Journal of Applied Psychology
2015, Vol. 100, No. 4, 1040 –1055
© 2014 American Psychological Association
0021-9010/15/$12.00 http://dx.doi.org/10.1037/apl0000002
Third Parties’ Reactions to the Abusive Supervision of Coworkers
Marie S. Mitchell
Ryan M. Vogel
University of Georgia
Pennsylvania State University–Erie
Robert Folger
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
University of Central Florida
This research examines 3rd parties’ reactions to the abusive supervision of a coworker. Reactions were
theorized to depend on 3rd parties’ beliefs about the targeted coworker and, specifically, whether the
target of abuse was considered deserving of mistreatment. We predicted that 3rd parties would experience
anger when targets of abuse were considered undeserving of mistreatment; angered 3rd parties would
then be motivated to harm the abusive supervisor and support the targeted coworker. Conversely, we
predicted that 3rd parties would experience contentment when targets of abuse were considered deserving
of mistreatment; contented 3rd parties would then be motivated to exclude the targeted coworker.
Additionally, we predicted that 3rd parties’ moral identity would moderate the effects of 3rd parties’
experienced emotions on their behavioral reactions, such that a strong moral identity would strengthen
ethical behavior (i.e., coworker support) and weaken harmful behavior (i.e., supervisor-directed deviance,
coworker exclusion). Moderated mediation results supported the predictions. Implications for theory and
practice are discussed.
Keywords: abusive supervision, exclusion beliefs, supervisor-directed deviance, coworker support,
coworker exclusion
attention to understanding how abusive supervision affects the
larger work environment and, specifically, to third party employees. This research has shown that witnessing abuse can motivate
different third party behavioral reactions, such as counterproductive behavior that is harmful to coworkers (Harris, Harvey, Harris,
& Cast, 2013) and prosocial behavior intended to help the target of
abuse (Priesemuth, 2013). What is less clear, though, is why third
parties react one way versus another and which factors influence
their different reactions.
Traditional applications of deontic justice theory (Folger, 1998,
2001) suggest that witnessing abusive supervision would anger third
parties. The experience of anger would then motivate these third
parties to redress the transgression by harming the perpetrator, particularly if these third parties held a weak moral identity (defined as
the degree to which moral principles are strongly aligned with one’s
sense of self; Aquino & Reed, 2002) because third parties with a weak
moral identity are less concerned about maintaining kind, caring, and
friendly behavior toward others (see Jennings, Mitchell, & Hannah, in
press, for a review). An implicit assumption to the traditional application of deontic justice theory, however, is that all third parties
become angered by another’s mistreatment because all third parties
believe mistreating others is unfair.
We present an alternative view and theorize that not all third
parties believe it is unfair to mistreat specific coworkers. Our
arguments are consistent with research that has shown that employees sometimes find coworkers to be provocative targets of
abuse (Kim & Glomb, 2010; Lam, Van der Vegt, Walter, &
Huang, 2011). Moreover, recent work has shown that third parties’
perceptions of fairness can be biased (Blader, Wiesenfeld, Fortin,
& Wheeler-Smith, 2013). This research has shown that third
parties’ reactions depend on their evaluation of the recipient of the
Many of us have been exposed to organizational authorities
acting cruelly or abusively toward employees. Both the popular
press (Pachter, 2003; Sutton, 2010) and academic studies (Cortina,
Magley, Williams, & Langhout, 2001; Neuman & Keashly, 2003)
have shown that employees are often victim to egregious and
demeaning acts by supervisors, termed abusive supervision (subordinates’ perceptions of sustained hostility against them by a
supervisor; Tepper, 2000). Given its prevalence, researchers have
sought to understand the consequences of abusive supervision and
found it has a significant toll on its victims. For instance, abusive
supervision has been found to heighten victims’ psychological
distress, emotional exhaustion, self-regulation impairment, and job
burnout (see Martinko, Harvey, Brees, & Mackey, 2013, and
Tepper, 2007, for reviews). Recently, researchers (e.g., Harris,
Harvey, Harris, & Cast, 2013; Priesemuth, 2013) have turned their
This article was published Online First September 22, 2014.
Marie S. Mitchell, Department of Management, Terry College of Business, University of Georgia; Ryan M. Vogel, Department of Management,
The Behrend College, Pennsylvania State University–Erie; Robert Folger,
Department of Management, College of Business Administration, University of Central Florida.
We would like to acknowledge the financial support provided by the
Terry–Sanford research grant from the Terry College of Business, University of Georgia. We would also like to thank members of the Behavioral
Ethics Group for their valuable feedback during the 2012 workshop series
in Orlando, Florida.
Correspondence concerning this article should be addressed to Marie S.
Mitchell, Department of Management, Terry College of Business, University of Georgia, 404 Brooks Hall, Athens, GA 30602. E-mail: msmitche@
uga.edu
1040
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
treatment. Third parties become angered by unfair outcomes given
to recipients who they positively evaluate and become contented
by unfair outcomes given to recipients who they negatively evaluate. Given these findings, we argue that it is possible for some
third parties not to experience anger as is implicit in traditional
deontic justice reasoning but instead may experience contentment
from the abuse. We theorize that the different emotions elicited
will motivate different behavioral reactions from third parties.
Our work, therefore, extends deontic principles to consider third
parties’ evaluation of the target of abuse. We develop and test a
model, which integrates principles from deontic justice theory
(Folger, 2001, 2012) with scope of justice theory (Deutsch, 1974;
Opotow, 1990, 1995), to explain differential third party reactions
to the abusive supervision of a coworker. Deontic justice theory
suggests that mistreating another (i.e., abusive supervision) motivates third parties’ to react according to what they believe is “just”
and fair. We draw from scope of justice theory to explain that third
parties’ beliefs about what is “just” depend on their perceptions of
the target of abuse. Specifically, we propose a critical factor that
influences third party reactions is their exclusion beliefs, defined
as individuals’ beliefs about the degree to which another person
deserves disrespectful or unfair treatment (Opotow, 2012).
We theorize that witnessing abusive supervision elicits different emotions and that the type of emotion evoked depends on
third parties’ exclusion beliefs about the target of abuse. Third
parties with lower exclusion beliefs perceive the target as
undeserving of mistreatment, and, thus, these third parties
should experience anger—a sense of hostility about the victim’s
abuse. Third parties with higher exclusion beliefs perceive the
target as deserving of mistreatment, and, thus, these third parties should experience contentment—a sense of satisfaction
with the victim’s abuse. Because each emotion motivates
unique behavioral responses, we predict third parties who experience anger will be motivated to harm the supervisor (i.e.,
supervisor-directed deviance) and offer support to the target of
abuse (i.e., coworker support). Third parties who experience
contentment will be motivated to exclude the target (i.e., coworker exclusion). Consistent with third party research (see
Skarlicki & Kulik, 2004) and moral self theory (Blasi, 1984,
1041
2004), we also predict that the effects of the experienced
emotions prompted from the abuse on third parties’ behavioral
reactions will be moderated by third parties’ moral identity.
Thus, we predict abusive supervision will indirectly influence
third parties’ behavioral reactions, as mediated by emotion and
moderated by their exclusion beliefs and moral identity (see
Figure 1).
Our work makes three primary contributions. First, our research
extends the literature by explaining why abusive supervision motivates different third party reactions. Much of the work on abusive
supervision has examined its effects from the victim’s or supervisor’s perspective (see Martinko et al., 2013 and Tepper, 2007, for
reviews). Although emerging work has demonstrated abuse can be
detrimental to third parties (e.g., Harris et al., 2013; Miner-Rubino
& Cortina, 2007), we suggest that not all third parties react to
abuse in the same way because not all third parties hold the same
beliefs about the person being abused. Second, we extend the
literature on third party reactions to mistreatment. Mitchell, Vogel,
and Folger (2012) called on researchers to extend deontic reasoning to consider a broader range of third party reactions. Our work
addresses this call. We provide an important clarification to deontic
justice reasoning, namely, that third party reactions depend on third
parties’ beliefs about the target of the treatment. Therefore, we refine
deontic justice theory to consider important boundary conditions that
impact third party reactions to others’ mistreatment. Third, our work
contributes to the workplace deviance literature in explaining how
abusive supervision can promote a spiral of destructive work behavior
(e.g., supervisor-directed deviance, coworker exclusion). Research
has shown that these types of behaviors are costly to organizations
(Detert, Treviño, Burris, & Andiappan, 2007; Dunlop & Lee, 2004).
Thus, it is important to understand how abusive supervision incites
third parties’ harmful behavior in order for decision makers to mitigate it.
Theoretical Overview
The Greek word deon means duty or obligation. Folger (2001)
thus proposed a deontic justice theory, which suggests that witnessing another’s mistreatment engenders third parties’ sense of
Figure 1. Theoretical model.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1042
MITCHELL, VOGEL, AND FOLGER
obligation to abide by what they believe is fair and right. The
theory specifies that witnessing mistreatment creates a deontic
motivation in which third parties experience other-focused emotions— emotions related to the person being abused—and that the
elicited emotion from the mistreatment creates action tendencies
that align with what the third parties believe is just.
Folger (2001) proposed that the primary emotion third parties
experience from witnessing another’s mistreatment is anger. Anger
is experienced as a moral emotion (Hoffman, 1989, 2000; Montada
& Schneider, 1989); it is an emotion elicited by considering the
interests and welfare of another person (Haidt, 2003). Anger
derives from third parties’ sense that the witnessed treatment
violated moral principles associated with individual rights and
liberty (Rozin, Lowery, Imada, & Haidt, 1999). Thus, anger reflects third parties’ outrage about what is considered an unjustifiable event. It is in this way that anger forms as an honorable
impulse; it is an automatic and convincing gut feeling about the
other person’s treatment (Haidt, 2003).
Although some third parties may experience anger, we theorize
that not all will because not all third parties may believe it is unfair
to mistreat certain individuals. In regard to this point, Blader,
Wiesenfeld, Rothman, and Wheeler-Smith (2010) argued that third
parties’ justice judgments are biased by their evaluation of the
observed target. They proposed that the degree to which objectively fair or unfair outcomes would be judged as fair or unfair
depends on whether these third parties hold positive or negative
attitudes about the recipient of the outcome. When third parties
hold a positive evaluation of the recipient, their reactions should be
congruent with their social affiliation with the recipient. Conversely, when third parties hold a negative evaluation of the
recipient, their reactions should be consistent with their aversion to
the recipient. In support of these ideas, Blader et al. (2013) found
that third parties who held a negative evaluation of the recipient of
an objectively unfair outcome judged the outcome as more fair and
were content with the outcome. By comparison, those who held a
positive evaluation of the recipient judged an objectively unfair
outcome as less fair and were angered by the outcome.
We draw from Blader et al.’s (2010, 2013) ideas and contend
that third parties may experience anger or contentment about
abusive supervision, depending on their evaluations of the target of
abuse. Indeed, scholars (Tepper, Moss, & Duffy, 2011) have
recently suggested that some supervisors view certain subordinates
as provocative targets for abuse; these targets are considered
difficult, aggravating, and annoying, and so abusive supervision is
merited. Further, some empirical work has shown that employees
sometimes believe certain coworkers deserve mistreatment because these coworkers engage in behavior that makes them likely
targets for abuse (Kim & Glomb, 2010; Lam et al., 2011). Thus, it
is possible for some third parties to find certain coworkers as
deserving of mistreatment.
