Professional Documents
Culture Documents
In this study, the authors examine the relationship between abusive supervision and employee workplace
deviance. The authors conceptualize abusive supervision as a type of aggression. They use work on
retaliation and direct and displaced aggression as a foundation for examining employees’ reactions to
abusive supervision. The authors predict abusive supervision will be related to supervisor-directed
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
deviance, organizational deviance, and interpersonal deviance. Additionally, the authors examine the
This document is copyrighted by the American Psychological Association or one of its allied publishers.
moderating effects of negative reciprocity beliefs. They hypothesized that the relationship between
abusive supervision and supervisor-directed deviance would be stronger when individuals hold higher
negative reciprocity beliefs. The results support this hypotheses. The implications of the results for
understanding destructive behaviors in the workplace are examined.
In the last decade, there has been increased interest in harmful standing employee reactions. From a justice perspective, employ-
or destructive behaviors in organizations. Much of this research ees react to the perceived unfairness of the abusive supervisor’s
focuses on deviant behaviors of employees. (See Bennett & Rob- behavior. When employees feel they are treated unfairly, positive
inson, 2003, for a review.) However, recently, research has exam- attitudes and behavior suffer (Tepper, 2000; Tepper et al., 1998).
ined destructive behaviors managers commit—specifically, abu- Researchers also have used reactance theory as a foundation for
sive supervision (e.g. Tepper, 2000; Tepper, Duffy, Hoobler, & understanding employee reactions to abusive supervision (Zellars
Ensley, 2004; Tepper, Duffy, & Shaw, 2001). In this article, we et al., 2002). Reactance theory suggests that individuals strive to
consider the relationship between these two types of destructive maintain personal control (Brehm & Brehm, 1981; Wright &
behavior. Brehm, 1982). Researchers suggest that employees dealing with an
Recent research by Tepper and his colleagues (Tepper, 2000; abusive supervisor usually feel little or no control. As a result,
Tepper et al., 2001, 2004; Zellars, Tepper, & Duffy, 2002) has individuals engage in behavior to restore personal autonomy (e.g.,
focused attention on abusive supervision. Tepper (2000) defined decrease organizational citizenship behaviors; Zellars et al., 2002).
abusive supervision as the “subordinates’ perceptions of the extent Both of these approaches are useful for understanding individ-
to which their supervisors engage in the sustained display of uals’ reactions to abusive supervision. However, they do not
hostile verbal and nonverbal behaviors, excluding physical con- capture the uniquely aggressive and hostile behavior that defines
tact” (p. 178).1 Ashforth (1997) described abusive managers as abusive supervision. In this article, we conceptualize abusive su-
those who callously and arbitrarily use their power and authority to pervision as a type of aggression (behaviors perceived by the
mistreat employees. Abusive supervisors are known to use derog-
employee as intentionally harmful; Baron, 2005). We use work on
atory names, yell and scream, intimidate, withhold needed infor-
retaliation and direct and displaced aggression as a foundation for
mation, and humiliate and ridicule their employees (Keashly,
examining employees’ reactions to abusive supervision. We inves-
1998). Empirical research usually examines abuse from the sub-
tigate the relationship between abusive supervision and employee
ordinate’s perspective (Ashforth, 1997; Tepper, 2000; Tepper,
deviance directed at the supervisor, the organization, and other
Eisenbach, Kirby, & Potter, 1998; Zellars et al., 2002), and we take
individuals.
that perspective in this research.
Further, we consider the moderating effects of negative reci-
Research on abusive supervision has generally taken either an
procity beliefs. A negative reciprocity orientation is the tendency
organizational justice or a reactance theory approach to under-
for an individual to return negative treatment for negative treat-
ment (Cropanzano & Mitchell, 2005). We suggest individuals with
Marie S. Mitchell and Maureen L. Ambrose, Department of Manage-
ment, University of Central Florida.
1
We thank Ben Tepper for his generous contributions, specifically data Tepper (2000) also suggests abusive supervision involves indifference,
for reducing the Abusive Supervision measure. We also thank Becky for example, speaking rudely to subordinates in order to elicit desired task
Bennett, Russell Cropanzano, and Marshall Schminke for their feedback on performance (p. 179). However, most research focuses on the more pro-
earlier versions of this article. active, willfully hostile behavior (e.g., Ashforth, 1994, 1997; Bies & Tripp,
Correspondence concerning this article should be addressed to Marie S. 1998a, 1998c; Namie & Namie, 2000). Indeed, researchers often refer to
Mitchell, who is now at the Department of Management, University of abusive supervisors as “managerial bullies” (Ashforth, 1994, 1997; Namie
Nebraska–Lincoln, P.O. Box 880491, Lincoln, NE 68588-0491. E-mail: & Namie, 2000; Salin, 2001). We focus explicitly on the active hostile
mmitchell@unlnotes.unl.edu behavior in this study.
