capture leadership processes that benefit the organization but of the leader” (De Hoogh & Den Hartog, 2008, p. 298)
disregard the welfare of their followers (e.g., tyrannical leadership; Evil Leadership “The leader and at least some followers commit
Einarsen et al., 2007) or organizational outsiders (e.g., insular leader atrocities. They use pain as an instrument of power.
The harm done to men, women, and children is severe
ship; Kellerman, 2004).
rather than slight. The harm can be physical,
Although there has been a great deal of rich conceptual work done to psychological, or both” (Kellerman, 2004, p. 46)
differentiate numerous styles of destructive leadership, it is important to Exploitative Leadership “Leadership with the primary intention to further the
build a cohesive literature that accounts for the theoretical foundations leader’s self-interest. Such leaders exploit others by
of various destructive leadership styles because destructive leadership is (1) acting egoistically, (2) exerting pressure and
manipulating followers, (3) overburdening followers,
deliberately conceptualized as a broad umbrella construct that includes
or, on the other hand, (4) consistently
these various types of destructive leadership (Krasikova et al., 2013). underchallenging followers, allowing no
Ultimately, Krasikova et al.’s definition of destructive leadership serves development” (Schmid, Verdorfer, & Peus, 2019, p.
as a useful framework for understanding the unifying factor of these 1404)
Insincere Leadership Occurs when leaders use “a diverse set of leadership
various leadership styles (i.e., leaders’ harmful methods of influencing
behaviours to achieve personal goals at the expense of
followers). However, it is important to determine whether our current others without direct confrontation but rather in the
empirical foundation is solid because research that examines destructive form of clandestine and deceitful tactics and
leadership has expanded exponentially in recent years. As a result, our strategies” (Schilling, 2009, p. 114)
empirical knowledge about destructive leadership is still incomplete Insular Leadership “The leader and at least some followers minimize or
disregard the health and welfare of the ‘other’ – that
because we still do not have a precise understanding of the magnitude of
is, those outside the group or organization for which
relationships within destructive leadership’s nomological network. they are directly responsible” (Kellerman, 2004, p. 45)
Leader Bullying (i.e., occurs “repeatedly over a period of time, and the
2.3. Destructive Leadership’s nomological network person confronted has to have difficulties defending
himself/herself”; Einarsen & Skogstad, 1996, p. 191)
Leader Exclusion Leaders deny followers “acceptance into meaningful
Extant research consistently demonstrates that destructive leader workplace relationships, activities or events” (Scott,
ship harms organizations and their members (Krasikova et al., 2013). 2007, p. 15)
For example, prior research has found that destructive leadership has Leader Incivility Leaders’ use of “low-intensity deviant behavior with
adverse effects on followers’ task performance (e.g., Tepper, Moss, & ambiguous intent to harm the target, in violation of
workplace norms for mutual respect. Uncivil
Duffy, 2011), voice (Carnevale, Huang, & Harms, 2018b), and work behaviors are characteristically rude and
place deviance (e.g., Vogel & Mitchell, 2017). Numerous theories and discourteous, displaying a lack of regard for others” (
frameworks have informed this literature, but the application of social Andersson & Pearson, 1999, p. 457)
psychological and resource-based theories has proven especially Leader Narcissism Leaders’ behaviors “principally motivated by their
own egomaniacal needs and beliefs, superseding the
insightful (Mackey et al., 2019). Social psychological theories explain
needs and interests of the constituents and institutions
why justice/fairness perceptions (e.g., Tepper, 2000) and social ex they lead” (Rosenthal & Pittinsky, 2006, p. 631)
change perceptions (e.g., Mitchell & Ambrose, 2007) affect followers’ Leader Undermining Leaders’ “behavior intended to hinder, over time, the
responses to destructive leadership. In contrast, resource-based theories ability to establish and maintain positive
explain why self-regulatory resources (e.g., McAllister, Mackey, & interpersonal relationships, work-related success, and
favorable reputation” (Duffy, Ganster, & Pagon, 2002,
Perrewé, 2018) and ego depletion (e.g., Lin, Ma, & Johnson, 2016) affect p. 332)
followers’ reactions to destructive leadership. The prevalent use of so Negative Leadership
cial psychological and resource-based theories has contributed to our (continued on next page)
understanding of numerous relationships within destructive
3
Table 1 (continued ) for future research and theory” (Snyder, 2019, p. 339) in this area by
Construct Definition generating a foundation for knowledge generation. Above, we provided
an overview of the conceptualization of destructive leadership and the
Leaders engage in “commonly disliked and denounced
behaviours ranging from ineffective to destructive
theoretical foundations that explain many of the findings within this
aspects” (Schilling, 2009, p. 103) literature. Below, we use meta-analysis to illuminate precise insights
Personalized Charismatic Leaders emphasize their own self-interest and into the magnitude of relationships within destructive leadership’s
Leadership purposefully create unbalanced relationships with nomological network so we can answer our Research Question:
their followers by manipulating and disempowering
them (Howell, 1988)
Pseudo-Transformational Occurs when “leaders advance their own self- Research Question: What is the magnitude of relationships within
Leadership interested agendas by dominating and controlling destructive leadership’s nomological network?
their followers. In focusing on self-interest, pseudo-
transformational leaders are more interested in 3. Method
becoming personal idols than in the collective ideals
that might benefit their followers” (Barling, Christie,
& Turner, 2008, p. 852) We conducted random-effects meta-analyses of relationships within
Petty Tyranny “Someone who uses their power and authority destructive leadership’s nomological network so we could answer our
oppressively, capriciously, and perhaps vindictively. Research Question. We closely followed best practice recommendations
It suggests, in short, someone who lords their power
(e.g., Geyskens, Krishnan, Steenkamp, & Cunha, 2009) for implement
over others” (Ashforth, 1997, p. 126)
Toxic Leadership “Individuals, who by dint of their destructive ing the meta-analysis reporting standards (e.g., Kepes, McDaniel,
behaviors and dysfunctional personal qualities Brannick, & Banks, 2013) while conducting our meta-analysis.
generate a serious and enduring poisonous effect on Accordingly, we are explicit and transparent about how we conducted
the individuals, families, organizations, communities, our study so our results are replicable, theoretically sound, and relevant
and even societies they lead” (Lipman-Blumen, 2005,
p. 30)
for policy and practice.
Tyrannical Leadership
“Tyrannical leaders may behave in accordance with 3.1. Systematic literature search
the goals, tasks, missions and strategies of the
organisation, but they typically obtain results not
We made systematic efforts to locate destructive leadership research
through, but at the cost of subordinates” (Einarsen
et al., 2007, p. 212) that was available as of January 2020 so we could synthesize and
compare evidence across studies. Our goal was to be as systematic as
possible so we could create a meta-analytic data set that was represen
leadership’s nomological network because certain relationships have tative of the empirical destructive leadership literature. We used eight
been important to examine within these theoretical frameworks. literature search strategies to find empirical primary studies written in
Many other theories have also been applied within the destructive English that included at least one destructive leadership variable.
leadership literature to supplement the findings central to the social First, we used Google Scholar to systematically search for studies that
psychological and resource-based theories noted above. For example, cited destructive leadership measure development papers (e.g., Larsson,
trait activation theory has played an important role in clarifying why Brandebo, & Nilsson, 2012; Shaw, Erickson, & Harvey, 2011; Thor
followers’ individual differences explain their reactions to destructive oughgood, Tate, Sawyer, & Jacobs, 2012). Second, we used Google
leadership (e.g., Wang, Harms, & Mackey, 2015). Further, affective Scholar to search for studies that cited measure development papers for
events theory has helped provide a nuanced understanding of how state specific styles of destructive leadership, such as abusive supervision
negative and positive affect, as well as a host of other emotional re (Tepper, 2000), aversive leadership (Pearce & Sims, 2002), despotic
actions, relate to followers’ perceptions of and reactions to destructive leadership (De Hoogh & Den Hartog, 2008), exploitative leadership
leadership (e.g., Han, Harms, & Bai, 2017). More direct connections (Schmid et al., 2019), narcissistic leadership (Rosenthal & Pittinsky,
have been established between followers’ perceptions of destructive 2006), and petty tyranny (Ashforth, 1997). We also searched for studies
leadership and its adverse effects on the quality of relationships between that used measures of leaders’ downward-directed bullying (Einarsen,
leaders and their followers (e.g., leader-member exchange; Xu, Huang, Hoel, & Notelaers, 2009), incivility (Cortina, Magley, Williams, &
Lam, & Miao, 2012). Langhout, 2001), undermining (Duffy et al., 2002), and narcissism
Finally, it is important to acknowledge the importance of followers’ (Rosenthal & Pittinsky, 2006).
perceptions of destructive leaders on their own behaviors. For example, Third, we conducted searches on Google Scholar for studies that
scholars have found that social learning theory helps explain why and examined the specific types of destructive leadership that Krasikova
how followers learn and replicate negative behaviors from their leaders et al. (2013) described, Schyns and Schilling (2013) identified, and we
(e.g., Liu, Liao, & Loi, 2012), as well as why they would direct these included in Table 1. Specifically, we searched for studies with abusive,
negative behaviors toward their coworkers (e.g., Lian, Ferris, & Brown, aversive, corrupt, derailed, despotic, destructive, evil, exploitative,
2012) and/or families (e.g., Hoobler & Brass, 2006). Other studies have insincere, insular, narcissistic, negative, personal(ized) charismatic,
focused on followers’ attempts to retaliate toward their leaders based on pseudo-transformational, psychopathic, toxic, and/or tyrannical lead
the central tenets of self-control theory (e.g., Lian, Brown, Ferris, Liang, ership or supervision in the title. Fourth, we searched the reference
Keeping, & Morrison, 2014). The theories noted above have had a strong sections of recent reviews (e.g., Krasikova et al., 2013; Martinko et al.,
influence on findings within the destructive leadership literature, but an 2013; Tepper, 2007; Tepper et al., 2017) and meta-analyses (e.g.,
array of other theories have also been applied in various studies too. Mackey et al., 2017, 2019; Park, Hoobler, Wu, Liden, Hu, & Wilson,
Overall, the theories noted above provide useful explanations for why 2019; Schyns & Schilling, 2013; Zhang & Bednall, 2016; Zhang et al.,
destructive leadership would be related to followers’ personalities, in 2019) of destructive leadership and related topics (e.g., abusive super
dividual differences, attitudes, perceptions, and behaviors. vision) for studies to include.
However, there is still uncertainty about our understanding of re Fifth, we searched scholarly databases (i.e., ProQuest Dissertations
lationships within destructive leadership’s nomological network and Theses, Web of Science, PsycINFO) for journal articles, dissertations,
because the fragmented research in this area is not adequately inte theses, book chapters, conference papers, technical reports, and working
grated across the many theoretical frameworks and destructive leader papers that included “destructive” or “abusive” and “leadership” any
ship styles that have been examined. We remedy this problem by where in the full text of the manuscript. We used the databases listed
conducting a systematic literature review that can “serve as the grounds above to conduct methodologically rigorous searches for unpublished
4
studies so we could limit the effects of the file drawer problem (i.e., the 3.3. Coding
suppression of weak and non-significant results; Rothstein, Sutton, &
Borenstein, 2005) and publication bias (i.e., a meta-analytic sample that Our goal was to create a low-inference coding system that would
is systematically unrepresentative of the literature; Kepes, Banks, reduce uncertainty, increase transparency, and limit the need for coders
McDaniel, & Whetzel, 2012) on our results. to make subjective judgment calls (Aguinis, Dalton, et al., 2011). We
Sixth, we searched for in press and online first articles available on developed a coding form in Microsoft Excel so we could standardize as
the websites for business research journals that were included in the much of the coding process as possible. The fourth author of this study
Financial Times 50 journal list. We also conducted these searches for listed the authors, years, study/sample numbers (when necessary),
journals ranked as 3, 4, or 4* journals in the “ETHICS-CSR-MAN”, “HRM destructive leadership styles, correlates, and notes for primary studies in
& EMP”, “ORG STUD”, and “PSYCH (WOP-OB)” sections of the Char the coding form. Then, we made the coding form available to under
tered Association of Business Schools Academic Journal Guide. Seventh, graduate research assistants in a folder on Google Drive, along with a
we searched through conference proceedings for the Academy of Man Word file that summarized the coding protocol and pdf files of the pri
agement, Society for Industrial and Organizational Psychology, and mary studies listed in the coding form. Next, the second and third au
Southern Management Association for the years between 2000 (i.e., thors trained the research assistants how to use the protocol and coding
when Tepper [2000] was published) and 2019 (i.e., the most recent data form. The research assistants tested the coding form to ensure that they
available as of January 2020) for conference papers that included understood the coding form’s functionality, which included data vali
“destructive” or “abusive” and “leadership” anywhere in the full text of dation and drop-down menus so we could enhance coding accuracy.