Scope of justice theory (Deutsch, 1974; Opotow, 1990, 1995)
explains why this might occur. The theory proposes that individuals’ exclusion beliefs— beliefs about whether another person
deserves to be treated disrespectfully and unfairly (Opotow,
2012)—influence reactions to witnessed mistreatment. An evaluation of deservingness indicates a sense of “just deserts.” The
person is viewed as holding certain characteristics or has acted in
a manner that suggests he or she has earned mistreatment. Consequently, when individuals are considered deserving of mistreat-
ment, the general opposition to acts that violate human rights
becomes suspended. Excluded individuals are considered nonentities, expendable, and undeserving of respect. Thus, individuals
who hold strong exclusion beliefs for another believe it is morally
permissible to harm the person. In contrast, low exclusion beliefs
suggest that the person should be treated in a manner similar to
how the third party should be treated: with respect and dignity.
Included individuals are regarded as approximate equals and,
therefore, are believed to be undeserving of mistreatment. Altogether, scope of justice theory proposes that third parties have
more supportive reactions for targets of abuse when third parties
hold weak exclusion beliefs about the target, and third parties have
less supportive reactions for targets of abuse when third parties
hold strong exclusion beliefs about the target.
Third Parties’ Exclusion Beliefs and Emotions Evoked
From Abusive Supervision
Other-focused emotions of anger and contentment depend on
the relative cognitive appraisal associated with each emotion
(Blader et al., 2010, 2013). These emotions are contingent on third
parties’ evaluation “of the person about whom that emotion is felt”
(Blader et al., 2010, p. 35). Blader et al. (2013) argued that
observers’ evaluation of the recipient serves as a necessary “starting point or anchor of their own assessment and reaction to the
target’s encounter” (p. 64). Based on scope of justice theory
(Opotow, 1990, 1995), we suggest that how observers respond to
another’s mistreatment depends on their exclusion beliefs. The
theory proposes that exclusion beliefs influence emotional reactions to witnessed mistreatment and that both anger and contentment are modulated by third parties’ evaluation of the deservedness of the person who is being mistreated (Deutsch, 1990;
Feather, 2006). Therefore, we posit that a critical determinant to
whether anger or contentment would be experienced is third parties’ beliefs regarding whether the target of abuse is considered
deserving of the mistreatment (i.e., exclusion beliefs).
Along these lines, Feather (2006) suggested that emotions elicited by witnessing another person’s mistreatment are based on
primitive constructions— good things should happen to good people, and bad things should happen to bad people. Consistent with
Feather’s arguments, moral emotions scholars (Hoffman, 2000;
Montada & Schneider, 1989) have argued that anger may occur in
an automatic and rapid fashion when third parties believe the
person is undeserving of mistreatment. Mullen and Skitka (2006)
argued that third parties experience anger from certain events
because the witnessed events are inconsistent with their point of
view. In our study, we suggest that anger is possible for third
parties who believe the target is undeserving of mistreatment. To
be sure, experimental evidence has shown that when third parties
view a victim as less deserving of mistreatment, they become
angry (see Hafer & Bégue, 2005, for a review).
Feather (2006) also argued that if the person is viewed as deserving of mistreatment, a sense of satisfaction may be possible, such as
schadenfreude (taking pleasure from another’s misfortune). This
sense of satisfaction is experienced because these third parties desire
to denigrate or cut down the focal person (Feather, 2008). The person
is viewed as inferior to the observer in some way (Jankowski &
Takahshi, 2014; Takahashi et al., 2009). As such, experienced contentment is egocentric (Montada & Schneider, 1989), representing an
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
unfortunate reflection of malevolence in human nature (Leach,
Spears, Branscombe, & Doosje, 2003). We theorize that high exclusion beliefs about the target of abuse will enhance feelings of contentment when these targets are abused by supervisors. Empirical
findings have supported these arguments and have shown that third
parties experience contentment when they believe a target is deserving of mistreatment (Feather, 2008).
Thus, we theorize that the emotion third parties experience from
witnessing abusive supervision will align with their exclusion
beliefs about the target of abuse. Lower exclusion beliefs suggest
that the person being abused should be treated like the third
party—in a dignified and respectful manner. Consequently, witnessing the abuse of a coworker who is not excluded would not sit
well with these third parties and would elicit anger. In contrast, higher
exclusion beliefs suggest that the third party believes it is appropriate
to harm the target. Excluded targets are presumed to have earned any
mistreatment that befalls them. Consequently, third parties who hold
high exclusion beliefs for targets should experience contentment from
the abuse because they would be satisfied that these targeted coworkers are getting their “just deserts.”
Hypothesis 1: Third parties’ exclusion beliefs about the targeted coworker will moderate the relationship between abusive supervision of the coworker and third parties’ emotions,
such that abusive supervision of the coworker will more
strongly and positively impact third parties’ (a) anger when
exclusion beliefs are lower rather than higher and (b) contentment when exclusion beliefs are higher rather than lower.
Third Parties’ Emotions and Behavioral Reactions to
Abusive Supervision of Coworkers
We further suggest that the motivational effects of exclusion beliefs
from witnessed abusive supervision will create specific action tendencies that are consistent with third parties’ beliefs about the target of
abuse. Emotions create a state of readiness that guides behavior
(Frijda, 1988). Folger (2001) argued that emotions elicited from
witnessing another’s mistreatment are based on what third parties
believe is appropriate; these emotions, then, motivate third parties to
engage in behavior that is consistent with what they feel is right. We
suggest that third parties who become angry by abusive supervision of
a coworker (or those who hold low exclusion beliefs about the target
of abuse) will become motivated to restore fairness for the target of
abuse and treat the target with dignity. Third parties who experience
contentment by the abuse (or those who hold high exclusion beliefs
about the target of abuse) will be motivated to perpetuate mistreatment against the target of abuse.
Indeed, theory suggests emotions provoke affect-driven behaviors (e.g., Frijda, 1988; Weiss & Cropanzano, 1996). Affect-driven
behaviors organize “behavior around the demands of the precipitating situation” (Weiss & Cropanzano, 1996, p. 54). The emotions
instigated from the situation generate involuntary and automatic
intentions and motivations. Both anger and contentment have
specific action tendencies (Haidt, 2003), which we suggest align
with third parties’ exclusion beliefs about the target of abuse.
Anger instigated from witnessing harm against another motivates two different action tendencies. The first action tendency
angered third parties experience is to “attack, humiliate, or otherwise get back at the person who is perceived to be acting unfairly
1043
or immorally” (Haidt, 2003, p. 856). Anger energizes third parties
to fight for and protect the individual who has been victimized
(Shaver, Schwartz, Kirson, & O’Connor, 1987). The greater the
anger experienced, the stronger the restorative impulses (Darley &
Pittman, 2003). For instance, experienced anger from perceived
injustices has been found to motivate third parties to harm transgressors (O’Gorman, Wilson, & Miller, 2005), even when doing so
incurs personal costs (e.g., Skarlicki & Rupp, 2010; Turillo,
Folger, Lavelle, Umphress, & Gee, 2002). The second action
tendency that angered third parties experience is to help and
support victims (van Zomeren & Lodwijkx, 2005). For example,
research has shown that anger influences third parties to engage in
prosocial behavior, such as demonstrating support and care for the
victim (Earnshaw, Pitpitan, & Chaudoir, 2011; van Zomeren &
Lodwijkx, 2005; van Zomeren, Spears, Fischer, & Leach, 2004).
Altogether these arguments suggest that third parties who experience anger from abusive supervision of a coworker (i.e., those who
hold weak exclusion beliefs about the target of abuse) may direct
harmful behaviors toward the supervisor (i.e., supervisor-directed
deviance) and may direct supportive behaviors toward the target of
abuse (i.e., coworker support).
In contrast, contentment experienced from another person’s mistreatment does not promote beneficial actions toward mistreated individuals. In fact, Haidt (2003) argued that positive emotions evoked
from another’s misfortune seem to have no prosocial tendencies
whatsoever. Instead, these emotions motivate behaviors that translate
into direct derogation and exclusion of the source of contentment
(Opotow & Weiss, 2000; Spears & Leach, 2004). For example,
experienced contentment from another’s misfortune motivates individuals to distance themselves from and socially exclude the target
(Feather, 2008; Sundie, Ward, Beal, Chin, & Geiger-Oneto, 2009).
Because third parties who experience contentment believe these
abused individuals are deserving of the mistreatment, they would
likely be motivated to perpetuate harm against the target of abuse by
excluding them (i.e., coworker exclusion).
The Influence of Moral Identity on Third Parties’
Behavioral Reactions
Scholars who integrate deontic justice reasoning into models
examining third party behavioral reactions often consider the influence of moral identity (see Jennings et al., in press). This is
because a strong moral identity has been theorized and demonstrated to be a trait that effectively regulates individuals’ moral
behavior (cf. Aquino, McFerran, & Laven, 2011; Aquino, Reed,
Stewart, & Shapiro, 2005). Moral identity is defined as the extent
to which moral characteristics, such as being caring, kind, and
friendly, are internalized into an individual’s self-concept (Aquino
& Reed, 2002). Moral self theory (Blasi, 1983; Erikson, 1964)
articulates that a strong moral identity generates a motivation that
“serves to reinforce individuals’ ethical stance and enables them to
react to different situations more effectively and ethically” (Jennings et al., in press, p. 49). A strong moral identity commands “a
commitment to one’s sense of self to lines of action that promote
and protect the welfare of others” (Hart, Atkins, & Ford, 1998, p.
515). The motivational potency of moral identity inspires individuals to maintain behavioral self-consistency and preserve the integrity of their moral self-construal (Aquino & Reed, 2002; Blasi,
1984). By contrast, individuals with a weak moral identity do not
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1044
MITCHELL, VOGEL, AND FOLGER
hold themselves to that moral benchmark and are, therefore, less
likely to consider the ethicality or implications of their behavior
toward others.
We theorize that the self-regulatory qualities of moral identity
are particularly important to consider in the context of affectdriven behavior. Scholars have argued and found that affective
reactions that occur as a result of observed transgressions (e.g.,
abusive supervision) happen rather quickly (Blasi, 2004; Monin,
Pizarro, & Beer, 2007). Emotions elicited by these situations are
automatic and represent a reaction rather than deliberation about
the situation (Monin et al., 2007). For this reason, Haidt (2001)
argued that emotional reactions to such events create a “moral
dumbfounding” effect, wherein the initial experience of emotion
overwhelms cognition and, consequently, circumvents rational
processes that might consider whether the person should be experiencing the emotion in the first place. Emotions elicited that
contrast with the person’s sense of morality (i.e., his or her moral
identity) are discomforting. This tension motivates individuals to
reduce affect-driven desires that are inconsistent with their sense
of self and motivates them to strive to regulate their behavior with
their sense of morality. That is, individuals who strongly define
their sense of self with the characteristics of being caring, kind,
and friendly (i.e., individuals with a strong moral identity) strive to
ensure that their actions are consistent with those core characteristics (Jordan, Mullen, & Murnighan, 2011).
Thus, although emotion exerts a strong motivational force, it alone
does not paint the entire picture of moral motivation (Blasi, 2004;
Hardy & Carlo, 2005). In fact, researchers have found that emotion
only moderately accounts for (un)ethical behavior (Eisenberg &
Miller, 1987). This implies that individuals are not blindly guided
only by their emotions. Moral self theory (Blasi, 1984, 2004) suggests
that moral identity exerts a strong influence in diffusing the tension
associated with emotions that motivate unethical behavior and one’s
moral sense of self. A strong moral identity encourages the regulation
of behavior to meet moral standards. It is in this way that a strong
moral identity assists moral motivation and helps third parties do the
right thing, regardless of the emotions they may experience. Hence,
third parties who hold a strong moral identity should be able to control
their desires to act on their affect-driven tendencies and resist emotionally charged desires to behave inappropriately; instead, a strong
moral identity should prompt individuals to engage in behavior
aligned with the principles of care and kindness.