1159
1160 RESEARCH REPORTS
stronger negative reciprocity beliefs are more likely to direct deviance and (nonsupervisory) interpersonal deviance as well as
deviant behavior toward the perceived source of harm. Thus, in the organizational deviance when considering employee reactions to
case of abusive supervision, stronger negative reciprocity beliefs abusive supervision.
should be associated with increased deviant behavior directed at
the supervisor. In the following sections, we review relevant Abusive Supervision and Employee Deviance
literature on abusive supervision, employee deviance, and negative
reciprocity beliefs. Interpersonal treatment is a driving factor in deviant behavior
(Robinson & Greenberg, 1998). Workplace experiences such as
Abusive Supervision frustration, injustices, and threats to self are primary antecedents to
employee deviance (Bennett & Robinson, 2003). Ashforth (1997)
Although abusive supervision is a low base-rate phenomenon, it suggested that abusive supervision promotes feelings of frustra-
has notable effects on employee attitudes (Tepper, 2000). Research tion, helplessness, and alienation. Tepper (2000) found that abu-
shows that abusive supervision is related to lower levels of satis- sive supervision negatively influences perceptions of justice. Thus,
faction, commitment, and justice perceptions, and higher levels of abusive supervision is a likely antecedent of employee deviance.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
turnover, role conflict, and psychological distress (Ashforth, 1997; As we noted above, we expect abusive supervision will be
Duffy, Ganster, & Pagon, 2002; Tepper, 2000). related to employee workplace deviance in two ways. First, em-
Fewer research studies have investigated the effects of abusive ployees may respond by directly retaliating against their supervi-
supervision on employee behaviors. Two studies by Tepper and sor. Second, employees may engage in “displaced” deviance by
colleagues (Tepper et al., 2004; Zellars et al., 2002) suggested that targeting the organization or other individuals. We discuss each of
abusive supervision negatively affects organizational citizenship these below.
behaviors. Employees subjected to an abusive supervisor engage in
fewer organizational citizenship behaviors (Zellars et al., 2002). Supervisor-Directed Deviance
Further, abusive supervision also negatively affects how employ-
ees perceive the genuineness of their peers’ organizational citizen- Retaliation plays an important role in research on aggression as
ship behaviors (Tepper et al., 2004). In addition to decreased well as research on workplace deviance. Retaliation involves the
positive behavior, we suggest abusive supervision will increase desire to punish an offender for unwarranted and malicious acts
negative behavior, specifically, employee workplace deviance. (Averill, 1982). Retaliation refers to behavior that seeks to “make
In considering the relationship between abusive supervision and the wrongdoer pay” for a transgression or event that harms or
employee deviance, we found research on aggression and retalia- jeopardizes the victim in some meaningful way (Skarlicki &
tion to be useful. This research suggests that interpersonal mis- Folger, 2004, p. 374).2
treatment (like abusive supervision) promotes retaliation and ag- Research on aggression demonstrates individuals may respond
gression displaced on other targets. For example, in their study on to the aggressive behavior of others by choosing to retaliate. For
injustice and retaliation, Skarlicki and Folger (1997) found that example, Brown (1968) found severe offensive behavior (social
conditions of multiple unfairness (distributive, procedural, and humiliation) resulted in strong retaliatory reactions, even at a
interactional) were associated with higher levels of organizational personal cost to the retaliator. Bies and Tripp (1996) found that
retaliatory behavior. Notably, these behaviors are characterized by individuals seek revenge against those who harm them. A recent
both direct and displaced methods of retaliation (e.g., disobeyed meta-analysis demonstrates that when individuals attribute respon-
supervisor’s instructions, left a mess unnecessarily, spread rumors sibility to a harmdoer, they respond with anger and retaliation
about coworkers). We suggest employees engage in deviant be- (Rudolph, Roesch, Greitemeyer, & Weiner, 2004).
havior to retaliate directly against their abusive supervisor, and Several researchers have focused on investigating antecedents
they may also engage in displaced deviant behavior. Moreover, we of retaliation in organizations (e.g., Allred, 1999; Bies & Tripp,
expect negative reciprocity beliefs to affect the relationship be- 1998b; Bies, Tripp, & Kramer, 1997; Folger & Baron, 1996;
tween abusive supervision and deviance. Skarlicki & Folger, 1997, 2004). Empirical evidence demonstrates
that individuals retaliate against perceived injustices (Greenberg &
Workplace Deviance Alge, 1998; Skarlicki & Folger, 1997; Skarlicki, Folger, & Tesluk,
1999), threats to identity (Aquino & Douglas, 2003), violations of
Workplace deviance is purposeful behavior that violates orga- trust (Bies & Tripp, 1996), and personal offense (Aquino, Tripp, &
nizational norms and is intended to harm the organization, its Bies, 2001). When individuals feel they have been mistreated,
employees, or both (Bennett & Robinson, 2003). Robinson and retaliation is a deliberate, rational response (Bies & Tripp, 1996).