the manuscript. Finally, we examined the reference lists of studies Then, two research assistants independently coded information from
located throughout our searches so we could identify studies that could each study.
be relevant for our meta-analysis. The coders recorded the publication types (e.g., journal article,
dissertation), study designs (i.e., self-report data or dyadic data), and
3.2. Inclusion criteria sample sizes (i.e., n) of the primary studies included in our meta-
analysis. The coders also recorded respondents’ nationalities. Next, the
We did not restrict sample inclusion based on when or where coders recorded several features of the destructive leadership measures,
empirical studies were conducted. However, we did require studies to including the low scale point, high scale point, number of items, and
meet the seven quality-based inclusion criteria described below to be response scale (i.e., agreement or frequency). Then, coders recorded the
included in our meta-analysis. The primary purpose for the inclusion rating source (i.e., self- or other-rating), mean (M), standard deviation
criteria was to standardize the focal variables as much as possible so we (SD), and the internal consistency (i.e., α; we exclusively used Cron
could enhance the validity of our findings (Aguinis, Dalton, Bosco, bach’s alpha) of the destructive leadership measures. Also, the coders
Pierce, & Dalton, 2011) and ability of others to replicate our meta- recorded the α and rating source of the correlates. Finally, the coders
analytic data set (Aytug, Rothstein, Zhou, & Kern, 2012). recorded the bivariate zero-order correlations (r) between destructive
First, we only included studies that were written in English. Second, leadership and its correlates. The coders primarily reported data from
we only included studies that surveyed employed respondents. Third, correlation tables unless there was a clear and obvious error, but they
we required that each study empirically measured destructive leader supplemented this data with information from the text of the studies that
ship in a manner that was consistent with our conceptualization of it (see we included in our meta-analyses. We defaulted to reporting informa
Table 1). Fourth, we only included studies that reported followers’ tion from correlation tables when the information within the text and
perceptions of destructive leadership, which has traditionally been tables was inconsistent.
operationalized from the follower’s perspective (Mackey et al., 2019; After the coders initially completed the coding form, we deleted data
Wang, Van Iddekinge, Zhang, & Bishoff, 2019). Fifth, we only included in cells that were inconsistent between coders. Then, we highlighted the
studies that reported destructive leadership variable(s) in correlation blank cells in red so the research assistants could recode the data prior to
matrices. Sixth, we required that all correlations from correlation submitting their final codes. Next, the second author calculated inter-
matrices were reported at the individual level of analysis. Finally, we rater agreement between the coders, which ranged from a low of 85%
required that the correlates we examined were consistent with the def for rating source to a high of 98% for publication type. Overall, we found
initions of constructs reported in Supplement A so we could standardize high levels of agreement because coders agreed at least 90% of the time
the correlates as much as possible. for 18 of the 19 coding categories. Then, the second author consulted the
Additionally, we examined the description of each sample of data primary studies to resolve all coding discrepancies. Next, the first author
that was included in our analyses to ensure that we did not violate the searched for abnormalities in the coding form (e.g., studies that reported
assumption of sample independence required to meaningfully use correlations in the opposite direction of expectations, extreme values),
Hunter and Schmidt’s (2004) meta-analytic technique. We defaulted to attempted to locate missing data, and prepared the data for analysis.
journal articles, dissertations, theses, conference papers, book chapters, When necessary, the first author used Mosier’s (1943) Equation 8 (see
technical reports, and working papers in that order when there was data Supplement C) to create composite variables for studies that reported
overlap across multiple primary studies. We included the earliest multiple effect sizes for a specific relationship, such as when a variable
available data for overlapping data across studies of the same type. was collected across multiple time periods or rating sources. Mosier’s
We enhanced the validity of our findings by strictly adhering to our method for creating composite variables is consistent with the use of
inclusion criteria. Ultimately, our search efforts identified 368 empirical Hunter and Schmidt’s (2004) meta-analytic technique for the nature of
studies that met our inclusion criteria. The 281 journal articles, 38 our research questions and theoretical framework. The third author
doctoral dissertations, 16 master’s theses, 32 conference papers, and one verified all of the final coding information reported throughout the
unpublished report are reported in Supplement B. The 368 studies supplements to ensure that there were no transcription errors.
included a total of 437 independent samples without data overlap that
met our inclusion criteria. We were able to incorporate 418 of these 3.4. Analyses
samples (k = 418, N = 123,511) into our final analyses because they
reported at least one correlation that was included in the analyses re We used Hunter and Schmidt’s meta-analysis program (Schmidt &
ported in our tables. Le, 2004) to run random-effects meta-analyses that weighted the results
by sample size. Random-effects models allow for population parameters
(i.e., ρ) to vary across studies because the models assume that the studies
included in the analyses are similar without requiring them to be
5
identical (Borenstein, Hedges, Higgins, & Rothstein, 2010). Thus, (k = 385, N = 114,196), petty tyranny (k = 2, N = 1,736), and general
random-effects models were appropriate for our analyses due to our measures of destructive leadership (k = 12, N = 4,488). Seven of the
study design and Research Question. studies included in the counts above (k = 7, N = 5,652) examined
We report several key results. First, we report the number of samples multiple styles of destructive leadership, which we labeled “composite”.
(k) and respondents (N) in each analysis. Then, we report weighted Overall, we identified 13 different styles of destructive leadership that
mean bivariate correlations (r) and population correlations that correct inform empirical findings within its literature. Our results demonstrate
for sampling error and measurement error (ρ), as well as their respective that the majority of empirical destructive leadership studies examine
SDs (i.e., SDr and SDρ). These corrections are important because “sample abusive supervision. We provide a summary of the destructive leader
sizes are never infinite” and “there are no perfectly reliable measures” ship style(s) that were examined in each primary study in Supplement F.
(Hunter & Schmidt, 2004, p. 31). Next, we report the percentage of Our Research Question asked: “What is the magnitude of relation
variance that is attributable to artifacts for the population correlation ships within destructive leadership’s nomological network?” We used a
estimates. Finally, we report the 80% credibility intervals and 95% multi-step process to answer our Research Question. First, we examined
confidence intervals for the population correlations. The credibility in the overall results of our random-effects meta-analyses. Table 2 provides
tervals report the approximate ρ distribution within which 80% of the an overview of the relationships in destructive leadership’s nomological
obtained estimates occur (i.e., ρ ± [1.28 × SDρ]), which is useful for network that have been included in at least five independent samples of
evaluating the precision of our point estimates (Edwards & Christian, empirical data. The correlates include followers’ demographic infor
2014). In contrast, the confidence intervals are used to indicate the mation, Big Five personality factors, individual differences, perceptions
approximate ρ distribution within which we can be 95% certain that the of leadership styles, attitudes and perceptions, and behaviors, as well as
true value exists (Hunter & Schmidt, 2004). a few leader variables. Below, we report our point estimates of effect
We used Cronbach’s alpha (α) to correct the reported correlations for sizes (i.e., ρ), the heterogeneity of our point estimates (i.e., SDρ), the
measurement error because we examined correlations at the construct precision of our point estimates (i.e., 80% credibility intervals; Edwards
level (Hunter & Schmidt, 2004; Viswesvaran, Ones, Schmidt, Le, & Oh, & Christian, 2014), and the sample sizes (i.e., k and N) for each analysis.
2014). We used the median α value available from the other studies in We found many interesting results. First, all of the followers’ de
our meta-analysis that examined each specific relationship (see Sup mographic variables that we examined were weakly associated with
plement D) when αs were not reported in a primary study and when one- destructive leadership (− 0.03 ≤ ρ ≤ 0.06) despite their widespread use
or two-item measures were used. We used median α values instead of as control variables (e.g., age: k = 190; sex: k = 206). However, there
mean α values because median values are less susceptible to systematic were some attitudes (e.g., job satisfaction; ρ = − 0.41, SDρ = 0.15, 80%
sources of error from outliers (Aguinis, Gottfredson, & Joo, 2013). The credibility interval: [− 0.60, − 0.22], k = 52, N = 17,717), perceptions (e.
median α estimates we used were similar to the αs found in the man g., perceived organizational support; ρ = − 0.43, SDρ = 0.17, 80%
agement literature (Greco, O’Boyle, Cockburn, & Yuan, 2018) and used credibility interval: [− 0.66, − 0.21], k = 19, N = 4,756), and behaviors
in meta-analyses of destructive leadership styles (e.g., abusive supervi (e.g., counterproductive work behavior [CWB]; ρ = 0.43, SDρ = 0.15,
sion; Mackey et al., 2017). We assumed that there was perfect reliability 80% credibility interval: [0.24, 0.62], k = 64, N = 21,893) that were
(i.e., α = 1.00) for demographic information. We did not make correc meaningfully associated with destructive leadership. Further, destruc
tions for rater source (DeSimone, 2014; LeBreton, Scherer, & James, tive leadership was strongly and negatively related to several other
2014). forms of leadership perceptions, such as ethical leadership (ρ = − 0.63,
We conducted meta-analyses for correlates that were included in five SDρ = 0.22, 80% credibility interval: [− 0.91, − 0.34], k = 18, N = 8,186)
or more empirical studies that reported at least one destructive leader and leader-member exchange (LMX; ρ = − 0.52, SDρ = 0.15, 80%
ship variable. We chose this threshold because Field (2005) identified credibility interval: [− 0.72, − 0.33], k = 32, N = 9,077). Finally,
five studies as the minimum sample size to have enough power to destructive leadership was meaningfully related to followers’ task per
accurately conduct random-effects meta-analyses that assume effect formance (ρ = − 0.23, SDρ = 0.16, 80% credibility interval: [− 0.43,
sizes are approximately normally distributed. We reported correlations − 0.03], k = 60, N = 16,379) and organizational citizenship behavior
and effect sizes throughout our study without statistical significance (OCB; ρ = − 0.24, SDρ = 0.13, 80% credibility interval: [− 0.41, − 0.07],
tests so we did not mix meta-analysis with null hypothesis significance k = 32, N = 7,281).
testing (Hunter & Schmidt, 2004). In tandem with Table 2, the comparison of studies shown in Table 3
demonstrates that we were able to include many more correlates, utilize
4. Results much larger sample sizes (i.e., ks), and provide more robust estimates (i.
e., ρ instead of r) with indices of heterogeneity (i.e., SDρ) than Schyns
Our meta-analytic results include data from 418 samples of empirical and Schilling’s (2013) assessment of the destructive leadership litera
data (k = 418, N = 123,511) from working adults. We report the 1,944 ture. Mackey et al. (2019) included large samples, but they conceptu
correlations and their corresponding input values for the analyses in alized CWB and OCB much more broadly than we did. In contrast, we
Supplement E so our data set is explicit, transparent, and replicable. were more narrowly focused and considered some of the types of extra-
Overall, our data set is representative of the extant empirical literature role behaviors they included in these estimates (e.g., creativity, voice) as
that examines destructive leadership in the workplace from followers’ distinct constructs. Thus, Mackey et al. provided estimates of eight
perspectives. broadly defined performance variables, whereas we provide estimates of
For the first part of our empirical literature review, we sought to 84 distinct relationships. In summary, our study replicates and extends
determine which styles of destructive leadership have been empirically findings from previous meta-analyses for a small portion of the re
examined and the extent to which extant empirical research utilizes each lationships we examine and generates new meta-analytic estimates for
of the destructive leadership styles so we could provide information over 70 new relationships, including antecedents, within destructive
about how destructive leadership has been empirically examined. We leadership’s nomological network.
found empirical studies that included aversive (k = 7, N = 1,926), Finally, we conducted supplementary analyses for studies that
derailed (k = 2, N = 4,187), despotic (k = 4, N = 2,166), exploitative (k examined abusive supervision (see Supplement G) because this was the
= 3, N = 852), negative (k = 1, N = 131), and tyrannical (k = 1, N = dominant measure of destructive leadership used throughout the pri
2,539) leadership. We also found empirical studies that included leader mary studies included in our meta-analysis. We offer both a broad view
bullying (k = 1, N = 252), exclusion (k = 1, N = 252), incivility (k = 4, N of destructive leadership and a narrow view of abusive supervision for
= 959), narcissism (k = 5, N = 1,244), and undermining (k = 18, N = the collective empirical evidence we synthesized. Thus, we provide not
4,929). Additionally, we found studies that included abusive supervision only a systematic literature review of the destructive leadership
6
Table 2
Meta-Analytic Results.
Analysis k N r SDr ρ SDρ 80% Credibility Interval 95% Confidence Interval % Variance Attributable to
(ρ) (ρ) Artifacts
Followers’ Demographic Information