Hypothesis 2: Moral identity will moderate the effects of emotions on behaviors such that (a) the positive relationship between
anger and supervisor-directed deviance will be stronger when
moral identity is lower rather than higher, (b) the positive relationship between anger and coworker support will be stronger
when moral identity is higher rather than lower, and (c) the
positive relationship between contentment and coworker exclusion will be stronger when moral identity is lower rather than
higher.
These arguments suggest that moral identity influences third parties’ motivation to act on their emotional reactions from witnessing
coworkers be abused by a supervisor. Accordingly, moral identity
should assist third parties in maintaining ethical reactions to witnessed
abusive supervision. In particular, we theorize that third parties who
hold lower exclusion beliefs about the target of abuse will likely
become angry from witnessing the abusive supervision; this anger
will heighten desires to harm the transgressor (i.e., supervisor-directed
deviance) and help the victim (i.e., coworker support). Third parties
who hold higher exclusion beliefs about the target of abuse should
experience contentment, which should motivate perpetuating mistreatment against the target of abuse through coworker exclusion. Although
offering support to the target of abuse aligns with moral principles of care
and kindness, the desire to harm the supervisor and exclude the coworker
are inconsistent with a strong moral identity. Therefore, third parties with
a strong moral identity should become particularly motivated to resist the
temptation to act badly and instead take the high road and behave more
ethically.
In sum, we predict that the influence of abusive supervision of
a coworker should indirectly impact third parties’ behavioral reactions via their emotions elicited by witnessing abusive supervision but that their reactions will depend on third parties’ exclusion
beliefs about the target of abuse and their moral identity. When
anger is elicited by abusive supervision (i.e., third parties’ exclusion beliefs are low), third parties with a strong moral identity will
be less likely to act on their desire to punish the supervisor (i.e.,
supervisor-directed deviance) and more likely to act on their desire
to help the target of abuse (i.e., coworker support) because doing
so would be consistent with the moral principles of being a kind,
caring, and friendly person. When contentment is elicited (i.e.,
third parties’ exclusion beliefs are high), third parties with a strong
moral identity will be less likely to act on their desires to exclude
the target of abuse because exercising restraint in that fashion
would also be consistent with the moral principles of being a kind,
caring, and friendly person.
Hypothesis 3(a): The indirect effect of abusive supervision of
the coworker via anger on supervisor-directed deviance will
be more strongly positive when exclusion beliefs are lower
(rather than higher) and moral identity is lower (rather than
higher).
Hypothesis 3(b): The indirect effect of abusive supervision of
the coworker via anger on coworker support will be more
strongly positive when exclusion beliefs are lower (rather than
higher) and moral identity is higher (rather than lower).
Hypothesis 3(c): The indirect effect of abusive supervision of
the coworker via contentment on coworker exclusion will be
more strongly positive when exclusion beliefs are higher
(rather than lower) and moral identity is lower (rather than
higher).
Method
Sample and Procedure
Data were collected from full-time employees of various organizations located in the United States in multiple industries, including finance, insurance, education, healthcare, information
technology, and retail. Surveys were administered via the Internet
and across two time periods. Undergraduate business administration students were asked to serve as organizational contacts for
course credit. Students were asked to recruit up to two adults who
worked full-time in different organizations and who worked in an
environment with other coworkers. The researchers contacted the
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
participants directly. This convenience sampling data collection
technique has been used successfully by a variety of researchers
(e.g., Grant & Mayer, 2009; Greenbaum, Mawritz, & Eissa, 2012).
We received the contact information for 416 potential study
participants, of whom 358 participated in the Time 1 survey
(86.1% response rate) and 322 participated in the Time 2 survey
(77.4% response rate). Given the importance of careful responding
(D. Berry et al., 1992; Clark, Gironda, & Young, 2003), we
followed procedures to identify and remove careless responders
(e.g., Meade & Craig, 2012). In particular, we embedded instructed response items within each of the surveys (a sample item
is “Please mark the No. 3”), assessed consistency indices, and
assessed the length of time it took participants to complete the
survey. Participants who incorrectly responded to the instructed
response items, who responded with the same answer across the
surveys (such as marking all No. 3s), or who completed the survey
within a very brief time frame were not included in the analyses.
In addition, we integrated a critical incident technique in our
survey design. Researchers have argued that general survey studies
make it difficult, if not impossible, to effectively examine the nature
of specific events related to mistreatment and respondents’ reactions
to those events (see Hershcovis, 2011). Scholars, therefore, recommend the use of a critical incident technique to focus the study on
actions of specific perpetrators and respondents’ idiosyncratic reactions (Flanagan, 1954; Hershcovis, 2011) with the focus on a specific
event (Korsgaard, Brodt, & Whitener, 2002). This technique has been
demonstrated to be valid and effective in assessing individuals’ perceptions of and reactions to mistreatment by and conflict with supervisors (e.g., Bobocel, 2013; Korsgaard et al., 2002; Wang et al.,
2013). Although research has shown that recalling events may dilute
the effects associated with self-reported emotions (Thomas & Diener,
1990), research has also demonstrated that asking participants to recall
contextual details (e.g., details associated with witnessed abusive
supervision of a coworker) aids in the accuracy and vividness of
retrospection (Lang, Kozak, Miller, Levin, & McLean, 1980; Robinson & Clore, 2001). Accordingly, at the beginning of the Time 1
survey and prior to completing the survey’s measures, participants
were asked to recall details regarding a time when a coworker had
been treated in an unfavorable, unfriendly, or contentious manner by
a supervisor. We asked participants to describe this situation for a
frame of reference; respondents also provided us with the initials of
the coworker and supervisor in that situation. The coworker and
supervisor initials were automatically populated into the Time 1 and
Time 2 survey items that focused on either the coworker or supervisor. As an example, when rating perceptions of abusive supervision
against the identified coworker, we asked respondents to refer to the
situation and individuals identified in the situation they provided (and
the initials of the identified coworker and supervisor were inserted as
a reference in the measure instructions). In addition, the critical
incident described by the respondent at the beginning of the Time 1
survey was provided to participants at the beginning of the Time 2
survey for reference. Given the purpose of our study, we evaluated the
critical incidents to verify that participants witnessed abuse. Seven
participants who indicated they had not directly witnessed abusive
supervision of a coworker were removed from the data and not
included in the analyses.
At Time 1, participants completed measures of abusive supervision
of the coworker, exclusion beliefs about the coworker, experienced
emotions, moral identity, control variables, and demographics. At
1045
Time 2, participants rated their behaviors (supervisor-directed deviance, coworker support, and coworker exclusion). Podsakoff, MacKenzie, and Podsakoff (2012) suggested that researchers use time
separation between predictor and criterion variables to reduce consistency motifs and demand characteristics associated in common
method variance (CMV) bias. Accordingly, the surveys were separated in time by 1 month, as research has demonstrated a 1-month
time lag significantly reduces CMV-related inflation (Ostroff, Kinicki, & Clark, 2002). Further, a time lag of 1 month is consistent with
published work examining the influence of perceptions on behavior
(e.g., Jensen, Patel, & Raver, 2014; Matthews, Wayne, & Ford, 2014;
Rodell, 2013).
Complete data for all variables in our theoretical model were
available from 221 matched Time 1–Time 2 observations from
eligible participants. Analysis of variance tests demonstrated that
no significant differences existed between demographics of the
individuals who responded at Time 2 and nonresponders from
Time 1. On average, participants were 41.5 years old (SD ⫽ 13.0
years) and had worked for their organization for 8.3 years (SD ⫽
9.0 years); 57% were female, 81.4% were White, and 54.8%
worked in nonsupervisory positions.
Measures
Abusive supervision of the targeted coworker. We measured abusive supervision of the coworker using Tepper’s (2000)
15-item abusive supervision measure. Third parties were asked to
rate how often they witnessed or were aware of the supervisor
(identified in the situation) engaging in each of the listed behaviors
against the coworker identified in the situation (e.g., “Ridiculed the
coworker,” “Put down the coworker in front of others”) on a
5-point scale (1 ⫽ Never, 5 ⫽ Always).
Exclusion beliefs about the targeted coworker. We followed measure development procedures (Hinkin, 1998) to create a
measure of exclusion beliefs.1 First, items were developed based
on Opotow’s (1990, 1995) work and definition of exclusion beliefs
(i.e., “It is OK to treat this coworker unfairly,” and “This coworker
1
EFA was conducted to explore the factor structure of three items
created for the exclusion beliefs measure from an Amazon Mechanical
Turk sample (N ⫽ 126; average age ⫽ 29.7 years [SD ⫽ 7.84] and
organizational tenure ⫽ 4.1 years [SD ⫽ 3.41]; 51.6% were female, 84.9%
were White, and 69.8% were in nonsupervisory positions). We dropped
one item, which loaded separately and had an unacceptably low loading
(.03). The remaining two items yielded one factor (␭ ⬎ 1.0), accounting for
54.6% of the variance. Each item loading exceeded .80. The measure’s alpha
was .77. Data to examine the discriminant validity of the measure were
collected via Amazon’s Mechanical Turk (N ⫽ 250; average age ⫽ 29.4 years
[SD ⫽ 8.72] and organizational tenure ⫽ 3.58 years [SD ⫽ 3.74]; 68% were
female, 82% were White, and 77.2% were in nonsupervisory positions). We
examined the distinctiveness of our measure to measures of trust in the
coworker (Mayer & Gavin, 2006; ␣ ⫽ .81), similarity to the coworker (Liden
et al., 1993; ␣ ⫽ .94), and dislike for the coworker (Wayne & Ferris, 1990;
␣ ⫽ .91). The correlations provide preliminary evidence of discriminant
validity: exclusion beliefs with trust in the coworker r ⫽ –.19, p ⬍ .01; with
similarity to the coworker r ⫽ –.24, p ⬍ .01; with dislike for the coworker r ⫽
.59, p ⬍ .01. CFA results also suggest the measures are distinct: the four-factor
measurement model fit the data reasonably well—␹2(100) ⫽ 372.46, p ⬍
.001; CFI ⫽ .91; RMSEA ⫽ .10; SRMR ⫽ .07—and was a better fit than all
nested alternatives.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1046
MITCHELL, VOGEL, AND FOLGER
deserves to be treated poorly”). Second, data from a separate
sample of working adults were used to conduct an exploratory
factor analysis (EFA), which revealed two items loaded on one
factor. Third, data from a separate sample of working adults
provided preliminary evidence of discriminant validity of the
measure. Given this evidence, we assessed exclusion beliefs with
our developed measure and asked third parties to rate the two items
on a 5-point scale (1 ⫽ Strongly disagree, 5 ⫽ Strongly agree).
Emotions. Anger and contentment were assessed with threeitem measures by Frederickson, Tugade, Waugh, and Larkin
(2003). Third parties rated the extent to which they felt each
emotion (for anger: angry, irritated, and annoyed; for contentment: content, serene, and peaceful) as a consequence of the
supervisor’s treatment of the coworker identified in the situation
on a 5-point scale (1 ⫽ Not at all, 5 ⫽ Extremely).
Moral identity. Moral identity was assessed with Aquino and
Reed’s (2002) five-item measure. Third parties read a series of
moral characteristics that could describe a person (e.g., caring,
kind, friendly) and were asked to rate their agreement with a series
of items about their internalization of these characteristics (e.g.,
“Being someone who has these characteristics is an important part
of who I am,” “It would make me feel good to be a person who has
these characteristics”) on a 5-point scale (1 ⫽ Strongly disagree,
5 ⫽ Strongly agree).