Bennett (1995) developed a widely accepted typology of work-
place deviance, which categorizes two basic types of deviance:
2
organizational and interpersonal. Organizational deviance is devi- Our use of the term retaliation stems from the aggression literature. In
ance directed toward the organization (e.g., shirking hours, pur- organizational research, scholars have also used the term revenge to de-
scribe behavior that is intended to punish another for an offense (see, for
posefully extending overtime), and interpersonal deviance is de-
example, Bies & Tripp, 1996, 1998a, 1998b, 1998c). Additionally, the term
viance directed toward individuals (e.g., verbal abuse, sexual retaliation has been used in equal employment opportunity law and the
harassment). Recent research suggests it is useful to distinguish whistle-blowing literature, in which retaliation is explicated as behavior
between two types of interpersonal deviance: deviant behaviors that seeks to punish an employee who engages in protected behavior
targeted against supervisors and those targeted at other individuals (Crockett & Gilmere, 1999). This use describes a specific instance of
(Hershcovis et al., 2007). Thus, we investigate supervisor-directed retaliatory behavior.
RESEARCH REPORTS 1161
Interpersonal mistreatment is a central component of abusive Hypothesis 1c: Abusive supervision will be positively related
supervision, and research indicates employees perceive supervi- to (nonsupervisory) interpersonal deviance.
sors as a dominant source of interpersonal mistreatment (Bies,
1999). Supervisors are reported to be the most prominent source of
bullying at work (Neuman & Keashly, 2003). Indeed, both theo- Moderating Effects of Negative Reciprocity Beliefs
retical and empirical research suggests abusive supervision is
Retaliation plays an important role in our conceptualization of
related to retaliation. For example, Folger (1993) proposed that
the relationship between abusive supervision and supervisor-
supervisors who fail to meet an acceptable standard of demeanor
directed deviance. The principle of retaliation emphasizes the
promote retaliation. Bies and Tripp (1998b) found that victims of
biblical injunction of “a life for a life, eye for eye, tooth for tooth
abusive bosses directly undermined their bosses in private as well
. . . bruise for bruise” (Exodus 21:23–25) and is a common theme
as openly ridiculed or challenged them. Baron and Neuman (1998) in deviance research (cf. Bennett & Robinson, 2003). Gouldner
reported 31.4% of their respondents displayed aggression against a (1960) captured this principle in his negative norm of reciprocity.
supervisor and felt justified doing so. Aquino, Tripp, and Bies Reciprocity encompasses quid pro quo behaviors, meaning that
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
(2006) demonstrated that lower level individuals are more likely to something given generates an obligation to return an equivalent
This document is copyrighted by the American Psychological Association or one of its allied publishers.
seek revenge than higher level individuals. Jones (2003) found that gesture. Most research focuses on positive reciprocity, which
interactional injustice from an authority was significantly related promotes stability in relationships through considerate, valued, and
to supervisor-directed retaliation. Further, a recent meta-analysis balanced exchanges. Favorable treatment generates favorable
by Hershcovis et al. (2007) found that unfair supervisor treatment treatment.
was a strong predictor of supervisor-targeted aggression. However, Gouldner (1960) also noted that individuals may
In the workplace deviance literature, retaliation is conceptual- endorse a negative norm of reciprocity, under which unfavorable
ized as an interpersonal form of deviance (Bennett & Robinson, treatment promotes “not the return of benefits but the return of
2003). By definition, retaliation involves deliberate actions against injuries” (p. 172). Indeed, individuals may be guided by negative
a perceived harmdoer. In the case of abusive supervision, these reciprocity beliefs whereby they believe that when someone mis-
behaviors would be targeted against the supervisor (i.e., treats them, it is acceptable to retaliate in return (Cropanzano &
supervisor-directed deviance). Therefore, we believe abusive su- Mitchell, 2005). Yet, Gouldner suggested that not all victims seek
pervision will be associated with supervisor-directed deviance. to retaliate. Some may feel it is acceptable to “turn the other
cheek.” Thus, individuals vary in their beliefs about the appropri-
Hypothesis 1a: Abusive supervision will be positively related ateness of negative reciprocity.