Age 190 63,879 − 0.02 0.10 − 0.02 0.09 (− 0.14, 0.10) (− 0.03, 0.00) 28%
Education 62 20,850 − 0.03 0.11 − 0.03 0.10 (− 0.16, 0.10) (− 0.06, 0.00) 24%
Hours Worked Per Week 14 10,221 0.06 0.04 0.06 0.01 (0.04, 0.07) (0.04, 0.08) 92%
Marital Status 8 4,514 − 0.03 0.07 − 0.03 0.06 (− 0.11, 0.04) (− 0.08, 0.02) 36%
Position 17 4,546 0.02 0.09 0.02 0.07 (− 0.07, 0.12) (− 0.02, 0.07) 44%
Sex 206 64,712 − 0.03 0.12 − 0.03 0.11 (− 0.17, 0.11) (− 0.05, − 0.01) 24%
Tenure with Job 25 8,795 0.00 0.08 0.00 0.07 (− 0.08, 0.08) (− 0.03, 0.03) 42%
Tenure with Leader 74 20,992 0.03 0.09 0.04 0.07 (− 0.06, 0.13) (0.01, 0.06) 41%
Tenure with Organization 100 25,963 0.01 0.08 0.01 0.05 (− 0.06, 0.07) (− 0.01, 0.02) 60%
Tenure with Work Group 5 1,258 0.05 0.22 0.05 0.23 (− 0.24, 0.34) (− 0.16, 0.25) 8%
Work Experience 8 2,720 0.01 0.09 0.01 0.08 (− 0.09, 0.10) (− 0.06, 0.07) 37%
Leader Variables
Leader Age 21 5,356 − 0.04 0.14 − 0.04 0.12 (− 0.20, 0.12) (− 0.10, 0.02) 22%
Leader Sex 35 7,561 − 0.06 0.14 − 0.06 0.12 (− 0.22, 0.09) (− 0.11, − 0.02) 25%
Leader Tenure with Org. 9 2,056 0.01 0.06 0.01 0.00 (0.01, 0.01) (− 0.03, 0.06) 100%
Leader Interactional Justice 6 1,262 − 0.43 0.18 − 0.47 0.18 (− 0.69, − 0.24) (− 0.61, − 0.32) 11%
Followers’ Big Five Personality Factors