Supervisor-directed deviance. Supervisor-directed deviance
was assessed with Mitchell and Ambrose’s (2007) 10-item measure. Third parties indicated how often they engaged in the listed
behaviors toward the supervisor identified in the situation (e.g.,
“Acted rudely toward the supervisor,” “Gossiped about the supervisor”) on a 5-point scale (1 ⫽ Never, 5 ⫽ Always).
Coworker support. Coworker support was assessed with
Westring and Ryan’s (2010) nine-item measure. Third parties rated
their agreement about whether they engaged in the listed behaviors
toward the coworker in the situation (e.g., “Provided the coworker
emotional support,” “Provided the coworker with help to handle
the situation with the supervisor”) on a 5-point scale (1 ⫽ Strongly
disagree, 5 ⫽ Strongly agree).
Coworker exclusion. Coworker exclusion was assessed with
Hitlan and Noel’s (2009) seven-item measure. Third parties indicated their agreement about whether they engaged in the listed
behaviors toward the coworker in the situation (e.g., “Gave the
coworker the ‘silent treatment,’” “Shut out the coworker from your
conversations”) on a 5-point scale (1 ⫽ Strongly disagree, 5 ⫽
Strongly agree).
Control variables. Because research on abusive supervision
suggests that third parties are more likely to agree with abusive
supervisors to whom they are similar (Tepper et al., 2011), we
controlled for third parties’ perceived similarity to the supervisor
with Liden, Wayne, and Stilwell’s (1993) six-item measure (e.g.,
“This supervisor and I are similar in terms of our outlook, perspective, and values,” “I think that this supervisor and I will see
things in the same way”). We also controlled for the third parties’
gender, the target of abuse’s gender, and the third parties’ work
tenure with the target of abuse, as research has shown that gender
(Malamut & Offermann, 2001; Miner-Rubino & Cortina, 2007)
and the length of time third parties know the victim (Eaton, 2013;
Priesemuth, 2013) may influence third parties’ reactions.
Results
Measurement Model Results
We conducted confirmatory factor analysis (CFA) to assess the
distinctiveness of the variables in our hypothesized model. Item
parcels for several of the latent constructs (i.e., abusive supervision
of the coworker, perceived similarity to the supervisor, supervisordirected deviance, coworker support, coworker exclusion) were
used, given the number of items compared with the study’s sample
size (Little, Cunningham, Shahar, & Widaman, 2002). For each of
these constructs, we created three parcels by successively assigning highest and lowest loading items across parcels (e.g., Major,
Turner, & Fletcher, 2006). Further, because there were two items
measuring exclusion beliefs, these items were averaged and specified to load onto their latent construct at the square root of alpha,
and its error variance was fixed to (1 ⫺ ␣)ⴱ(variance), following
recommendations by Kline (2005). Results indicated the hypothesized measurement model was a good fit to the data:—␹2(289) ⫽
457.82, p ⬍ .001; comparative fit index (CFI) ⫽ .96; root-meansquare error of approximation (RMSEA) ⫽ .05; standardized
root-mean-square residual (SRMR) ⫽ .05—and was a better fit
compared with nested alternative models.2
Hypothesis Tests Results
Table 1 shows the means, standard deviations, and correlations
among the study variables. We tested our hypotheses using Edwards and Lambert’s (2007) path analytic procedures for moderated indirect effects models (see also MacKinnon, Lockwood,
Hoffman, West, & Sheets, 2002). We tested all effects with biascorrected confidence intervals from 1,000 bootstrapped estimates
(Shrout & Bolger, 2002). Additionally, in all equations predicting
the dependent variables, we controlled for the direct effects of the
independent variable, abusive supervision of the coworker. Tables
2 and 3 summarize the path analytic results of the model. Predictor
variables were mean-centered, and interactions were plotted at
high and low levels (⫾ 1 SD) of each moderator (Aiken & West,
1991).
Hypothesis 1(a) predicted that third parties’ exclusion beliefs
about the coworker would moderate the relationship between
abusive supervision of the coworker and third parties’ anger
such that the effects would be more strongly positive when
exclusion beliefs were lower than higher. The Abusive
Supervision⫻Exclusion Beliefs interaction term was significant
and negative on anger (b ⫽ ⫺0.36, p ⬍ .05), and the relationship between abusive supervision and anger was more strongly
positive when exclusion beliefs were lower (b ⫽ .76, p ⬍ .001)
rather than higher (b ⫽ 0.30, p ⬍ .05; see Figure 2a). Therefore,
Hypothesis 1(a) was supported.
Hypothesis 1(b) predicted that third parties’ exclusion beliefs about the
coworker would moderate the relationship between abusive supervision
of the coworker and third parties’ contentment, such that the effects
would be more strongly positive when exclusion beliefs were higher than
lower. The Abusive Supervision ⫻ Exclusion Beliefs interaction term
2
Details of these comparison tests can be obtained by contacting the first
author.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
1047
Table 1
Descriptive Statistics and Correlationsa
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Variables
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Abusive supervision of the coworker
Exclusion beliefs
Anger
Contentment
Moral identity
Supervisor-directed deviance
Coworker support
Coworker exclusion
Supervisor similarity
Third party genderb
Coworker genderb
Third party/coworker tenure
M
SD
1
2
3
4
5
6
7
8
9
10
11
2.40
1.34
3.18
1.20
4.53
1.35
3.62
1.28
2.57
1.57
1.61
4.48
0.74
0.63
1.12
0.54
0.54
0.39
0.74
0.55
0.92
0.49
0.49
5.06
(.91)
⫺.10
.51
⫺.02
.04
.27
.18
⫺.02
⫺.49
.04
.03
.09
(.75)
⫺.17
.15
⫺.24
.04
⫺.17
.28
.22
⫺.20
⫺.08
⫺.01
(.91)
⫺.26
.25
.18
.21
⫺.15
⫺.47
.15
.06
.12
(.84)
⫺.23
.03
.09
.26
.19
⫺.16
⫺.08
.01
(.84)
⫺.25
.13
⫺.32
⫺.05
.29
.18
.13
(.81)
.13
.13
⫺.24
⫺.08
⫺.06
⫺.06
(.94)
⫺.26
⫺.28
.02
.03
.09
(.95)
.20
⫺.19
⫺.09
⫺.03
(.94)
.01
⫺.06
⫺.06
—
.41
⫺.09
—
⫺.25
Note. Coefficient alpha values for the variables are shown in italics along the diagonal. Correlations greater than |.13| are significant at p ⬍ .05 and those
greater than |.17| are significant at p ⬍ .01, two-tailed.
a
N ⫽ 221. b 1 ⫽ male, 2 ⫽ female.
was significant and positive on contentment (b ⫽ 0.23, p ⬍ .05), and the
relationship between abusive supervision and contentment was more
strongly positive when exclusion beliefs were higher (b ⫽ 0.24, p ⬍ .05)
rather than lower (b ⫽ ⫺0.04, ns; see Figure 2b). These results support
Hypothesis 1(b).
Hypothesis 2(a) predicted that third parties’ moral identity would
moderate the relationship between third parties’ anger and supervisordirected deviance, such that the effects would be more strongly
positive when moral identity was lower rather than higher. The
Anger ⫻ Moral Identity interaction term was significant and negative
on supervisor-directed deviance (b ⫽ ⫺0.11, p ⬍ .05), and the
relationship between anger and supervisor-directed deviance was
more strongly positive when moral identity was lower (b ⫽ 0.11,
p ⬍ .05) rather than higher (b ⫽ ⫺0.02, ns; see Figure 3a). These
results support Hypothesis 2(a).
Hypothesis 2(b) predicted that third parties’ moral identity
would moderate the relationship between third parties’ anger and
coworker support, such that the effects would be more strongly
positive when moral identity was higher than lower. The Anger ⫻
Moral Identity interaction term was significant and positive on
coworker support (b ⫽ 0.22, p ⬍ .05), and the relationship be-
tween anger and coworker support was more strongly positive
when moral identity was higher (b ⫽ 0.15, p ⬍ .05) rather than
lower (b ⫽ ⫺0.09, ns; see Figure 3b). Thus, Hypothesis 2(b) was
supported.
Hypothesis 2(c) predicted that third parties’ moral identity would
moderate the relationship between third parties’ contentment and
coworker exclusion, such that the effects would be more strongly
positive when moral identity was lower rather than higher. The
Contentment ⫻ Moral Identity interaction term was significant and
negative on coworker exclusion (b ⫽ ⫺0.29, p ⬍ .01), and the
relationship between contentment and coworker exclusion was more
strongly positive when moral identity was lower (b ⫽ 0.27, p ⬍ .01)
rather than higher (b ⫽ – 0.04, ns; see Figure 3c). These results
support Hypothesis 2(c).
We used the information from Tables 2 and 3 to compute the
indirect effects of abusive supervision of the coworker on the
dependent variables (via the mediators, anger and contentment) at
higher and lower values of the moderators (exclusion beliefs and
moral identity). Hypothesis 3(a) predicted that the indirect effects
of abusive supervision of the coworker on supervisor-directed deviance (via anger) would be more strongly positive when exclusion
Table 2
Path Analytic Results on Anger and Contentmenta
Anger
Contentment
Step 1
Step 2
Step 1
Step 2
Predictors
b
SE
b
SE
b
SE
Supervisor similarity
Third party genderb
Coworker genderb
Third party/coworker tenure
Abusive supervision of the coworker
Exclusion beliefs
Abusive supervision of the Coworker ⫻ Exclusion Beliefs
Equation R2
⌬R2
⫺0.35ⴱⴱ
0.34ⴱⴱ
⫺0.04
0.02
0.54ⴱⴱ
⫺0.07
0.08
0.13
0.14
0.01
0.11
0.12
⫺0.34ⴱⴱ
0.34ⴱⴱ
⫺0.03
0.02
0.50ⴱⴱ
⫺0.15
⫺0.36ⴱ
0.37ⴱⴱ
0.02ⴱ
0.08
0.13
0.14
0.01
0.10
0.10
0.15
0.13ⴱⴱ
⫺0.17ⴱⴱ
0.01
0.00
0.08
0.06
0.04
0.07
0.07
0.01
0.05
0.07
Note. Beta values are unstandardized regression coefficients.
a
N ⫽ 221. b 1 ⫽ male, 2 ⫽ female.
†
p ⬍ .10. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01.
0.35ⴱⴱ
0.08ⴱⴱ
b
0.13ⴱⴱ
⫺0.17ⴱ
0.00
⫺0.00
0.10†
0.12
0.23ⴱ
0.10ⴱⴱ
0.02ⴱ
SE
0.04
0.07
0.07
0.01
0.05
0.08
0.11
MITCHELL, VOGEL, AND FOLGER
1048
Table 3
Path Analytic Results on the Dependent Variablesa
Supervisor-directed deviance
Step 1
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Predictors
Supervisor similarity
Third party genderb
Coworker genderb
Third party/coworker tenure
Abusive supervision of the coworker
Anger
Contentment
Moral identity
Anger ⫻ Moral Identity
Contentment ⫻ Moral Identity
Equation R2
⌬R2
b
Coworker support
Step 2
SE
b
Step 1
SE
ⴱ
b
SE
ⴱⴱ
⫺0.05
⫺0.01
⫺0.03
⫺0.01
0.09ⴱ
0.04†
0.03
0.06
0.06
0.01
0.04
0.03
⫺0.06
⫺0.00
⫺0.04
⫺0.01
0.09ⴱ
0.05†
0.03
0.06
0.06
0.01
0.04
0.03
⫺0.18
⫺0.03
⫺0.01
⫺0.00
0.04
0.04
0.07
0.11
0.12
0.01
0.08
0.05
⫺0.19ⴱⴱ
0.07
⫺0.23ⴱⴱ
⫺0.11ⴱ
0.08
0.05
0.15
0.13
0.17ⴱⴱ
0.20ⴱⴱ
0.03ⴱ
0.10ⴱⴱ
Coworker exclusion
Step 2
b
Step 1
SE
ⴱ
0.07
0.11
0.12
0.01
0.08
0.05
0.22†
0.22ⴱ
0.12
0.09
⫺0.17
⫺0.05
⫺0.01
⫺0.00
0.04
0.03
0.12ⴱⴱ
0.02ⴱ
b
Step 2
SE
b
SE
ⴱⴱ
0.12
⫺0.12
0.02
0.00
0.07
0.04
0.08
0.09
0.01
0.05
0.11
⫺0.13
0.06
0.00
0.05
0.04
0.08
0.08
0.01
0.05
0.20ⴱ
⫺0.25ⴱⴱ
0.09
0.07
0.12†
⫺0.20ⴱⴱ
0.07
0.06
⫺0.29ⴱⴱ
0.22ⴱⴱ
0.04ⴱⴱ
0.10
0.18ⴱⴱ
ⴱ
Note. Beta values are unstandardized regression coefficients.
a
N ⫽ 221. b 1 ⫽ male, 2 ⫽ female.
†
p ⬍ .10. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01.
beliefs were lower and moral identity was lower. The results support
Hypothesis 3(a); the indirect effect of abusive supervision on
supervisor-directed deviance was more strongly positive when exclusion beliefs were lower and moral identity was lower (␳ ⫽ .08,
p ⬍ .05) rather than when exclusion beliefs were higher and moral
identity was higher (␳ ⫽ –.01, ns), when exclusion beliefs were higher
and moral identity was lower (␳ ⫽ .03, ns), or when exclusion beliefs
were lower and moral identity was higher (␳ ⫽ –.01, ns).
Hypothesis 3(b) predicted that the indirect effects of abusive
supervision of the coworker on coworker support (via anger)
would be more strongly positive when exclusion beliefs were
lower and moral identity was higher. The results support Hypothesis 3(b); the indirect effect of abusive supervision on coworker
support was more strongly positive when exclusion beliefs were
lower and moral identity was higher (␳ ⫽ .11, p ⬍ .05) rather than
when exclusion beliefs were higher and moral identity was higher
(␳ ⫽ .04, ns), when exclusion beliefs were higher and moral
identity was lower (␳ ⫽ –.03, ns), or when exclusion beliefs were
lower and moral identity was lower (␳ ⫽ –.07, ns).
Hypothesis 3(c) predicted that the indirect effects of abusive
supervision of the coworker-on-coworker exclusion (via contentment) would be more strongly positive when exclusion
beliefs were higher and moral identity was lower. The results
support Hypothesis 3(c); the indirect effect of abusive supervision on coworker exclusion was more strongly positive when
exclusion beliefs were higher and moral identity was lower
(␳ ⫽ .07, p ⬍ .05) rather than when exclusion beliefs were
higher and moral identity was higher (␳ ⫽ –.01, ns), when
Figure 2. Abusive supervision of the Coworker ⫻ Exclusion Beliefs interaction on emotion. a: Abusive
Supervision ⫻ Exclusion Beliefs on Anger. b: Abusive Supervision ⫻ Exclusion Beliefs on Contentment.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
1049
Figure 3. Moral identity interactions. a: Anger ⫻ Moral Identity on supervisor-directed deviance. b: Anger ⫻
Moral Identity on coworker support. c: Contentment ⫻ Moral Identity on coworker exclusion.
exclusion beliefs were lower and moral identity was higher
(␳ ⫽ .00, ns), or when exclusion beliefs were lower and moral
identity was lower (␳ ⫽ –.01, ns).
Supplemental Model Tests Results
Some scholars have speculated that moral identity might mitigate third parties’ emotional responses to another person’s mistreatment (e.g., O’Reilly & Aquino, 2011). Although our placement of moral identity at the second stage of our model is
consistent with moral self theory (Blasi, 1984, 2004) and research
(see Jennings et al., in press), we examined alternative models to
test the effects of moral identity as a moderator at the first stage of
our theoretical model. We also tested the possibility that exclusion
beliefs moderated the effects of felt emotions on the behavioral
reactions. Following protocol recommended by Edwards and Lambert (2007), we compared the results of our theoretical model
(Figure 1) to results of alternative models. In the first alternative
model, we specified moral identity as a moderator at the first stage
of the hypothesized model, between abusive supervision of the
coworker and the emotions. The results showed the Abusive Supervision of the Coworker ⫻ Moral Identity interaction term was
not significant on anger or contentment. In the second alternative
model, we specified exclusion beliefs as a moderator at the second
stage of the hypothesized model, between the emotions and the
dependent variables. The results showed the Anger ⫻ Exclusion
Beliefs interaction term was not significant on supervisor-directed
deviance or coworker support. The Contentment ⫻ Exclusion
Beliefs interaction term was significant on coworker exclusion,
such that the positive relationship was stronger when exclusion
beliefs were higher than lower. Thus, the indirect effect of abusive
supervision of the coworker-on-coworker exclusion (via contentment) may be even stronger than what we reported earlier. Overall,
MITCHELL, VOGEL, AND FOLGER
1050
the supplemental analyses lend further support for our theorized
model.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Discussion
The goal of our research was to examine third parties’ reactions
to abusive supervision of a coworker. Extant work has demonstrated that third parties may react quite differently to witnessing a
coworker be abused (Harris et al., 2013; Priesemuth, 2013). Our
work sought to identify factors that motivate different reactions.
The results of our study show that witnessed abusive supervision
indirectly influences third parties’ behavioral reactions at higher
and lower levels of their exclusion beliefs (whether third parties
believe the target of abuse was deserving or not of mistreatment)
and their moral identity. Exclusion beliefs about the target of abuse
evoked different emotions from third parties. When exclusion
beliefs were low, witnessing abusive supervision evoked anger.
When exclusion beliefs were high, witnessing abusive supervision
elicited contentment. The emotions experienced by our respondents motivated specific action tendencies. Third parties’ behavioral reactions were moderated by their moral identity,
however, such that a strong moral identity assisted these third
parties in resisting their desires to engage in harmful behavior
and, instead, enhanced the likelihood that they engaged in
ethically oriented acts. The results show that third parties with
a stronger rather than weaker moral identity were less likely to
direct harmful behavior toward the supervisor (i.e., supervisordirected deviance) or targeted coworker (i.e., coworker exclusion) and were more likely to help the target of abuse (i.e.,
coworker support).
Theoretical Implications
Our work has a number of implications for theory and research.
For instance, our work extends the abusive supervision literature to
consider the impact of abusive supervision on third parties (cf.
Folger, 2001). Our study suggests that abusive supervision elicits
different reactions from third parties—those who become angry try
to help the victim and harm the supervisor, and those who become
contented perpetrate abusive behavior against targeted coworkers
by excluding them. Consequently, this work demonstrates the
importance of considering the larger social environment in which
abusive supervision occurs and suggests third parties’ beliefs about
the target of abuse matter. These beliefs influence third parties’
emotionally charged reactions from witnessing abusive supervision, which then motivates harmful behavior. Overall, the study
suggests supervisor abuse initiates a spiral of destructive and
costly behavior (cf. Andersson & Pearson, 1999).
This study also extends deontic principles by pointing out that a
critical boundary condition to third party reactions is third parties’
beliefs about the target of mistreatment. Our work directly addresses the underlying assumption implied in prior work (see
Skarlicki & Kulik, 2005)—that all third parties become angered by
another’s abuse because they find mistreating others unfair—and
demonstrates that third parties’ judgments about the abused coworker exert a strong influence on the nature of their reactions.
Consistent with deontic reasoning, we show that third parties are
guided by what they believe is “right.” However, our work extends
the traditional deontic view by explaining why not all third parties
believe what is “right” is to restore justice on behalf of the
victimized coworker by harming the supervisor and providing the
coworker with support. In contrast, some third parties believe that
abusing the targeted coworker is “right” and, therefore, see continued abuse against that coworker (viz., exclusion) as appropriate.
Not all third parties become angered by the mistreatment; instead,
some feel contented about the abuse. This expanded view of
deontic justice offers researchers the opportunity for a more comprehensive understanding of third parties’ reactions to mistreatment.
Our research also adds to the growing literature on victimization
and victims as provocative targets. Prior work has shown that
supervisors target abuse against employees who are aggravating
and annoying (Tepper et al., 2011). Coworkers seemingly do the
same. There may be many reasons for coworkers to hold strong
exclusion beliefs about another employee. Theory (Deutsch, 1990;
Feather, 2006) suggests that exclusion beliefs originate from a
social comparison, wherein the target is considered inferior to the
third party. It represents a “just deserts” ideal, such that mistreatment is perceived to be deserved because of some characteristic or
behavior of the target. Our results suggest that when third parties
hold high exclusion beliefs about particular coworkers, they are
less bothered by negative events that befall the targets of abuse, but
if third parties do not hold high exclusion beliefs, they are willing
to “go to bat” for the targets of abuse and attempt to restore justice
on the victims’ behalf.
Furthermore, the results of our research add to the literature on
emotions and their influence on employees’ work behavior. Emotions have specific action tendencies that can impair employees’
abilities to engage in functional work behavior (Weiss & Cropanzano, 1996). Consistent with these ideas, the results show that
anger influenced supervisor-directed deviance and contentment
influenced coworker exclusion. Coworker support only emerged
when third parties held a strong moral identity; the direct effect of
anger on coworker support was not significant (b ⫽ 0.03, ns).
Further, abusive supervision of a coworker did not directly influence third parties’ contentment. It was only when third parties
believed the targeted coworker was deserving of mistreatment (i.e.,
high exclusion beliefs) that third parties experienced contentment
and were motivated to exclude the coworker. Thus, certain aspects
of the situation and the individual provide important boundary
conditions regarding the effects of emotion on affect-driven behaviors.
Limitations and Future Research Directions
As with all studies, ours is not without limitations. For instance,
researchers have discussed potential issues that can arise from
same-source data, with specific concerns about the nature of common method variance (CMV) bias. We argue, however, that given
our research goal, a same-source data design was necessary. First,
it was important to use the critical incident technique to focus
respondents on the research purpose, namely, their reactions to the
abusive supervision of a specific coworker. This design allowed us
to examine the influence of exclusion beliefs, which was theorized
as the focal variable that influences third parties’ differential
reactions to abusive supervision. Asking focal respondents about
their perceptions of the event and the targeted coworker, their
emotions regarding the event, their own level of moral identity,
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
and their engaged behaviors resulting from the event could only be
effectively assessed from the third party respondent. Supporting
this idea, meta-analysis comparing self-reports to other-reports of
sensitive data (such as abusive supervision, exclusion beliefs,
engaged deviance, and engaged exclusion) demonstrate that selfreported data are more accurate than data collected from otherreports (C. Berry, Carpenter, & Barratt, 2012). Further, Lee (1993)
argued that self-reports are more accurate because coworker and
supervisors are often unaware of employees’ destructive work
behavior (e.g., supervisor-directed deviance, coworker exclusion)
until the employee has been caught and reprimanded. Second,
asking the focal third party respondent to involve a coworker (who
may be the target of abuse) or supervisor (who may be the
abusive supervisor) to also rate his or her perceptions of the
event or of the focal third party respondent’s attitudes or
behaviors about the event (a particularly sensitive work situation) might unduly put the focal third party respondent at risk at
work. Doing so might also have been problematic vis-à-vis
institutional review board protocols and ethical guidelines of
the American Psychological Association (2010).