Agreeableness 23 6,911 − 0.13 0.09 − 0.15 0.09 (− 0.26, − 0.04) (− 0.19, − 0.11) 38%
Conscientiousness 27 7,779 − 0.15 0.12 − 0.18 0.12 (− 0.33, − 0.02) (− 0.23, − 0.13) 24%
Extraversion 18 5,409 − 0.03 0.09 − 0.03 0.08 (− 0.13, 0.06) (− 0.08, 0.01) 45%
Neuroticism 28 7,948 0.17 0.17 0.20 0.18 (− 0.03, 0.44) (0.13, 0.27) 12%
Openness to Experience 12 4,150 − 0.06 0.11 − 0.08 0.11 (− 0.22, 0.06) (− 0.15, − 0.01) 27%
Followers’ Individual Differences

Core Self-Evaluation 11 2,545 − 0.20 0.12 − 0.24 0.11 (− 0.38, − 0.09) (− 0.31, − 0.16) 30%
Moral Identity 8 1,930 − 0.18 0.14 − 0.22 0.14 (− 0.40, − 0.04) (− 0.33, − 0.11) 22%
Narcissism 6 1,238 0.06 0.14 0.08 0.15 (− 0.11, 0.26) (− 0.05, 0.21) 24%
Negative Affectivity 45 14,754 0.32 0.13 0.36 0.13 (0.19, 0.53) (0.32, 0.40) 15%
Positive Affectivity 16 3,544 − 0.17 0.14 − 0.19 0.14 (− 0.37, − 0.02) (− 0.27, − 0.12) 23%
Power Distance Orientation 15 4,473 0.09 0.14 0.10 0.16 (− 0.10, 0.31) (0.02, 0.19) 16%
Proactive Personality 5 2,951 − 0.05 0.05 − 0.06 0.03 (− 0.09, − 0.02) (− 0.10, − 0.01) 74%
Psychological Capital 7 3,212 − 0.26 0.21 − 0.29 0.22 (− 0.57, 0.00) (− 0.46, − 0.12) 4%
Self-Esteem 15 7,431 − 0.15 0.10 − 0.17 0.11 (− 0.31, − 0.03) (− 0.23, − 0.11) 18%
Social Desirability 6 1,459 0.03 0.17 0.04 0.19 (− 0.20, 0.28) (− 0.12, 0.20) 15%
Trait Anger 5 1,391 0.12 0.04 0.13 0.00 (0.13, 0.13) (0.08, 0.18) 100%
Followers’ Perceptions of Leadership Styles

Authoritarian Leadership 8 1,190 0.42 0.15 0.47 0.16 (0.27, 0.68) (0.35, 0.59) 19%
Ethical Leadership 18 8,186 − 0.57 0.21 − 0.63 0.22 (− 0.91, − 0.34) (− 0.73, − 0.52) 2%
Leader-Member Ex. (LMX) 32 9,077 − 0.47 0.15 − 0.52 0.15 (− 0.72, − 0.33) (− 0.58, − 0.47) 10%
Transactional Leadership 7 2,156 0.11 0.29 0.12 0.33 (− 0.31, 0.54) (− 0.13, 0.37) 4%
Transformational 15 3,922 − 0.32 0.19 − 0.34 0.20 (− 0.60, − 0.09) (− 0.45, − 0.24) 9%
Leadership
Followers’ Attitudes and Perceptions

Anger 10 2,512 0.39 0.18 0.42 0.18 (0.18, 0.65) (0.30, 0.54) 9%
Anxiety 11 2,961 0.24 0.09 0.27 0.08 (0.16, 0.37) (0.21, 0.33) 39%
Burnout 9 3,816 0.41 0.10 0.46 0.08 (0.36, 0.57) (0.40, 0.52) 25%
Commitment
Affective Commitment 21 7,722 − 0.25 0.12 − 0.29 0.13 (− 0.46, − 0.13) (− 0.35, − 0.23) 16%
Organizational Commit. 10 2,859 − 0.28 0.11 − 0.31 0.10 (− 0.43, − 0.19) (− 0.38, − 0.24) 29%
Depression 11 4,863 0.28 0.07 0.31 0.06 (0.24, 0.39) (0.27, 0.36) 42%
Emotional Exhaustion 42 13,953 0.37 0.12 0.42 0.13 (0.26, 0.58) (0.38, 0.46) 15%
Engagement 15 3,608 − 0.22 0.10 − 0.24 0.08 (− 0.34, − 0.14) (− 0.29, − 0.19) 41%
Fear of Leader 5 1,427 0.46 0.16 0.52 0.18 (0.29, 0.76) (0.36, 0.69) 8%
Frustration 5 1,339 0.33 0.10 0.40 0.11 (0.26, 0.54) (0.29, 0.51) 27%
Job Insecurity 5 5,547 0.29 0.07 0.35 0.08 (0.25, 0.46) (0.28, 0.43) 16%
Job Satisfaction 52 17,717 − 0.37 0.15 − 0.41 0.15 (− 0.60, − 0.22) (− 0.45, − 0.37) 11%
Job Tension 23 6,467 0.31 0.10 0.36 0.09 (0.24, 0.47) (0.31, 0.40) 31%
Justice Perceptions
Distributive Justice 12 4,572 − 0.26 0.10 − 0.28 0.09 (− 0.40, − 0.16) (− 0.34, − 0.22) 23%
Interactional Justice 17 4,170 − 0.49 0.17 − 0.54 0.17 (− 0.76, − 0.32) (− 0.62, − 0.46) 9%
Interpersonal Justice 10 2,723 − 0.56 0.15 − 0.61 0.17 (− 0.83, − 0.40) (− 0.72, − 0.51) 7%
Procedural Justice 15 4,937 − 0.31 0.10 − 0.34 0.10 (− 0.47, − 0.21) (− 0.40, − 0.28) 22%
Organizational Justice 9 2,269 − 0.36 0.21 − 0.41 0.20 (− 0.66, − 0.16) (− 0.54, − 0.28) 9%
Negative Affect 21 5,557 0.39 0.14 0.43 0.14 (0.25, 0.60) (0.36, 0.49) 15%
Organizational 7 2,739 − 0.27 0.11 − 0.31 0.11 (− 0.45, − 0.16) (− 0.40, − 0.22) 18%
Identification
Org.-Based Self-Esteem 7 2,111 − 0.27 0.12 − 0.29 0.13 (− 0.45, − 0.12) (− 0.39, − 0.18) 16%
Ostracism 5 2,678 0.61 0.19 0.63 0.19 (0.39, 0.87) (0.46, 0.80) 2%
Perceived Org. Support 19 4,756 − 0.39 0.17 − 0.43 0.17 (− 0.66, − 0.21) (− 0.52, − 0.35) 10%
Positive Affect 9 3,041 − 0.16 0.10 − 0.18 0.09 (− 0.29, − 0.07) (− 0.25, − 0.11) 31%
Psych. Contract Breach 9 2,224 0.37 0.11 0.42 0.11 (0.27, 0.56) (0.33, 0.50) 23%
(continued on next page)
7
Table 2 (continued )
Analysis k N r SDr ρ SDρ 80% Credibility Interval 95% Confidence Interval % Variance Attributable to
(ρ) (ρ) Artifacts
Psychological Distress 8 3,794 0.33 0.14 0.37 0.14 (0.19, 0.55) (0.27, 0.47) 9%
Resource Management 5 867 − 0.30 0.10 − 0.35 0.07 (− 0.44, − 0.26) (− 0.44, − 0.26) 55%
Ability
Self-Efficacy 7 2,265 − 0.16 0.11 − 0.18 0.11 (− 0.32, − 0.05) (− 0.27, − 0.10) 25%
Supervisor Org. 7 1,246 − 0.29 0.14 − 0.32 0.12 (− 0.48, − 0.16) (− 0.42, − 0.21) 27%
Embodiment
Trust in Leader 11 3,560 − 0.46 0.14 − 0.51 0.15 (− 0.71, − 0.32) (− 0.61, − 0.42) 10%
Turnover Intention 54 18,868 0.35 0.15 0.40 0.15 (0.21, 0.58) (0.35, 0.44) 11%
Work-Family Conflict 17 7,604 0.29 0.16 0.33 0.17 (0.11, 0.54) (0.25, 0.41) 8%
Followers’ Behaviors
Aggression 6 1,810 0.44 0.14 0.49 0.16 (0.28, 0.69) (0.35, 0.62) 10%
Bullying 8 2,920 0.56 0.19 0.60 0.19 (0.36, 0.84) (0.47, 0.73) 4%
Creativity 16 5,080 − 0.18 0.12 − 0.20 0.12 (− 0.36, − 0.04) (− 0.27, − 0.13) 19%
Counterproductive Work Behavior (CWB)
CWB-General 64 21,893 0.39 0.16 0.43 0.15 (0.24, 0.62) (0.39, 0.47) 10%
CWB-Interpersonal 30 9,244 0.34 0.16 0.39 0.15 (0.19, 0.58) (0.33, 0.44) 12%
CWB-Organizational 55 18,158 0.38 0.15 0.43 0.15 (0.24, 0.62) (0.39, 0.47) 11%
CWB-Leader 30 9,618 0.49 0.14 0.56 0.15 (0.37, 0.75) (0.50, 0.61) 9%
Helping Behavior 5 1,132 − 0.19 0.14 − 0.22 0.13 (− 0.38, − 0.06) (− 0.34, − 0.09) 25%
Organizational Citizenship Behavior (OCB)
OCB-General 32 7,281 − 0.21 0.13 − 0.24 0.13 (− 0.41, − 0.07) (− 0.29, − 0.19) 22%
OCB-Interpersonal 15 4,342 − 0.17 0.14 − 0.19 0.14 (− 0.37, − 0.01) (− 0.26, − 0.11) 17%
OCB-Organizational 15 3,727 − 0.22 0.11 − 0.25 0.11 (− 0.39, − 0.11) (− 0.31, − 0.19) 29%
Silence 9 2,509 0.27 0.10 0.31 0.09 (0.19, 0.43) (0.24, 0.38) 32%
Spousal Undermining 5 878 0.35 0.09 0.37 0.06 (0.30, 0.45) (0.29, 0.45) 58%
Task Performance 60 16,379 − 0.20 0.15 − 0.23 0.16 (− 0.43, − 0.03) (− 0.27, − 0.18) 15%
Voice 9 2,931 − 0.23 0.15 − 0.26 0.15 (− 0.46, − 0.07) (− 0.37, − 0.16) 14%
Work Effort 10 2,786 − 0.21 0.10 − 0.23 0.09 (− 0.35, − 0.11) (− 0.30, − 0.16) 32%
Note. k = number of studies included in the analysis. N = total sample size of all studies included in the analysis. r = average weighted bivariate correlation across
studies. SDr = standard deviation of the average weighted bivariate correlations across studies. ρ = the population estimate that corrects the zero-order bivariate
correlation for measurement and sampling error across studies. SDρ = standard deviation of the population correlation estimates across studies. Ex. = Exchange. Org. =
Organizational. Commit. = Commitment. Psych. = Psychological. Sex was coded such that male = 0 and female = 1. Marital status was coded such that single = 0 and
married = 1.
literature, but we also update Mackey et al.’s (2017) meta-analysis on improves our understanding of destructive leadership. Our explicit,
abusive supervision. The results for abusive supervision studies were transparent, and replicable data set included the broad array of
similar to the overall results. Mackey et al.’s study searches ended three destructive leadership styles examined in prior empirical research (see
years before ours and did not include unpublished studies, which Table 1) so we could examine the magnitude of relationships within its
enabled us to meaningfully update their findings with a much larger and nomological network (see Table 2). Our study makes an important
more inclusive meta-analytic data set (Mackey et al.: k = 140; our k = contribution to business research and the destructive leadership litera
384 for abusive supervision) so we could examine 84 relationships ture because “building your research on and relating it to existing
within abusive supervision’s nomological network instead of only the 36 knowledge is the building block of all academic research activities”
relationships that Mackey et al. included in their study. (Snyder, 2019, p. 333). Our results support the prevailing wisdom that
Additionally, we conducted sub-group moderator analyses (Gonza destructive leadership is adversely associated with followers’ percep
lez-Mulé & Aguinis, 2018) for the relationships reported in Supplement tions and behaviors, but our addition of the last decade’s worth of data
G. Specifically, we report the results for each relationship based on (1) to the empirical foundation for this literature also generates more pre
Tepper’s (2000) full 15-item measure of abusive supervision and (2) cise estimates of relationships within destructive leadership’s nomo
Mitchell and Ambrose’s (2007) five-item measure. Although Tepper’s logical network, which refines our understanding of destructive
measure includes a wide variety of abusive supervisory behaviors, leadership.
Mitchell and Ambrose’s widely-used adaptation only focuses on active- Our systematic approach to examining the destructive leadership
aggressive abusive supervision. Thus, there is a subtle, but important, literature enabled us to answer our Research Question: “What is the
difference between the content domain of the abridged and full versions magnitude of relationships within destructive leadership’s nomological
of the abusive supervision scale that are often used in this stream of network?” We also uncovered numerous areas for further development
research. These differences manifested in a few key differences between that can lead to meaningful theoretical development. For example, we
results obtained across the five-item and 15-item measures, such as the found that numerous styles of destructive leadership have been empir
results for self-esteem (ρ15-item = − 0.13; ρ5-item = − 0.34), leader- ically examined. However, it was clear that empirical research examined
member exchange (ρ15-item = − 0.57; ρ5-item = − 0.34), interactional fewer styles of destructive leadership than conceptual research has
justice (ρ15-item = − 0.59; ρ5-item = − 0.36), and OCBs toward organiza described, which means that the breadth and depth of the conceptuali
tions (ρ15-item = − 0.30; ρ5-item = − 0.07). However, the results across the zation of destructive leadership may not be adequately integrated into
sub-group moderator analyses were similar across most variables. its measurement across studies. Also, the heterogeneity (i.e., SDρ) in our
results suggests that broadly assuming destructive leadership is “bad”
5. Discussion for followers and organizations is insufficient for generating a precise
understanding of why, how, and the extent to which destructive lead
We drew data from 418 empirical studies (k = 418, N = 123,511) ership impacts followers. Thus, our study advances destructive leader
that examined destructive leadership to conduct random-effects meta- ship research by providing the precision, scope, and depth necessary to
analyses so we could engage in an empirical literature review that identify and explain findings within its literature.
8
Table 3 Morgan, Kolev, & Mcnamara, 2018; Rauch, Rosenbusch, Unger, & Frese,
A Comparison of Results between Our Study and Prior Meta-Analyses. 2016) because meta-analytic results estimate true relationships that are
Our Study Schyns & Mackey et al. of scientific and practical interest. Indeed, meta-analytic findings have
Schilling (2013) (2019) “become essential in the evolution of knowledge about management”
k ρ(SDρ) k r k ρ(SDρ) (Combs, Crook, & Rauch, 2019, p. 1) because they generate impactful
insights that do not suffer from the methodological limitations present in
Individual all primary studies (Hunter & Schmidt, 2004). Up-to-date meta-analyses
Differences
Negative Affectivity 45 0.36 15 0.34
are important because their findings tend to be highly influential; also,
(0.13) techniques for explicitly and transparently reporting meta-analytic
Positive Affectivity 16 − 0.19 8 − 0.09 procedures keep improving in ways that enhance the robustness and
(0.14) validity of meta-analytic findings (Aytug et al., 2012). Plus, future
Attitudes and Perceptions research in this burgeoning area can leverage the impressive body of
Job Satisfaction 51 − 0.41 21 − 0.34 knowledge summarized in our results to determine how to move this
(0.15) field forward, rather than rely on single primary studies to provide in
Job Tension 23 0.36 24 0.24
sights. Thus, our systematic evaluation of the magnitude of relationships
(0.09)
Justice Perceptions 12 − 0.32 within its nomological network provides an important update necessary
Distributive Justice 12 − 0.28 for this literature to move toward its unrealized potential.
(0.09) Second, we make a contribution by illuminating that abusive su
Interactional Justice 17 − 0.54
pervision and the theories extensively applied within this literature
(0.17)
Procedural Justice 15 − 0.34 provide the primary theoretical foundation for our understanding of
(0.10) destructive leadership. Our study provides empirical evidence that much
Organizational 9 − 0.41 of destructive leadership research actually examines subordinates’
Justice (0.20) negative perceptions of their supervisors instead of how destructive
Turnover Intention 54 0.40 11 0.31
leaders impact followers and their organizations. Thus, our study dem
(0.15)
onstrates that much of our understanding of destructive leadership is
Behaviors
rooted in the relationship between subordinates and their supervisors
CWB-General 64 0.43 19 0.38 83 0.45
(0.15) (0.14)
instead of the process of influence that traditionally dominates leader
CWB-Interpersonal 30 0.39 31 0.38 ship research (Ashford & Sitkin, 2019). This contribution generates a
(0.15) (0.11) solid foundation that facilitates theory development (Snyder, 2019).
CWB-Organizational 55 0.43 55 0.40 Thus, our study makes important empirical and theoretical contribu
(0.15) (0.15)
tions that resolve debates about relationships within destructive lead
CWB-Leader 30 0.56 31 0.56
(0.15) (0.13) ership’s nomological network while also identifying debates that need to
OCB-General 32 − 0.24 64 − 0.23 be started or given additional attention.
(0.13) (0.13) We are able to make important theoretical and empirical contribu
OCB-Interpersonal 15 − 0.19 22 − 0.25 tions because our meta-analytic literature review re-stabilizes the
(0.14) (0.12)
OCB-Organization 15 − 0.25 37 − 0.24
foundation for the destructive leadership literature. Despite the utility
(0.11) (0.14) and influence of Schyns and Schilling’s (2013) meta-analysis, it is
Task Performance 60 − 0.23 12 − 0.20 54 − 0.23 important to update this aging foundation (e.g., Schyns & Schilling: k =
(0.16) (0.16) 57; our study: k = 418) because the cumulative evidence in this litera
Note. k = number of studies included in the analysis.r = average weighted ture has amassed to the point where fragmentation could inhibit cohe
bivariate correlation across studies. ρ = the population correlation estimate that sive knowledge production. Further, our unprecedented scope and
corrects the zero-order bivariate correlation for measurement and sampling systematic searches enabled us to examine narrow variables instead of
error across primary studies. SDρ = standard deviation of population correlation lumping them into composites like prior meta-analyses did (e.g., Mackey
estimates. CWB = counterproductive work behavior. OCB = organizational et al., 2017, 2019; Schyns & Schilling, 2013). Thus, we substantively
citizenship behavior. Some of Mackey et al.’s (2019) estimates have higher improve the foundation for our understanding of destructive leader
sample sizes than ours because they conceptualized CWB and OCB much more
ship’s nomological network broadly and abusive supervision’s nomo
broadly than we did. In contrast, we were more narrowly focused and considered
logical network narrowly. Accordingly, our results make practical
some of the types of extra-role behaviors they included in these estimates (e.g.,
creativity, voice) as distinct constructs.
contributions that stem from good science that generates meaningful
insights for evidence-based practice.
Throughout the sections below, we expand on our study’s contri
5.2. Limitations
butions to theory, business research, policy, and practice so scholars and
practitioners can push our understanding of destructive leadership for
We discuss our study’s limitations below so we can properly
ward toward reaching its unrealized potential.
contextualize our findings. First, our meta-analytic data were limited by
the quality and accuracy of the information reported in the primary
5.1. Contributions to theory and business research studies that we included in our analyses (Banks et al., 2016; O’Boyle,
Banks, & Gonzalez-Mulé, 2017). Thus, we carefully evaluated studies
We make important contributions by conducting a systematic liter during the literature search, coding process, and analyses to ensure our
ature review that provides novel insight into the magnitude of re results were not biased by misreported information or transcription
lationships within destructive leadership’s nomological network. First, errors.