Given the need to collect data from the same source, we took
several proactive steps to address the concerns related to CMV.
First, we adopted design elements recommended by researchers
(Podsakoff et al., 2012) that have been shown to reduce CMV.
Specifically, we collected our criterion variables at a later time
than the predictor and mediating variables. Researchers have demonstrated that CMV cannot present a bias to data when moderators
are considered (Siemsen, Roth, & Oliveira, 2010); this research
has shown that CMV would weaken, not strengthen, interactions.
Thus, scholars (Podsakoff et al., 2012; Siemsen et al., 2010) have
concluded that if an interaction effect is found, it should be taken
as strong evidence that an interaction exists and that CMV did not
present a bias. Notably, our results show significant interaction
effects and that the indirect effects varied over levels of the two
moderator variables. Last, we tested for CMV statistically. Following Williams, Hartman, and Cavazotte’s (2010) procedures, we
used a latent marker variable to examine the extent to which CMV
existed in our data.3 The results show CMV was not present and
did not bias the findings.
A second area of concern relates to causality. As with all field
and survey studies, causality cannot be conclusively determined.
However, we took steps to enhance the nature of causality in our
study. First, we ordered the variables in the data collection efforts
in such a way to ensure the causal chain was accurately represented. The ordering of variables in the model was based on strong
theory (e.g., Folger, 2012; Weiss & Cropanzano, 1996), as well as
experimental research (e.g., Blader et al., 2013; Turillo et al.,
2002) about the temporal experience of emotion and behavior.
Second, we assessed our dependent variables at a separate time
from our predictor variables, which scholars contend enhances
causal inferences (Podsakoff et al., 2012). Nonetheless, we concede that examining the influence of exclusion beliefs in association with third parties’ reactions to witnessed abuse is challenging
and, therefore, urge researchers to consider untangling the dynamics within an experimental setting.
Despite these limitations, our work can assist researchers in
extending the literature further. In particular, because our work
demonstrates that exclusion beliefs play a critical role in understanding how third parties react to abusive supervision of a co-
1051
worker, researchers might examine how exclusion beliefs form
within the work environment. Scope of justice theory (Opotow,
1990, 1995) suggests that there are many reasons why individuals
may exclude others (e.g., race, gender, personality, behavior). Our
work did not focus on a specific reason for exclusion, but it is
possible for different reasons of exclusion to differentially influence third parties’ reactions. We are hopeful researchers continue
this line of work and explore the development of and further
consequences of exclusion beliefs.
Additionally, our expanded view of deontic justice principles
provides a foundation for research, such as examining other emotions that may emerge from witnessing another person be mistreated. The emotions studied in our article may be viewed as two
extreme forms of emotional reactions: anger and contentment. The
means for anger and contentment suggest that anger is a more
prominent emotional experience, which is a good thing. Third
parties, thus, are more distressed by—rather than cynical about—
witnessing a coworker’s abuse. Still, third parties may also experience other types of emotions, such as being sad, scared, or
intimidated about witnessing abusive supervision. It may also be
possible for third parties to experience sympathy, compassion, or
empathy for the abused coworker. Our work points to the need to
study other emotions that might be evoked in third parties when
they witness abusive supervision.
It would also be interesting for researchers to consider other
third party behavioral reactions. For example, Mitchell et al.
(2012) theorized that angered third parties might try to avenge the
victim by approaching the supervisor to ask that the abuse stop or
by reporting the supervisor to authorities; contented third parties
may try to encourage others to exclude and abuse the victim or
engage in more aggressive behaviors against the target of abuse
(e.g., aggression, undermining). More work is needed to fully
explore third parties’ reactions.
Additionally, we see value in researchers exploring the effects
of contentment from another’s mistreatment. Research on
schadenfreude—a sense of joy or pleasure experienced because of
another’s mistreatment or misfortune (Feather, 2006)— has received limited attention within organizational research. The results
of our work suggest that schadenfreude motivates destructive
behavior against the source of contentment. Feather (2006) argued
that work environment pressures may make the experience of
schadenfreude more likely to occur. Given this, there is much to
learn about the organizational and interpersonal work factors that
contribute to schadenfreude or related emotions, such as envy or
jealousy.
Although our research shows that abusive supervision influences third parties’ reactions, it is possible for the third party to
also be a target of abusive supervision and that the supervisor may
abuse the entire work unit. If this were the case, the abuse may
3
A latent marker variable was used to control for and partial out the
variance that was due to a common method (Williams et al., 2010). The
first procedural step is to compare a baseline constrained model to one in
which the substantive items were allowed to load on the unconstrained
method factor. If the chi-square difference test between these two models
is significant, then CMV is said to be present, and further tests are needed
to determine the extent to which CMV biases the results. The chi-square
difference test involved in the first step of testing was not significant,
suggesting CMV was not present or biased the parameter estimates of our
hypothesized model.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1052
MITCHELL, VOGEL, AND FOLGER
suggest to third parties that the abuse of a coworker is not targeted.
Given this, it is possible for third parties who are also abused to
become even more angry about abuse of a peer—someone, like
themselves, who is undeserving of abuse (i.e., low exclusion
beliefs). Yet, third parties who are also abused might be less
angered because they see the abuse as simply a climate issue
(Ogunfowora, 2013), which may temper affective and behavioral
reactions. Researchers should explore these ideas further.
Finally, we asked participants in our study to refer to a specific
coworker (who was abused) and a specific supervisor (the abuser).
This design was necessary to examine third parties’ exclusion
beliefs about the target of abuse. However, research may consider
other designs to explore third party reactions. For example, examining third party reactions more generally may reduce the risk
placed on the third party to recruit other sources to participate in
the study. Researchers might also consider exploring third party
reactions within experimental settings. We understand, however,
that testing a full range of third party reactions in a lab may be
impossible, as lab settings may be insufficient to elicit realistic
(taboo) reactions (King, Hebl, Morgan, & Ahmad, 2013), such as
experienced contentment from witnessing abusive supervision.
Further, some aspects of organizations may be difficult to capture
or might be overlooked within a lab setting (Heath & Sitkin, 2001).
Thus, researchers might explore the effects over a course of studies
that may infer the process by which behavioral reactions occur (cf.
Spencer, Zanna, & Fong, 2005).
Implications for Practice
Our findings underscore the importance for organizations to
control the incidence of abusive supervision, as our results show
abusive supervision impacts the larger workgroup and, specifically, third party witnesses. Prior estimates of abusive supervision
suggest that abusing subordinates produces high costs to organizations through its impact on victims and victims’ behavioral
reactions (e.g., lawsuits, health costs; Tepper, Duffy, Henle, &
Lambert, 2006). These estimates, however, did not consider the
consequences imposed on or by third parties. Taking third parties’
reactions into account suggests that the costs of abusive supervision to organizations are even greater, as our study demonstrates
third parties can perpetuate destructive behavior within organizations from witnessing abuse (i.e., supervisor-directed deviance,
coworker exclusion). Although on the face of things, third parties
directing deviance at an abusive supervisor might seem to be a
positive reaction, these types of behaviors (and coworker exclusion) have been found to be quite costly to organizations (e.g.,
Detert et al., 2007; Dunlop & Lee, 2004). They wreak havoc on the
interpersonal dynamics needed to foster productive work behavior
among employees and within business units (cf. Sutton, 2010).
A positive takeaway, then, is that some third parties engaged in
more constructive reactions by way of coworker support. These
types of reactions can clearly help organizational decision makers
uncover incidents of abusive supervision. Moreover, they may
assist with fostering a compassionate and supportive business
environment. Yet, our results suggest that harnessing third parties’
empathic reactions to aid the abused coworker may be challenging.
That is, the results of our study demonstrate that third parties who
were angered by witnessing the abuse only engaged in supportive
behaviors when they also held a strong moral identity. Thus, our
study underscores the importance of organizational decision makers in fostering ethical behaviors that more easily enable employees’ moral identity. Organizations can do so by communicating the
importance of ethical behavior and emphasizing accountability to
these standards, as research has shown that increasing awareness to
moral standards strengthens moral identity (Aquino et al., 2011).
Conclusion
Our overall theoretical model and results suggest that third
parties’ reactions may contribute to the costs of abusive supervision. Although prosocial behaviors (i.e., coworker support) did
emerge, more destructive behaviors were also probable. Evoked
anger from witnessed abuse motivated supervisor-directed deviance, whereas evoked contentment from witnessed abuse motivated coworker exclusion. These effects were even stronger when
third parties held a weak moral identity. Because these behaviors
can be a source of a financial drain on organizations (e.g., Dunlop
& Lee, 2004), decision makers should focus efforts on reducing
the incidence of abusive supervision and enhancing relational
skills that can assist third parties in fostering effective and ethical
interpersonal dynamics for organizations.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting interactions. Thousand Oaks, CA: Sage.
American Psychological Association. (2010). Ethical principles of psychologists and code of conduct (including 2010 amendments). Retrieved
from http://www.apa.org/ethics/code/index.aspx
Andersson, L., & Pearson, C. (1999). Tit for tat? The spiraling effect of
incivility in the workplace. Academy of Management Review, 24, 452–
471.
Aquino, K., McFerran, B., & Laven, M. (2011). Moral identity and the
experience of moral elevation in response to acts of uncommon goodness. Journal of Personality and Social Psychology, 100, 703–718.
doi:10.1037/a0022540
Aquino, K., & Reed, A. (2002). The self-importance of moral identity.
Journal of Personality and Social Psychology, 83, 1423–1440. doi:
10.1037/0022-3514.83.6.1423
Aquino, K., Reed, A., Stewart, M., & Shapiro, D. (2005). Self-regulatory
identity theory and reactions toward fairness enhancing organizational
policies. In S. Gilliland, D. Steiner, D. Skarlicki, & K. van den Bos
(Eds.), What motivates fairness in organizations? (pp. 129 –148). Charlotte, NC: Information Age.
Berry, C. M., Carpenter, N. C., & Barratt, C. L. (2012). Do other-reports
of counterproductive work behavior provide an incremental contribution
over self-reports? A meta-analytic comparison. Journal of Applied Psychology, 97, 613– 636. doi:10.1037/a0026739
Berry, D., Wetter, M. W., Baer, R., Larsen, L., Clark, C., & Monroe, K.
(1992). MMPI–2 random responding indices: Validation using a selfreport methodology. Psychological Assessment, 4, 340 –345. doi:
10.1037/1040-3590.4.3.340
Blader, S. L., Wiesenfeld, B. M., Fortin, M., & Wheeler-Smith, S. L.
(2013). Fairness lies in the heart of the beholder: How the social
emotions of third parties influence reactions to injustice. Organizational
Behavior and Human Decision Processes, 121, 62– 80. doi:10.1016/j
.obhdp.2012.12.004
Blader, S. L., Wiesenfeld, B. M., Rothman, N. B., & Wheeler-Smith, S. L.
(2010). Social emotions and justice: How the emotional fabric of groups
determines justice enactment and reactions. In B. Mannix, M. Neale, &
E. Mullen (Eds.), Fairness and groups: Research on managing groups
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
and teams (Vol. 13, pp. 29 – 62). Bingley, United Kingdom: Emerald
Group. doi:10.1108/S1534-0856(2010)0000013005
Blasi, A. (1983). Moral cognition and moral action: A theoretical perspective.