we make an empirical contribution by conducting the most complete Second, it was evident that not all studies followed best practice
meta-analysis of destructive leadership research to date. Our motivation recommendations for data screening (DeSimone & Harms, 2018; DeSi
for this contribution is to advance the conversation in the destructive mone, Harms, & DeSimone, 2015; Wood, Harms, Lowman, & DeSimone,
leadership literature by underpinning its fragmented foundation with a 2017) and data preparation (Aguinis, Hill, & Bailey, in press) because
solid layer of empirical knowledge. Our approach has been successfully information was inconsistently reported across the text and tables in
applied to other literatures in other meta-analyses (e.g., Hughes- some studies. We standardized our coding process by defaulting to
9
information reported in correlation tables. However, inconsistent debates about why substantial heterogeneity in findings exists across
reporting is usually problematic for meta-analyses, especially ours some of our estimates, especially for relationships that have large SDr
because nonresponse bias is prevalent in destructive leadership and values, large SDρ values, and/or wide 80% credibility intervals. For
other CWB research, which results in lower response rates for this type of example, why is there so much heterogeneity in the observed relation
research that can attenuate observed relationships reported in primary ship between destructive leadership and followers’ task performance (ρ
studies (Greco, O’Boyle, & Walter, 2015). Additionally, destructive = − 0.23, SDρ = 0.16, 80% credibility interval: [− 0.43, − 0.03], k = 60,
leadership and other related variables, such as abusive supervision (e.g., N = 16,379)? Future research is needed to offer a parsimonious theo
Mackey et al., 2017) and workplace deviance (e.g., Mackey, McAllister, retical explanation for this finding.
Ellen, & Carson, in press) tend to be low base-rate phenomena with To this end, we recommend examining destructive leaders’ motives
problematic skew and response bias. All of these issues affect the mea for engaging in behavior that is perceived as destructive. Prior research
surement of destructive leadership, which impacts the resultant has argued that destructive leaders sometimes engage in strategic ex
knowledge we generated from research that examines it. pressions of hostility (Tepper, Duffy, & Breaux-Soignet, 2012) and/or
Third, it is possible that our meta-analytic data set was not accurately strategic bullying (Ferris, Zinko, Brouer, Buckley, & Harvey, 2007) with
representative of the empirical destructive leadership literature if un performance promotion motives (Liu et al., 2012). Thus, it is possible
published data that were not available to us systematically differed from that some destructive leadership perceptions capture leaders who
the studies that we were able to include in our analyses (i.e., the file engage in destructive actions with constructive intentions. This might
drawer problem; Kepes et al., 2012; Rothstein et al., 2005). We strived to explain why we do not see a stronger relationship between destructive
limit publication bias and the file drawer problem by including a large leadership and followers’ task performance (ρ = − 0.23), as well as why
sample of studies and conducting extensive searches for unpublished there is so much heterogeneity in this relationship (SDρ = 0.16). There
data. This enabled us to incorporate 38 doctoral dissertations, 16 mas may be some level of destructive leadership that is actually functional as
ter’s theses, 32 conference papers, and one unpublished report into our a means of motivating followers if properly implemented. Future
study. research that examines curvilinear effects of destructive leadership on
Fourth, we did not correct for direct or indirect range restriction followers’ task performance could examine this possibility.
because we did not have enough information to determine if either Additionally, future research could generate a nuanced under
existed or what effect they would have had on our findings (Hunter, standing of how much of destructive leadership is in the eye of the
Schmidt, & Le, 2006). Fifth, our data precluded us from making in beholder. Our findings demonstrate that followers’ trait negative
ferences about causality because the inclusion criteria limited our data affectivity (ρ = 0.36, SDρ = 0.13, 80% credibility interval: [0.19, 0.53],
set to studies with field data instead of experimental manipulations. k = 45, N = 14,754) and state negative affect (ρ = 0.43, SDρ = 0.14, 80%
Accordingly, we emphasize that our results are intended to be charac credibility interval: [0.25, 0.60], k = 21, N = 5,557) are both related to
teristic of the observed correlations reported across a representative destructive leadership perceptions. Thus, some subordinates may be
sample of empirical studies that examine employed followers’ percep primed to perceive destructive leadership. Accordingly, it is important
tions of destructive leadership. Sixth, we found evidence of heteroge for future research to examine how followers’ personality traits affect
neity across some of the relationships we examined. However, the their perceptions of destructive leadership, especially because we found
magnitude of many of our meta-analytic results was consistent with that the Big Five personality factors were generally weak predictors of
estimates from other meta-analyses of social psychological phenomena destructive leadership (− 0.18 ≤ ρ ≤ 0.20). Instead, we encourage future
(Paterson, Harms, Steel, & Credé, 2016). Altogether, the limitations research to explore the role of the Dark Triad (i.e., Machiavellianism,
noted above are common for meta-analyses. narcissism, and psychopathy), as well as other “dark” personality traits,
Finally, our focal variable (i.e., destructive leadership) had some such as schadenfreude (Van Dijk, Ouwerkerk, Goslinga, & Nieweg,
limitations that likely impacted our conclusions. We included many 2005) and sadism (Min, Pavisic, Howald, Highhouse, & Zickar, 2019).
different measures of destructive leadership into our study, and each has Next, additional empirical destructive leadership research is needed
important conceptual and empirical distinctions. It is clear from our to examine the nuances of abusive supervision measures. Specifically,
findings that this literature is reliant on Tepper’s (2000) measure of additional research is needed to understand the theoretical, conceptual,
abusive supervision to assess destructive leadership. However, we found and empirical differences between findings from Mitchell and Ambro
studies that used other measures of destructive leadership, so variations se’s (2007) five-item measure of active-aggressive abusive supervision
in the measurement of destructive leadership across studies could versus Tepper’s (2000) full 15-item measure of abusive supervision. The
impact the validity of our inferences about findings across this literature. results of sub-group moderator analyses in Supplement G show that
It is possible that differences across these measures impacted the pre results from studies that employ these different measures are usually
cision (i.e., 80% credibility intervals) and/or heterogeneity (i.e., SDr and similar (e.g., job satisfaction: ρ15-item = − 0.34; ρ5-item = − 0.35), but that
SDρ) of our results. Ultimately, it is important to consider the alignment they are sometimes notably different (e.g., LMX: ρ15-item = − 0.57; ρ5-item
between the conceptualization and operationalization (Heggestad, = − 0.34). Thus, we call for additional research to generate meaningful
Scheaf, Banks, Hausfeld, Tonidandel, & Williams, 2019) of destructive nuances to our knowledge by determining which relationships and to
leadership throughout this literature while evaluating our findings. what extent our understanding of abusive supervision (specifically) and
destructive leadership (broadly) is impacted by the use of different
5.3. Actionable agenda for future business research measures.
Additionally, examining styles of destructive leadership other than
It is evident from the thousands of studies we searched through and abusive supervision is important because abusive supervision focuses on
hundreds of empirical destructive leadership studies we found that this supervisors/managers, whereas (destructive) leadership is a process of
area is of considerable scholarly and practical interest. Below, we influence that can come from formal/designated and informal/non-
describe how researchers can use the meta-analytic building blocks we designated sources (Ashford & Sitkin, 2019). Although we were able
provide to improve the field (DeSimone, Köhler, & Schoen, 2019). to include empirical studies with 13 different styles of destructive
Overall, we encourage researchers to leverage the data available in our leadership in our meta-analysis, many of them were only examined in a
supplements to further engage in the storytelling process that explains few studies. We encourage future research to compare results across
why observed relationships exist (Shepherd & Suddaby, 2017) as they destructive leadership styles to evaluate how findings from specific
provide depth that thickens the plotlines in new and existing theoretical destructive leadership styles (e.g., leader narcissism; Carnevale, Huang,
stories. & Harms, 2018a) generalize across broad assessments of destructive
First, it would be informative for future research to instigate needed leadership. The rampant construct proliferation that is present in this
10
area (see Table 1) has contributed to the inconsistent study designs power because their goals align with their leaders’ goals) contribute to
applied in the destructive leadership literature. the destructive leadership process (Thoroughgood et al., 2012).
The important conceptual distinctions between various forms of Similarly, empirical research has not yet developed a solid under
destructive leadership are evident when examining the definitions of the standing of how the defining features of conducive environments
various styles of destructive leadership reported in Table 1. For example, contribute to destructive leadership processes. Thoroughgood et al.
destructive leadership can be relatively covert (e.g., leader exclusion) or (2018) provided a useful conceptual summary of the key features that
overt (e.g., evil leadership). Further, leaders can engage in behaviors produce conducive organizational environments, which include lack of
with ambiguous intent (e.g., leader incivility) or clear intent (e.g., internal and external checks and balances, organizational instability,
aversive leadership). Additionally, destructive leadership can be ongoing complexity and dynamism, and reduced scrutiny. However, the
conceptualized in specific (e.g., lying, cheating, and stealing in corrupt scant empirical research that examines how susceptible followers and
leadership) or broad (e.g., insincere leadership) terms. Thus, the conducive organizational environments contribute to destructive lead
conceptualization and operationalization of destructive leadership has ership processes that enable, tolerate, and sometimes even promote
important implications for study findings. The five-item and 15-item destructive leadership is problematic for this literature. Although there
measures of abusive supervision are, by far, the most widely used are some examples of scholars examining the Toxic Triangle in extreme
measures in this stream of research. However, the results in Supplement contexts (e.g., the Penn State scandal; Thoroughgood & Padilla, 2013),
G demonstrate that even these two measures do not always generate we advocate for research that validates measures of conducive envi
consistent results. Thus, additional research is needed to determine the ronments and various types of susceptible followers that can be applied
extent to which the conceptualizations and operationalizations of broadly so we can systematically improve our understanding of the
destructive leadership variables are valid, as well as the extent to which Toxic Triangle as it operates in a variety of organizational settings.
they overlap. Ultimately, it would be useful to generate evidence-based
recommendations for how to create a cohesive body of research. 5.4. Implications for practice
Also, we recommend that researchers leverage our findings when
evaluating potential substantive and control variables for future studies. Our meta-analytic results have clear and important implications for
We hope that our findings help researchers identify followers’ work evidence-based practice (Aguinis, Banks, Rogelberg, & Cascio, 2020;
place attitudes (e.g., turnover intention), perceptions (e.g., job tension), Ones, Viswesvaran, & Schmidt, 2017) and consulting (De Fuentes &
and behaviors (e.g., CWB) that are meaningfully associated with Porcuna, 2016). Prior business research has reported adverse associa
destructive leadership and theoretically relevant to their research tions between destructive leadership and workplace outcomes. Our re
questions. sults support the prevailing wisdom, but we also provide refined
Next, we encourage future research to continue to expand on some of estimates of the magnitude of these relationships. Our findings are
the promising avenues that have emerged recently. For example, important for practitioners and consultants because they stem from
research is beginning to explore how traditionally positive leader traits, collective empirical evidence that is less subject to context-specific
attitudes, and behaviors can have destructive consequences (e.g., hu biases than the results from individual primary studies. Further, we
mility and humor; Yam, Christian, Wei, Liao, & Nai, 2018; Zapata & provide important information about how consistently our results apply
Hayes-Jones, 2019). Conversely, additional research is needed to across studies (i.e., when SDr and SDρ values are low and 80% credibility
expand on what we know about the positive sides of traditionally dark intervals are narrow) so we can demonstrate that destructive leadership
leader traits (e.g., the Dark Triad and beyond; Harms & Spain, 2015; has many meaningful effects on followers’ workplace attitudes, per
Spain, Harms, & LeBreton, 2014), attitudes, and behaviors (e.g., moral ceptions, and behaviors that are of utility and interest for practitioners
cleansing; Liao, Yam, Johnson, Liu, & Song, 2018). Additionally, and consultants.
research is needed to improve our limited understanding of the ante We encourage practitioners to consider that it likely is difficult to
cedents of leaders’ destructive behaviors, such as their envy (e.g., Yu, make sweeping generalizations about which employees will likely
Duffy, & Tepper, 2018), stress (e.g., Harms, Credé, Tynan, Leon, & perceive that they experienced destructive leadership because we found
Jeung, 2017), socioeconomic status (e.g., Martin, Côté, & Woodruff, weak results for demographic variables (− 0.03 ≤ ρ ≤ 0.06). Instead, our
2016), and state levels of cognitive resources (e.g., sleepiness; Barnes, results suggest that followers’ attitudes, perceptions, and behaviors are
Lucianetti, Bhave, & Christian, 2015). Our findings reported in Table 2 more meaningfully associated with destructive leadership, so this may
can provide a foundation for examining leader and follower character be a useful starting point for evaluating follower characteristics that are
istics that could be used to build more robust theories of the antecedents associated with destructive leadership. For example, we found that
of destructive leadership, as well as integrate findings from these destructive leadership was strongly associated with CWBs toward
emerging areas. leaders (ρ = 0.56, SDρ = 0.15, 80% credibility interval: [0.37, 0.75], k =
Finally, we recommend that researchers improve our understanding 30, N = 9,618) and coworkers (ρ = 0.39, SDρ = 0.15, 80% credibility
of the full experience of destructive leadership, which encompasses interval: [0.19, 0.58], k = 30, N = 9,244), so CWBs may be an indicator
leaders, followers, and organizations. This combination of contextual of the presence of followers’ perceptions of destructive leadership.
features is labeled the “Toxic Triangle”, which consists of destructive Finally, we encourage organizational leaders to be aware of the mean
leaders, susceptible followers, and conducive environments (Padilla, ingful relationships between destructive leadership and followers’ task
Hogan, & Kaiser, 2007; Thoroughgood, Sawyer, Padilla, & Lunsford, performance (ρ = − 0.23, SDρ = 0.16, 80% credibility interval: [− 0.43,
2018). At a broad conceptual level, “authoritarians”, “lost souls”, “by − 0.03], k = 60, N = 16,379) and OCBs (ρ = − 0.24, SDρ = 0.13, 80%
standers”, “opportunists”, and “acolytes” are theorized to either coop credibility interval: [− 0.41, − 0.07], k = 32, N = 7,281) because these
erate (i.e., collude) with destructive leaders or comply (i.e., conform) behaviors are critical for effective organizational functioning.
with them (Thoroughgood, Padilla, Hunter, & Tate, 2012) in unstable
organizations that lack institutional checks and balances (Padilla et al., 6. Conclusion
2007). However, empirical research has not yet developed a solid un
derstanding of how the defining features of “authoritarians” (i.e., fol Our motivation for this study was to conduct a systematic literature
lowers influenced by legitimate power), “lost souls” (i.e., followers who review that improves our understanding of the magnitude of relation
have unmet needs that are influenced by referent power), “bystanders” ships within destructive leadership’s nomological network. We drew
(i.e., manipulative followers who are influenced by coercive power), from 418 empirical studies (k = 418, N = 123,511) that examined
“opportunists” (i.e., ambitious followers who are influenced by reward destructive leadership so we could conduct meta-analyses that enabled
power), and “acolytes” (i.e., followers who are influenced by expert us to improve our understanding of destructive leadership. Overall, we
11
found that numerous styles of destructive leadership are investigated in Carnevale, J. B., Huang, L., & Harms, P. D. (2018a). Leader consultation mitigates the
harmful effects of leader narcissism: A belongingness perspective. Organizational
this literature, destructive leadership has many meaningful relationships
Behavior and Human Decision Processes, 146, 76–84.
with antecedents and outcomes throughout its nomological network, Carnevale, J., Huang, L., & Harms, P. (2018b). Speaking up to the “emotional vampire”:
and that the majority of empirical studies rely on examining abusive A conservation of resources perspective. Journal of Business Research, 91, 48–59.
supervision. Darth Vader likely would have warned us that we did not Carreyrou, J. (2018). Bad blood: Secrets and lies in a Silicon Valley startup. New York:
Vintage Books.
truly know “the power of the dark side” of destructive leadership prior to Combs, J. G., Crook, T. R., & Rauch, A. (2019). Meta-analytic research in management:
our study. However, our systematic literature review builds the solid Contemporary approaches, unresolved controversies, and rising standards. Journal of
empirical foundation necessary to advance knowledge within the Management Studies, 56(1), 1–18.
Cortina, L. M., Magley, V. J., Williams, J. H., & Langhout, R. D. (2001). Incivility in the
destructive leadership literature by generating building blocks for workplace: Incidence and impact. Journal of Occupational Health Psychology, 6(1),
nuanced theory development, identifying useful directions for future 64–80.
business research, and facilitating evidence-based recommendations for De Fuentes, C., & Porcuna, R. (2016). Main drivers of consultancy services: A meta-
analytic approach. Journal of Business Research, 69(11), 4775–4780.
policy and practice that leverage and integrate the many insights from De Hoogh, A. H. B., & Den Hartog, D. N. (2008). Ethical and despotic leadership,
this literature. relationships with leader’s social responsibility, top management team effectiveness
and subordinates’ optimism: A multi-method study. The Leadership Quarterly, 19(3),
297–311.
DeSimone, J. A. (2014). When it’s incorrect to correct: A brief history and cautionary
Declaration of Competing Interest note. Industrial and Organizational Psychology, 7(4), 527–531.
DeSimone, J. A., & Harms, P. D. (2018). Dirty data: The effects of screening respondents
who provide low-quality data in survey research. Journal of Business and Psychology,
The authors declare that they have no known competing financial 33, 559–577.
interests or personal relationships that could have appeared to influence DeSimone, J. A., Harms, P. D., & DeSimone, A. J. (2015). Best practice recommendations
the work reported in this paper. for data screening. Journal of Organizational Behavior, 36(2), 171–181.
DeSimone, J. A., Köhler, T., & Schoen, J. L. (2019). If it were only that easy: The use of
meta-analytic research by organizational scholars. Organizational Research Methods,
Acknowledgements 22(4), 867–891.
Duffy, M. K., Ganster, D. C., & Pagon, M. (2002). Social undermining in the workplace.
Academy of Management Journal, 45(2), 331–351.
We thank Sami Berrada, Erin Branca, Nate McLachlan, Katie Miller, Edwards, J. R., & Christian, M. S. (2014). Using accumulated knowledge to calibrate
Ryan Peck, Nataliya Potapenko, Tomas Riga, Lindsey Yonish, and Wil theoretical propositions. Organizational Psychology Review, 4(3), 279–291.
Einarsen, S., Aasland, M. S., & Skogstad, A. (2007). Destructive leadership behaviour: A
helmina Zwennes for their help coding data for our study. We also thank
definition and conceptual model. The Leadership Quarterly, 18(3), 207–216.
Joel Carnevale and three anonymous reviewers for the helpful feedback Einarsen, S., Hoel, H., & Notelaers, G. (2009). Measuring exposure to bullying and
they provided as we improved our manuscript throughout the review harassment at work: Validity, factor structure and psychometric properties of the
process. Negative Acts Questionnaire-Revised. Work & Stress, 23(1), 24–44.
Einarsen, S., & Skogstad, A. (1996). Bullying at work: Epidemiological findings in public
and private organizations. European Journal of Work & Organizational Psychology, 5
Appendix A. Supplementary material (2), 185–201.
Ferris, G. R., Zinko, R., Brouer, R. L., Buckley, M. R., & Harvey, M. G. (2007). Strategic
bullying as a supplementary, balanced perspective on destructive leadership. The
Supplementary data to this article can be found online at https://doi. Leadership Quarterly, 18(3), 195–206.
org/10.1016/j.jbusres.2020.10.037. Field, A. P. (2005). Is the meta-analysis of correlation coefficients accurate when
population correlations vary? Psychological Methods, 10(4), 444–467.
Geyskens, I., Krishnan, R., Steenkamp, J.-B.-E.-M., & Cunha, P. V. (2009). A review and
References evaluation of meta-analysis practices in management research. Journal of
Management, 35(2), 393–419.
Gonzalez-Mulé, & Aguinis, H. (2018). Advancing theory by assessing boundary
Aguinis, H., Banks, G. C., Rogelberg, S. G., & Cascio, W. F. (2020). Actionable
conditions with metaregression: A critical review and best-practice
recommendations for narrowing the science-practice gap in open science.
recommendations. Journal of Management, 44(6), 2246–2273.
Organizational Behavior and Human Decision Processes, 158, 27–35.
Greco, L. M., O’Boyle, E. H., Cockburn, B. S., & Yuan, Z. (2018). Meta-analysis of
Aguinis, H., Dalton, D. R., Bosco, F. A., Pierce, C. A., & Dalton, C. M. (2011). Meta-
coefficient alpha: A reliability generalization study. Journal of Management Studies,
analytic choices and judgment calls: Implications for theory building and testing,
55(4), 583–618.
obtained effect sizes, and scholarly impact. Journal of Management, 37(1), 5–38.
Greco, L. M., O’Boyle, E. H., & Walter, S. L. (2015). Absence of malice: A meta-analysis of
Aguinis, H., Gottfredson, R. K., & Joo, H. (2013). Best-practice recommendations for
nonresponse bias in counterproductive work behavior research. Journal of Applied
defining, identifying, and handling outliers. Organizational Research Methods, 16(2),
Psychology, 100(1), 75–97.
270–301.
Han, G. H., Harms, P. D., & Bai, Y. (2017). Nightmare bosses: The impact of abusive
Aguinis, H., Hill, N. S., & Bailey, J. R. (2020). Best practices in data collection and
supervision on employees’ sleep, emotions, and creativity. Journal of Business Ethics,
preparation: Recommendations for reviewers, editors, and authors. Organizational
145, 21–31.
Research Methods. https://doi.org/10.1177/1094428119836485 (in press).
Harms, P. D., Credé, M., Tynan, M., Leon, M., & Jeung, W. (2017). Leadership and stress:
Andersson, L. M., & Pearson, C. M. (1999). Tit for tat? The spiraling effect of incivility in
A meta-analytic review. The Leadership Quarterly, 28(1), 178–194.
the workplace. Academy of Management Review, 24(3), 452–471.
Harms, P. D., & Spain, S. (2015). Beyond the bright side: Dark personality at work.
Ashford, S. J., & Sitkin, S. B. (2019). From problems to progress: A dialogue on prevailing
Applied Psychology: An International review, 64(1), 15–24.
issues in leadership research. The Leadership Quarterly, 30(4), 454–460.
Hartmans, A., & Leskin, P. (2020). The rise and fall of Elizabeth Holmes, the Theranos
Ashforth, B. (1997). Petty tyranny in organizations: A preliminary examination of
founder awaiting trial on federal charges of ‘massive fraud’. Retrieved from Business
antecedents and consequences. Canadian Journal of Administrative Sciences, 14(2),
Insider.
126–140.
Heggestad, E. D., Scheaf, D. J., Banks, G. C., Hausfeld, M. M., Tonidandel, S., &
Aytug, Z. G., Rothstein, H. R., Zhou, W., & Kern, M. C. (2012). Revealed or concealed?
Williams, E. B. (2019). Scale adaptation in organizational science research: A review
Transparency of procedures, decisions, and judgment calls in meta-analyses.
and best-practice recommendations. Journal of Management, 45(6), 2596–2627.
Organizational Research Methods, 15(1), 103–133.
Hershcovis, M. S. (2011). “Incivility, social undermining, bullying…oh my!”: A call to
Banks, G. C., O’Boyle, E. H., Pollack, J. M., White, C. D., Batchelor, J. H., Whelpley, C. E.,
reconcile constructs within workplace aggression research. Journal of Organizational
Abston, K. A., Bennett, A. A., & Adkins, C. L. (2016). Questions about questionable
Behavior, 32(3), 499–519.
research practices in the field of management. Journal of Management, 42(1), 5–20.
Hoobler, J. M., & Brass, D. J. (2006). Abusive supervision and family undermining as
Barling, J., Christie, A., & Turner, N. (2008). Pseudo-transformational leadership:
displaced aggression. Journal of Applied Psychology, 91(5), 1125–1133.
Towards the development and test of a model. Journal of Business Ethics, 81(4),
Howell, J. M. (1988). Two faces of charisma: Socialized and personalized leadership in
851–861.
organizations. In J. Conger, & R. Kanungo (Eds.), Charismatic leadership: The illusive
Barnes, C. M., Lucianetti, L., Bhave, D. P., & Christian, M. S. (2015). You wouldn’t like
factor in organizational effectiveness (pp. 213–236). San Francisco, CA: Jossey-Bass.
me when I’m sleepy: Leader sleep, daily abusive supervision, and work unit
Hughes-Morgan, M., Kolev, K., & McNamara, G. (2018). A meta-analytic review of
engagement. Academy of Management Journal, 58(5), 1419–1437.
competitive aggressiveness research. Journal of Business Research, 85, 73–82.
Bligh, M. C., Kohles, J. C., Pearce, C. L., Justin, J. E., & Stovall, J. F. (2007). When the
Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: Correcting error and bias in
romance is over: Follower perspectives of aversive leadership. Applied Psychology: An
research findings (2nd ed.). Thousand Oaks, CA: Sage Publications Inc.
International Review, 56(4), 528–557.
Hunter, J. E., Schmidt, F. L., & Le, H. (2006). Implications of direct and indirect range
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2010). A basic
restriction for meta-analysis methods and findings. Journal of Applied Psychology, 91
introduction to fixed-effect and random-effects models for meta-analysis. Research
(3), 594–612.
Synthesis Methods, 1(2), 97–111.
12
Kellerman, B. (2004). Bad leadership: What it is, how it happens, why it matters. Boston, Rauch, A., Rosenbusch, N., Unger, J., & Frese, M. (2016). The effectiveness of cohesive
MA: Harvard Business School Press. and diversified networks: A meta-analysis. Journal of Business Research, 69(2),
Kepes, S., Banks, G. C., McDaniel, M., & Whetzel, D. L. (2012). Publication bias in the 554–568.
organizational sciences. Organizational Research Methods, 15(4), 624–662. Rosenthal, S., & Pittinsky, T. (2006). Narcissistic leadership. The Leadership Quarterly, 17
Kepes, S., McDaniel, M. A., Brannick, M. T., & Banks, G. C. (2013). Meta-analytic reviews (6), 617–633.
in the organizational sciences: Two meta-analytic schools on the way to MARS (the Rothstein, H. R., Sutton, A. J., & Borenstein, M. (2005). Publication bias in meta-analysis:
meta-analytic reporting standards). Journal of Business and Psychology, 28(2), Prevention, assessment and adjustments. West Sussex, England: John Wiley and Sons.
123–143. Schilling, J. (2009). From ineffectiveness to destruction: A qualitative study on the
Krasikova, D. V., Green, S. G., & LeBreton, J. M. (2013). Destructive leadership: A meaning of negative leadership. Leadership, 5(1), 102–128.
theoretical review, integration, and future research agenda. Journal of Management, Schmid, E. A., Verdorfer, A. P., & Peus, C. (2019). Shedding light on leaders’ self-interest:
39(5), 1308–1338. Theory and measurement of exploitative leadership. Journal of Management, 45(4),
Larsson, G., Brandebo, M. F., & Nilsson, S. (2012). Destrudo-L: Development of a short 1401–1433.
scale designed to measure destructive leadership behaviours in a military context. Schmidt, F. L., & Le, H. (2004). Software for the Hunter-Schmidt meta-analysis methods.