Developmental Review, 3, 178 –210. doi:10.1016/0273-2297(83)90029-1
Blasi, A. (1984). Moral identity: Its role in moral functioning. In W.
Kurtines & J. Gewirtz (Eds.), Morality, moral behavior, and moral
development (pp. 128 –139). New York, NY: Wiley.
Blasi, A. (2004). Moral functioning: Moral understanding and personality.
In D. Lapsley & D. Narvaez (Eds.), Moral development, self, and
identity (pp. 335–347). Mahwah, NJ: Erlbaum.
Bobocel, D. R. (2013). Coping with unfair events constructively and
destructively: The effects of overall justice and self– other orientation.
Journal of Applied Psychology, 98, 720 –731. doi:10.1037/a0032857
Clark, M. E., Gironda, R. J., & Young, R. W. (2003). Detection of back
random responding: Effectiveness of MMPI–2 and personality assessment inventory validity indices. Psychological Assessment, 15, 223–234.
doi:10.1037/1040-3590.15.2.223
Cortina, L. M., Magley, V. J., Williams, J. H., & Langhout, R. D. (2001).
Incivility in the workplace: Incidence and impact. Journal of Occupational Health Psychology, 6, 64 – 80. doi:10.1037/1076-8998.6.1.64
Darley, J. M., & Pittman, T. S. (2003). The psychology of compensatory
and retributive justice. Personality and Social Psychology Review, 7,
324 –336. doi:10.1207/S15327957PSPR0704_05
Detert, J. R., Treviño, L. K., Burris, E. R., & Andiappan, M. (2007).
Managerial modes of influence and counterproductivity in organizations: A longitudinal business-unit-level investigation. Journal of Applied Psychology, 92, 993–1005. doi:10.1037/0021-9010.92.4.993
Deutsch, M. (1974). Awaking the sense of injustice. In M. Lerner & M.
Ross (Eds.), The quest for justice: Myth, reality, ideal (pp. 19 – 42).
Toronto, ON, Canada: Holt, Rinehart & Winston.
Deutsch, M. (1990). Psychological roots of moral exclusion. Journal of
Social Issues, 46, 21–25. doi:10.1111/j.1540-4560.1990.tb00269.x
Dunlop, P. D., & Lee, K. (2004). Workplace deviance, organizational
citizenship behavior, and business unit performance: The bad apples do
spoil the whole barrel. Journal of Organizational Behavior, 25, 67– 80.
doi:10.1002/job.243
Earnshaw, V., Pitpitan, E., & Chaudoir, S. (2011). Intended responses to
rape as functions of attitudes, attributions of fault, and emotions. Sex
Roles, 64, 382–393. doi:10.1007/s11199-010-9920-1
Eaton, J. (2013). The effects of third-party validation and minimization on
judgments of the transgressor and the third party. British Journal of
Social Psychology, 52, 273–289. doi:10.1111/j.2044-8309.2011.02066.x
Edwards, J., & Lambert, L. (2007). Methods for integrating moderation and
mediation: A general analytical framework using moderated path analysis.
Psychological Methods, 12, 1–22. doi:10.1037/1082-989X.12.1.1
Eisenberg, N., & Miller, P. A. (1987). The relation of empathy to prosocial
and related behaviors. Psychological Bulletin, 101, 91–119. doi:
10.1037/0033-2909.101.1.91
Erikson, E. H. (1964). Insight and responsibility. New York, NY: Norton.
Feather, N. (2006). Deservingness and emotions: Applying the structural
model of deservingness to the analysis of affective reactions to outcomes. European Review of Social Psychology, 17, 38 –73. doi:10.1080/
10463280600662321
Feather, N. (2008). Effects of observer’s own status on reactions to a high
achiever’s failure: Deservingness, resentment, schadenfreude, and sympathy. Australian Journal of Psychology, 60, 31– 43. doi:10.1080/
00049530701458068
Flanagan, J. C. (1954). The critical incident technique. Psychological
Bulletin, 51, 327–358. doi:10.1037/h0061470
Folger, R. (1998). Fairness as moral virtue. In M. Schminke (Ed.), Managerial ethics: Moral management of people and processes (pp. 13–34).
Mahwah, NJ: Erlbaum.
1053
Folger, R. (2001). Fairness as deonance. In S. Gilliland, D. Steiner, & D.
Skarlicki (Eds.), Theoretical and cultural perspectives on organizational
justice (pp. 3–33). Greenwich, CT: Information Age.
Folger, R. (2012). Deonance: Behavioral ethics and moral obligation. In D.
DeCremer & A. Tenbrunsel (Eds.), Behavioral business ethics: Ideas on
an emerging field (pp. 123–142). London, United Kingdom: Taylor &
Francis.
Frederickson, B. L., Tugade, M. M., Waugh, C. E., & Larkin, G. R. (2003).
What good are positive emotions in crisis? A prospective study of
resilience and emotions following terrorist attacks on the United States
on September 11th, 2001. Journal of Personality and Social Psychology,
84, 365–376. doi:10.1037/0022-3514.84.2.365
Frijda, N. H. (1988). The laws of emotion. American Psychologist, 43,
349 –358. doi:10.1037/0003-066X.43.5.349
Grant, A. M., & Mayer, D. M. (2009). Good soldiers and good actors:
Prosocial and impression management motives as interactive predictors
of affiliative citizenship behaviors. Journal of Applied Psychology, 94,
900 –912.
Greenbaum, R. L., Mawritz, M. B., & Eissa, G. (2012). Bottom-line
mentality as an antecedent of social undermining and the moderating
roles of core self-evaluations and conscientiousness. Journal of Applied
Psychology, 97, 343–359. doi:10.1037/a0025217
Hafer, C., & Bégue, L. (2005). Experimental research on just-world theory:
Problems, developments, and future challenges. Psychological Bulletin,
131, 128 –167. doi:10.1037/0033-2909.131.1.128
Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108, 814 – 834.
Haidt, J. (2003). The moral emotions. In R. J. Davidson, K. R. Scherer, &
H. H. Goldsmith (Eds.), Handbook of affective sciences (pp. 852– 870).
Oxford, United Kingdom: Oxford University Press.
Hardy, S. A., & Carlo, G. (2005). Identity as a source of moral motivation.
Human Development, 48, 232–256. doi:10.1159/000086859
Harris, K. J., Harvey, P., Harris, R. B., & Cast, M. (2013). An investigation
of abusive supervision, vicarious abusive supervision, and their joint
impacts. Journal of Social Psychology, 153, 38 –50. doi:10.1080/
00224545.2012.703709
Hart, D., Atkins, R., & Ford, D. (1998). Urban America as a context for the
development of moral identity in adolescence. Journal of Social Issues,
54, 513–530. doi:10.1111/j.1540-4560.1998.tb01233.x
Heath, C., & Sitkin, S. B. (2001). Big-B versus Big-O: What is organizational about organizational behavior? Journal of Organizational Behavior, 22, 43–58. doi:10.1002/job.77
Hershcovis, M. S. (2011). “Incivility, social undermining, bullying . . . oh
my!”: A call to reconcile constructs within workplace aggression research. Journal of Organizational Behavior, 32, 499 –519. doi:10.1002/
job.689
Hinkin, T. (1998). A brief tutorial on the development of measures for use
in survey questionnaires. Organizational Research Methods, 1, 104 –
121. doi:10.1177/109442819800100106
Hitlan, R. T., & Noel, J. (2009). The influence of workplace exclusion and
personality on counterproductive work behaviours: An interactionist
perspective. European Journal of Work and Organizational Psychology,
18, 477–502. doi:10.1080/13594320903025028
Hoffman, M. (1989). Empathic emotions and justice in society. Social
Justice Research, 3, 283–311. doi:10.1007/BF01048080
Hoffman, M. (2000). Empathy and moral development: Implications for
caring and justice. New York, NY: Cambridge University Press. doi:
10.1017/CBO9780511805851
Jankowski, K. F., & Takahashi, H. (2014). Cognitive neuroscience of
social emotions and implications for psychopathology: Examining embarrassment, guilt, envy, and schadenfreude. Psychiatry and Clinical
Neurosciences, 68, 319 –336. doi:10.1111/pcn.12182
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1054
MITCHELL, VOGEL, AND FOLGER
Jennings, P. L., Mitchell, M. S., & Hannah, S. T. (in press). The moral self:
A review and integration of the literature. Journal of Organizational
Behavior. doi:10.1002/job.1919
Jensen, J. M., Patel, P., & Raver, J. (2014). Is it better to be average? High
and low performance as predictors of employee victimization. Journal of
Applied Psychology, 99, 296 –309. doi:10.1037/a0034822
Jordan, J., Mullen, E., & Murnighan, J. K. (2011). Striving for the moral
self: The effect of recalling past moral actions on future moral behavior.
Personality and Social Psychology Bulletin, 37, 701–713. doi:10.1177/
0146167211400208
Kim, E., & Glomb, T. M. (2010). Get smarty pants: Cognitive ability,
personality, and victimization. Journal of Applied Psychology, 95, 889 –
901. doi:10.1037/a0019985
King, E. B., Hebl, M. R., Morgan, W. B., & Ahmad, A. S. (2013). Field
experiments on sensitive organizational topics. Organizational Research
Methods, 16, 501–521. doi:10.1177/1094428112462608
Kline, R. (2005). Principles and practices of structural equation modeling.
New York, NY: Guilford Press.
Korsgaard, M. A., Brodt, S. E., & Whitener, E. M. (2002). Trust in the face
of conflict: The role of managerial trustworthy behavior and organizational context. Journal of Applied Psychology, 87, 312–319. doi:
10.1037/0021-9010.87.2.312
Lam, C. K., Van der Vegt, G. S., Walter, F., & Huang, X. (2011). Harming
high performers: A social comparison perspective of interpersonal harming in work teams. Journal of Applied Psychology, 96, 588 – 601. doi:
10.1037/a0021882
Lang, P., Kozak, M., Miller, G., Levin, D., & McLean, A. (1980). Emotional imagery: Conceptual structure and pattern of somato-visceral
response. Psychophysiology, 17, 179 –192. doi:10.1111/j.1469-8986
.1980.tb00133.x
Leach, C., Spears, R., Branscombe, N., & Doosje, B. (2003). Malicious
pleasure: Schadenfreude at the suffering of another group. Journal of
Personality and Social Psychology, 84, 932–943. doi:10.1037/00223514.84.5.932
Lee, R. M. (1993). Doing research on sensitive topics. London, United
Kingdom: Sage.