Leadership & Organization Development Journal, 62(5), 1556–1582. Iowa City: Department of Management and Organizations, University of Iowa.
LeBreton, J. M., Scherer, K. T., & James, L. R. (2014). Corrections for criterion reliability Schyns, B., & Schilling, J. (2013). How bad are the effects of bad leaders? A meta-analysis
in validity generalization: A false prophet in a land of suspended judgment. Industrial of destructive leadership and its outcomes. The Leadership Quarterly, 24(1), 138–158.
and Organizational Psychology, 7(4), 478–500. Scott, K. D. (2007). The development and test of an exchange-based model of interpersonal
Lian, H., Brown, D. J., Ferris, D. L., Liang, L. H., Keeping, L. M., & Morrison, R. (2014). workplace exclusion (Unpublished doctoral dissertation). Lexington, Kentucky:
Abusive supervision and retaliation: A self-control framework. Academy of University of Kentucky.
Management Journal, 57(1), 116–139. Shaffer, J. A., DeGeest, D., & Li, A. (2016). Tackling the problem of construct
Lian, H., Ferris, D. L., & Brown, D. J. (2012). Does power distance exacerbate or mitigate proliferation: A guide to assessing the discriminant validity of conceptually related
the effects of abusive supervision? It depends on the outcome. Journal of Applied constructs. Organizational Research Methods, 19(1), 80–110.
Psychology, 97(1), 107–123. Shaw, J. B., Erickson, A., & Harvey, M. (2011). A method for measuring destructive
Liao, Z., Yam, K. C., Johnson, R. E., Liu, W., & Song, Z. (2018). Cleansing my abuse: A leadership and identifying types of destructive leaders in organizations. The
reparative response model of perpetrating abusive supervisor behavior. Journal of Leadership Quarterly, 22(4), 575–590.
Applied Psychology, 103(9), 1039–1056. Shepherd, D. A., & Suddaby, R. (2017). Theory building: A review and integration.
Lin, S. H. J., Ma, J., & Johnson, R. E. (2016). When ethical leader behavior breaks bad: Journal of Management, 43(1), 59–86.
How ethical leader behavior can turn abusive via ego depletion and moral licensing. Siddaway, A. P., Wood, A. M., & Hedges, L. V. (2018). How to do a systematic review: A
Journal of Applied Psychology, 101(6), 815–830. best practice guide for conducting and reporting narrative reviews, meta-analyses,
Lipman-Blumen, J. (2005). Toxic leadership: When grand illusions masquerade as noble and meta-syntheses. Annual Review of Psychology, 70, 747–770.
visions. Leader to Leader, 36, 29–36. Snyder, H. (2019). Literature review as a research methodology: An overview and
Liu, D., Liao, H., & Loi, R. (2012). The dark side of leadership: A three-level investigation guidelines. Journal of Business Research, 104, 333–339.
of the cascading effect of abusive supervision on employee creativity. Academy of Spain, S. M., Harms, P., & LeBreton, J. M. (2014). The dark side of personality at work.
Management Journal, 55(5), 1187–1212. Journal of Organizational Behavior, 35(2), S41–S60.
Long, C., & Hays, T. (2018). ‘Pharma Bro’ Martin Shkreli cries in court, is sentences to 7 Tepper, B. J. (2000). Consequences of abusive supervision. Academy of Management
years for securities fraud. Retrieved from Chicago Tribune. Journal, 43(2), 178–190.
Mackey, J. D., Frieder, R. E., Brees, J. R., & Martinko, M. J. (2017). Abusive supervision: Tepper, B. J. (2007). Abusive supervision in work organizations: Review, synthesis, and
A meta-analysis and empirical review. Journal of Management, 43(6), 1940–1965. research agenda. Journal of Management, 33(3), 261–289.
Mackey, J. D., McAllister, C. P., Ellen, B. P., III, & Carson, J. E. (2020). A meta-analysis of Tepper, B. J., Duffy, M. K., & Breaux-Soignet, D. M. (2012). Abusive supervision as
interpersonal and organizational workplace deviance research. Journal of political activity: Distinguishing impulsive and strategic expressions of downward
Management. https://doi.org/10.1177/0149206319862612 (in press). hostility. In G. R. Ferris, & D. C. Treadway (Eds.), Politics in organizations: Theory and
Mackey, J. D., McAllister, C. P., Maher, L. P., & Wang, G. (2019). Leaders and followers research considerations (pp. 191–212). New York: Routledge: Taylor & Francis Group.
behaving badly: A meta-analytic examination of curvilinear relationships between Tepper, B. J., Duffy, M. K., Henle, C. A., & Lambert, L. S. (2006). Procedural injustice,
destructive leadership and followers’ workplace behaviors. Personnel Psychology, 72 victim precipitation, and abusive supervision. Personnel Psychology, 59(1), 101–123.
(1), 3–47. Tepper, B. J., & Henle, C. A. (2011). A case for recognizing distinctions among constructs
Martin, S. R., Côté, S., & Woodruff, T. (2016). Echoes of our upbringing: How growing up that capture interpersonal mistreatment in work organizations. Journal of
wealthy or poor relates to narcissism, leader behavior, and leader effectiveness. Organizational Behavior, 32(3), 487–498.
Academy of Management Journal, 59(6), 2157–2177. Tepper, B. J., Moss, S. E., & Duffy, M. K. (2011). Predictors of abusive supervision:
Martinko, M. J., Harvey, P., Brees, J. R., & Mackey, J. (2013). A review of abusive Supervisor perceptions of deep-level dissimilarity, relationship, conflict, and
supervision research. Journal of Organizational Behavior, 34(S1), S120–S137. subordinate performance. Academy of Management Journal, 54(2), 279–294.
Martinko, M. J., Harvey, P., & Mackey, J. D. (2014). Conceptual and empirical confounds Tepper, B. J., Simon, L., & Park, H. M. (2017). Abusive supervision. Annual Review of
in the organizational sciences: An explication and discussion. Journal of Organizational Psychology and Organizational Behavior, 4, 123–152.
Organizational Behavior, 35(8), 1052–1063. Thoroughgood, C. N., & Padilla, A. (2013). Destructive leadership and the Penn State
McAllister, C. P., Mackey, J. D., & Perrewé, P. L. (2018). The role of self-regulation in the scandal: A toxic triangle perspective. Industrial and Organizational Psychology, 6(2),
relationship between abusive supervision and job tension. Journal of Organizational 144–149.
Behavior, 39(4), 416–428. Thoroughgood, C. N., Padilla, A., Hunter, S. T., & Tate, B. W. (2012). The susceptible
Min, H., Pavisic, I., Howald, N., Highhouse, S., & Zickar, M. J. (2019). A systematic circle: A taxonomy of followers associated with destructive leadership. The
comparison of three sadism measures and their ability to explain workplace Leadership Quarterly, 23(5), 897–917.
mistreatment over and above the dark triad. Journal of Research in Personality, 82, Thoroughgood, C. N., Sawyer, K. B., Padilla, A., & Lunsford, L. (2018). Destructive
1–13. leadership: A critique of leader-centric perspectives and toward a more holistic
Mitchell, M. S., & Ambrose, M. L. (2007). Abusive supervision and workplace deviance definition. Journal of Business Ethics, 151(3), 627–649.
and the moderating effects of negative reciprocity beliefs. Journal of Applied Thoroughgood, C. N., Tate, B. W., Sawyer, K. B., & Jacobs, R. (2012). Bad to the bone:
Psychology, 92(4), 1159–1168. Empirically defining and measuring destructive leader behavior. Journal of
Mosier, C. I. (1943). On the reliability of a weighted composite. Psychometrika, 8(3), Leadership & Organizational Studies, 19(2), 230–255.
161–168. Van Dijk, W. W., Ouwerkerk, J. W., Goslinga, S., & Nieweg, M. (2005). Deservingness
O’Boyle, E. H., Banks, G. C., & Gonzalez-Mulé, E. (2017). The Chrysalis Effect: How ugly and schadenfreude. Cognition and Emotion, 19(6), 933–939.
initial results metamorphosize into beautiful articles. Journal of Management, 43(2), Viswesvaran, C., Ones, D. S., Schmidt, F. L., Le, H., & Oh, I.-S. (2014). Measurement error
376–399. obfuscates scientific knowledge: Path to cumulative knowledge requires corrections
Ones, D. S., Viswesvaran, C., & Schmidt, F. L. (2017). Realizing the full potential of for unreliability and psychometric meta-analysis. Industrial and Organizational
psychometric meta-analysis for a cumulative science and practice of human resource Psychology, 7(4), 507–518.
management. Human Resource Management Review, 27(1), 201–215. Vogel, R. M., & Mitchell, M. S. (2017). The motivational effects of diminished self-esteem
Padilla, A., Hogan, R., & Kaiser, R. B. (2007). The toxic triangle: Destructive leaders, for employees who experience abusive supervision. Journal of Management, 43(7),
susceptible followers, and conducive environments. The Leadership Quarterly, 18(3), 2218–2251.
176–194. Wang, G., Harms, P. D., & Mackey, J. D. (2015). Does it take two to tangle? Subordinates’
Park, H., Hoobler, J. M., Wu, J., Liden, R. C., Hu, J., & Wilson, M. S. (2019). Abusive perceptions of and reactions to abusive supervision. Journal of Business Ethics, 131,
supervision and employee deviance: A multifoci justice perspective. Journal of 487–503.
Business Ethics, 158(4), 1113–1131. Wang, G., Van Iddekinge, C. H., Zhang, L., & Bishoff, J. (2019). Meta-analytic and
Paterson, T. A., Harms, P. D., Steel, P., & Credé, M. (2016). An assessment of the primary investigations of the role of followers in ratings of leadership behavior in
magnitude of effect sizes: Evidence from 30 years of meta-analysis in management. organizations. Journal of Applied Psychology, 104(1), 70–106.
Journal of Leadership & Organizational Studies, 23(1), 66–81. Wood, D., Harms, P. D., Lowman, G. H., & DeSimone, J. A. (2017). Response speed and
Pearce, C. L., & Sims, H. P. (2002). Vertical versus shared leadership as predictors of the response consistency as mutually validating indicators of data quality in online
effectiveness of change management teams: An examination of aversive, directive, samples. Social Psychological and Personality Science, 8(4), 454–464.
transactional, transformational, and empower leader behaviors. Group Dynamics: Xu, E., Huang, X., Lam, C. K., & Miao, Q. (2012). Abusive supervision and work
Theory, Research, and Practice, 6(2), 172–197. behaviors: The mediating role of LMX. Journal of Organizational Behavior, 33(4),
531–543.
13
Yam, K. C., Christian, M. S., Wei, W., Liao, Z., & Nai, J. (2018). The mixed blessing of Managerial Sciences from Georgia State University, and a B.S. in Civil Engineering from
leader sense of humor: Examining costs and benefits. Academy of Management Auburn University. His research has been published in numerous academic journals,
Journal, 61(1), 348–369. including the Journal of Management, Journal of Organizational Behavior, Journal of Man
Yu, L., Duffy, M. K., & Tepper, B. J. (2018). Consequences of downward envy: A model of agement Studies, and The Leadership Quarterly. Professor Ellen researches organizational
self-esteem threat, abusive supervision, and supervisory leader self-improvement. behavior topics related to social influence in organizations. His primary focus areas are
Academy of Management Journal, 61(6), 2296–2318. leadership and organizational politics. He has taught courses on organizational behavior,
Zapata, C. P., & Hayes-Jones, L. C. (2019). The consequences of humility for leaders: A negotiation, and leadership.
double-edged sword. Organizational Behavior and Human Decision Processes, 152,
47–63.
Dr. Charn P. McAllister is an Assistant Professor in the Department of Management in the
Zhang, Y., & Bednall, T. C. (2016). Antecedents of abusive supervision: A meta-analytic
W. A. Franke College of Business at Northern Arizona University. He earned a Ph.D. in
review. Journal of Business Ethics, 139(3), 455–471.
Business Administration from Florida State University, M.A. in Humanities from American
Zhang, Y., Liu, X., Xu, S., Yang, L.-Q., & Bednall, T. C. (2019). Why abusive supervision
Military University, and a B.S. in Engineering Psychology from the United States Military
impacts employee OCB and CWB: A meta-analytic review of competing mediating
Academy at West Point. His research has been published in numerous academic journals,
mechanisms. Journal of Management, 45(6), 2474–2497.
including the Journal of Applied Psychology, Journal of Management, and Personnel Psy
chology. His current research interests include interpersonal mistreatment, political skill,
Dr. Jeremy D. Mackey is a Dean’s Fellow and Associate Professor of Management in the and stress. Professor McAllister’s research focuses on the various ways employees attempt
Harbert College of Business at Auburn University. He earned a Ph.D. in Business Admin to navigate and thrive in the workplace. He has taught courses on leadership, negotiations,
istration from Florida State University and a B.S. in Business Administration from Virginia and organizational behavior.
Tech. His research has been published in numerous academic journals, including the
Journal of Applied Psychology, Strategic Management Journal, Journal of Management, and
Ms. Katherine C. Alexander is a Ph.D. student in the Department of Management in the
Personnel Psychology. His current research interests include destructive leadership, inter
Harbert College of Business at Auburn University. She earned a M.B.A. from Auburn
personal mistreatment, meta-analysis, and measure development. Professor Mackey’s
University and a B.S. in Economics, Philosophy, and Political Economy from the University
research examines why and how interpersonal mistreatment affects professional re
of Alabama at Birmingham. Her research has been published in the Journal of Business
lationships and counterproductive work behavior. Much of his current research focuses on
Ethics. Her current research interests include interpersonal mistreatment, insubordination,
how to use meta-analysis and measure development to meaningfully advance our under
destructive leadership, and meta-analysis. Overall, Katie’s research examines destructive
standing of interpersonal mistreatment at work. He has taught courses about organiza
leadership and its impacts in organizations. Her research has resulted in the development
tional behavior and human resource management.
of a measure of insubordination that can facilitate the development of a new research
stream. She also conducts meta-analyses of interpersonal mistreatment topics that refine
Dr. B. Parker Ellen III is an Assistant Professor in the Management and Organizational our understanding of its true impacts on perpetrators and victims. She has taught courses
Development Group of the D’Amore-McKim School of Business at Northeastern Univer on the principles of management.
sity. He earned a Ph.D. in Business Administration from Florida State University, M.S. in
14

The Dark Side of Leadership: A Systematic Literature Review and Meta-Analysis of Destructive Leadership Research

Uploaded by

Copyright:

Available Formats

The Dark Side of Leadership: A Systematic Literature Review and Meta-Analysis of Destructive Leadership Research

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Dark Side of Leadership: A Systematic Literature Review and Meta-Analysis of Destructive Leadership Research

Uploaded by

Copyright:

Available Formats

Journal of Business Research xxx (xxxx) xxx

Contents lists available at ScienceDirect

Journal of Business Research