Liden, R. C., Wayne, S. J., & Stilwell, D. (1993). A longitudinal study on
the early development of leader–member exchanges. Journal of Applied
Psychology, 78, 662– 674. doi:10.1037/0021-9010.78.4.662
Little, T., Cunningham, W., Shahar, G., & Widaman, K. (2002). To parcel
or not to parcel: Exploring the question, weighing the merits. Structural
Equation Modeling, 9, 151–173. doi:10.1207/S15328007SEM0902_1
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., &
Sheets, V. (2002). A comparison of methods to test mediation and other
intervening variable effects. Psychological Methods, 7, 83–104. doi:
10.1037/1082-989X.7.1.83
Major, D. A., Turner, J. E., & Fletcher, T. D. (2006). Linking proactive
personality and the Big Five to motivation to learn and development
activity. Journal of Applied Psychology, 91, 927–935. doi:10.1037/
0021-9010.91.4.927
Malamut, A., & Offermann, L. (2001). Coping with sexual harassment:
Personal, environmental, and cognitive determinants. Journal of Applied
Psychology, 86, 1152–1166. doi:10.1037/0021-9010.86.6.1152
Martinko, M. J., Harvey, P., Brees, J. R., & Mackey, J. (2013). A review
of abusive supervision research. Journal of Organizational Behavior,
34(S), S120 –S137. doi:10.1002/job.1888
Matthews, R. A., Wayne, J. H., & Ford, M. T. (2014). A work–family
conflict/subjective well-being process model: A test of competing theories of longitudinal effects. Journal of Applied Psychology. Advance
online publication. doi:10.1037/a0036674
Mayer, R. C., & Gavin, M. (2005). Trust in management and performance:
Who minds the shop while the employees watch the boss? Academy of
Management Journal, 48, 874 – 888. doi:10.5465/AMJ.2005.18803928
Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in
survey data. Psychological Methods, 17, 437– 455. doi:10.1037/
a0028085
Miner-Rubino, K., & Cortina, L. M. (2007). Beyond targets: Consequences
of vicarious exposure to misogyny at work. Journal of Applied Psychology, 92, 1254 –1269. doi:10.1037/0021-9010.92.5.1254
Mitchell, M. S., & Ambrose, M. L. (2007). Abusive supervision and
workplace deviance and the moderating effects of negative reciprocity
beliefs. Journal of Applied Psychology, 92, 1159 –1168. doi:10.1037/
0021-9010.92.4.1159
Mitchell, M. S., Vogel, R. M., & Folger, R. (2012). Beyond the consequences to the victim: The impact of abusive supervision on third party
observers. In R. A. Giacalone & M. D. Promislo (Eds.), Handbook of
unethical work behavior: Implications for well-being (pp. 21– 43). Armonk, NY: M. E. Sharpe.
Monin, B., Pizarro, D. A., & Beer, J. S. (2007). Deciding versus reacting:
Conceptions of moral judgment and the reason-affect debate. Review of
General Psychology, 11, 99 –111. doi:10.1037/1089-2680.11.2.99
Montada, L., & Schneider, A. (1989). Justice and emotional reactions to
the disadvantaged. Social Justice Research, 3, 313–344. doi:10.1007/
BF01048081
Mullen, E., & Skitka, L. J. (2006). Exploring the psychological underpinnings of the moral mandate effect: Motivated reasoning, group differentiation, or anger? Journal of Personality and Social Psychology, 90,
629 – 643. doi:10.1037/0022-3514.90.4.629
Neuman, J., & Keashly, L. (2003, April). Workplace bullying: Persistent
patterns of workplace aggression. Paper presented at the meeting of the
Society for Industrial and Organizational Psychology, Orlando, FL.
O’Gorman, R., Wilson, R. S., & Miller, R. R. (2005). Altruistic punishing
and helping differ in sensitivity to relatedness, friendship, and future
interactions. Evolution and Human Behavior, 26, 375–387. doi:10.1016/
j.evolhumbehav.2004.12.006
Ogunfowora, B. (2013). When the abuse is unevenly distributed: The
effects of abusive supervision variability on work attitudes and behaviors. Journal of Organizational Behavior, 34, 1105–1123. doi:10.1002/
job.1841
Opotow, S. (1990). Moral exclusion and injustice: An introduction. Journal of Social Issues, 46, 1–20. doi:10.1111/j.1540-4560.1990.tb00268.x
Opotow, S. (1995). Drawing the line: Social categorization, moral exclusion, and the scope of justice. In B. Bunker & J. Rubin (Eds.), Conflict,
cooperation, and justice (pp. 347–369). San Francisco, CA: Jossey–
Bass.
Opotow, S. (2012). Moral exclusion. In D. J. Christie (Ed.), Encyclopedia
of peace psychology (pp. 672– 676). Hoboken, NJ: Wiley-Blackwell.
Opotow, S., & Weiss, L. (2000). Denial and the process of moral exclusion
in environmental conflict. Journal of Social Issues, 56, 475– 490. doi:
10.1111/0022-4537.00179
O’Reilly, J., & Aquino, K. (2011). A model of third parties’ morally
motivated responses to mistreatment in organizations. Academy of Management Review, 36, 526 –543. doi:10.5465/AMR.2011.61031810
Ostroff, C., Kinicki, A. J., & Clark, M. A. (2002). Substantive and
operational issues of response bias across levels of analysis: An example
of climate–satisfaction relationships. Journal of Applied Psychology, 87,
355–368. doi:10.1037/0021-9010.87.2.355
Pachter, B. (2003). Bosses behaving badly. Harvard Business Review, 81,
14 –15.
Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. (2012). Sources of
method bias in social science research and recommendations on how to
control it. Annual Review of Psychology, 63, 539 –569. doi:10.1146/
annurev-psych-120710-100452
Priesemuth, M. (2013). Stand up and speak out: Employees’ prosocial
reactions to observed abusive supervision. Business & Society, 52,
649 – 665. doi:10.1177/0007650313490559
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
THIRD PARTIES’ REACTIONS TO ABUSIVE SUPERVISION
Robinson, M. D., & Clore, G. L. (2001). Simulation, scenarios, and
emotional appraisal: Testing the convergence of real and imagined
reactions to emotional stimuli. Personality and Social Psychology Bulletin, 27, 1520 –1532. doi:10.1177/01461672012711012
Rodell, J. B. (2013). Finding meaning through volunteering: Why do
employees volunteer, and what does it mean for their jobs? Academy of
Management Journal, 56, 1274 –1294. doi:10.5465/amj.2012.0611
Rozin, P., Lowery, L., Imada, S., & Haidt, J. (1999). The CAD triad
hypothesis: A mapping between three moral emotions (contempt, anger,
disgust) and three moral codes (community, autonomy, divinity). Journal of Personality and Social Psychology, 76, 574 –586. doi:10.1037/
0022-3514.76.4.574
Shaver, P., Schwartz, J., Kirson, D., & O’Connor, C. (1987). Emotion
knowledge: Further exploration of a prototype approach. Journal of
Personality and Social Psychology, 52, 1061–1086. doi:10.1037/00223514.52.6.1061
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods, 7, 422– 445. doi:10.1037/1082-989X.7.4.422
Siemsen, E., Roth, A., & Oliveira, P. (2010). Common method bias in
regression models with linear, quadratic, and interaction effects. Organizational Research Methods, 13, 456 – 476. doi:10.1177/1094428109351241
Skarlicki, D. P., & Kulik, C. T. (2004). Third party reactions to employee
(mis)treatment: A justice perspective. Research in Organizational Behavior, 26, 183–229. doi:10.1016/S0191-3085(04)26005-1
Skarlicki, D. P., & Rupp, D. E. (2010). Dual processing and organizational
justice: The role of rational versus experiential processing in third party
reactions to workplace mistreatment. Journal of Applied Psychology, 95,
944 –952. doi:10.1037/a0020468
Spears, R., & Leach, C. W. (2004). Intergroupschadenfreude: Conditions
and consequences. In L. Tiedens & C. W. Leach (Eds.), The social life
of emotions (pp. 336 –355). Cambridge, United Kingdom: Cambridge
University Press. doi:10.1017/CBO9780511819568.018
Spencer, S. J., Zanna, M. P., & Fong, G. T. (2005). Establishing a causal
chain: Why experiments are often more effective than mediational
analyses in examining psychological processes. Journal of Personality
and Social Psychology, 89, 845– 851. doi:10.1037/0022-3514.89.6.845
Sundie, J. M., Ward, J. C., Beal, D. J., Chin, W. W., & Geiger-Oneto, S.
(2009). Schadenfreude as a consumption-related emotion: Feeling happiness about the downfall of another’s product. Journal of Consumer
Psychology, 19, 356 –373. doi:10.1016/j.jcps.2009.02.015
Sutton, R. (2010). Good boss, bad boss. New York, NY: Hachette Book
Group.
Takahashi, H., Kato, M., Matsuura, M., Mobbs, D., Suhara, T., & Okubo,
Y. (2009, February 13). When your gain is my pain, and your pain is my
gain: Neural correlates of envy and schadenfreude. Science, 323, 937–
939. doi:10.1126/science.1165604
Tepper, B. J. (2000). Consequences of abusive supervision. Academy of
Management Journal, 43, 178 –190. doi:10.2307/1556375
1055
Tepper, B. J. (2007). Abusive supervision in work organizations: Review,
synthesis, and research agenda. Journal of Management, 33, 261–289.
doi:10.1177/0149206307300812
Tepper, B. J., Duffy, M. K., Henle, C. A., & Lambert, L. S. (2006).
Procedural injustice, victim precipitation, and abusive supervision. Personnel Psychology, 59, 101–123. doi:10.1111/j.1744-6570.2006
.00725.x
Tepper, B. J., Moss, S. E., & Duffy, M. K. (2011). Predictors of abusive
supervision: Supervisor perceptions of deep-level dissimilarity, relationship conflict, and subordinate performance. Academy of Management
Journal, 54, 279 –294. doi:10.5465/AMJ.2011.60263085
Thomas, D. L., & Diener, E. (1990). Memory accuracy in the recall of
emotions. Journal of Personality and Social Psychology, 59, 291–297.
doi:10.1037/0022-3514.59.2.291
Turillo, C. J., Folger, R., Lavelle, J. J., Umphress, E. E., & Gee, J. O.
(2002). Is virtue its own reward? Self-sacrificial decisions for the sake of
fairness. Organizational Behavior and Human Decision Processes, 89,
839 – 865. doi:10.1016/S0749-5978(02)00032-8
van Zomeren, M., & Lodwijkx, H. (2005). Motivated responses to “senseless” violence: Explaining emotional and behavioural responses through
person and position identification. European Journal of Social Psychology, 35, 755–766. doi:10.1002/ejsp.274
van Zomeren, M., Spears, R., Fischer, A., & Leach, C. (2004). Put your
money where your mouth is! Explaining collection action tendencies
through group-based anger and group efficacy. Journal of Personality
and Social Psychology, 87, 649 – 664. doi:10.1037/0022-3514.87.5.649
Wang, M., Liu, S., Liao, H., Gong, Y., Kammeyer-Mueller, J., & Shi, J.
(2013). Can’t get it out of my mind: Employee rumination after customer
mistreatment and negative mood in the next morning. Journal of Applied
Psychology, 98, 989 –1004. doi:10.1037/a0033656
Wayne, S. J., & Ferris, G. R. (1990). Influence tactics, affect, and exchange
quality in supervisor–subordinate interactions: A laboratory experiment
and field study. Journal of Applied Psychology, 75, 487– 499. doi:
10.1037/0021-9010.75.5.487
Weiss, H. M., & Cropanzano, R. (1996). Affective events theory: A
theoretical discussion of the structure, causes and consequences of
affective experiences at work. In B. M. Straw and L. L. Cummings
(Eds.), Research in organizational behavior (Vol. 18, pp. 1–74). Greenwich, CT: JAI Press.
Westring, A. F., & Ryan, A. M. (2010). Personality and inter-role conflict
and enrichment: Investigating the mediating role of support. Human
Relations, 63, 1815–1834. doi:10.1177/0018726710371236
Williams, L., Hartman, N., & Cavazotte, F. (2010). Method variance and
marker variables: A review and comprehensive CFA marker technique.
Organizational Research Methods, 13, 477–514. doi:10.1177/
1094428110366036
Received February 9, 2014
Revision received August 20, 2014
Accepted August 21, 2014 䡲