to supervisor-directed deviance. Individuals who endorse negative reciprocity believe retribution
is the correct and proper response to unfavorable treatment (Eisen-
Displaced Deviance berger, Lynch, & Aselage, 2004). Those who hold strong negative
reciprocity beliefs are more likely to seek retaliation than avoid-
We also expect that abusive supervision will be related to other ance (McLean Parks, 1998). Those who do not hold strong nega-
types of deviance. That is, in addition to targeting the source of the tive reciprocity beliefs are less likely to engage in retaliatory
abuse, employees will react toward other targets. They will engage behavior. Indeed, research demonstrates that individuals vary in
in deviance directed toward the organization (organizational devi- their beliefs about the appropriateness of negative reciprocity.
ance) or individuals other than the supervisor (interpersonal devi- Moreover, individuals’ negative reciprocity beliefs influence be-
ance). The theory of displaced aggression guides our thinking here havioral choices (Gallucci & Perugini, 2003; Perugini, Gallucci,
(Dollard, Doob, Miller, Mowrer, & Sears, 1939) Presaghi, & Ercolani, 2003).
Research on displaced aggression suggests that individuals who In sum, individuals with strong negative reciprocity beliefs
become angry and frustrated by a harmdoer may displace their consider retaliation an appropriate response to negative treatment.
aggression on individuals who are not the source of the harm We suggest negative reciprocity beliefs will moderate the relation-
(Dollard et al., 1939). Dollard et al. (1939) offered two reasons ship between abusive supervision and employee deviance. How-
why individuals displace aggression. First, the harmdoer may not ever, because negative reciprocity is a quid pro quo belief, the
be available to retaliate against. Second, the victim may fear focus of retaliation should be the source of the mistreatment (the
further retaliation from the harmdoer. Should either of these con- abusive supervisor). Thus, we suggest that for individuals with
straints occur, direct retaliation is curbed (Baron, 1971) and ag- stronger negative reciprocity beliefs, the relationship between per-
ceived abusive supervision and supervisor-directed deviance will
gressive behaviors may be redirected or displaced on less powerful
be stronger than for those who do not endorse negative reciprocity.
or more available targets (e.g., coworkers; Miller, 1941). Thus,
However, we do not expect negative reciprocity beliefs to affect
retaliation is only one behavior employees may choose to engage
the relationship between abusive supervision and deviance toward
in as a consequence of perceived abuse; victims may also displace
the organization or toward individuals other than the supervisor.
their hostilities on others. We suggest that individuals subjected to
abusive supervision may displace aggression toward the organiza-
tion (i.e., organizational deviance)3 and individuals other than the 3
Some researchers suggest that individuals may target the organization
supervisor (i.e., interpersonal deviance). in an effort to retaliate against a supervisor (Ambrose, Seabright, &
Schminke, 2002). In our conceptualization, this is still displaced aggres-
Hypothesis 1b: Abusive supervision will be positively related sion, because the retaliatory attempt is indirect rather than direct. The target
to organizational deviance. of the aggression differs from the source of the harm.
1162 RESEARCH REPORTS
Hypothesis 2: Negative reciprocity beliefs will moderate the deviance, which asked respondents to indicate behaviors targeted
relationship between abusive supervision and supervisor- against their current supervisor.5 These items are shown in Ap-
directed deviance but will not moderate the relationship be- pendix C.
tween abusive supervision and other forms of deviance (or- Negative reciprocity beliefs. Negative reciprocity beliefs were
ganizational or interpersonal). Abusive supervision will be assessed with a 14-item measure developed by Eisenberger et al.
more strongly related to supervisor-directed deviance when (2004). The measure contains statements concerning the advisabil-
individuals more strongly believe in negative reciprocity. ity of retribution for unfavorable treatment. Respondents were
asked to rate their agreement on a 7-point scale (1 ⫽ strongly
Method agree, 7 ⫽ strongly disagree).
Controls. Past research suggests that trait anger influences
Sample and Procedure deviant reactions (Eisenberger et al., 2004; Fox & Spector, 1999).
Therefore, we controlled for trait anger in our analyses. We used
Surveys were distributed to individuals called for jury duty by a the anger subscale of Buss and Perry’s (1992) Aggression Ques-
county circuit court in the Southeastern United States. The re-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
searchers addressed potential jurors at the beginning of the day as in everyday life. Participants expressed agreement on a 5-point
they waited to learn if they would be required to serve. We scale (1 ⫽ very slightly true of me, 5 ⫽ very highly true of me). We
explained that the survey had nothing to do with the jury or court controlled for age, because research also suggests that age is
system, but rather, that we sought to understand more about related to deviant reactions (Aquino & Douglas, 2003; Grasmick &
sensitive issues that affect individuals at work. Therefore, we Kobayashi, 2002). Last, we controlled for employees’ tenure with
indicated that willing participants had to be currently employed. the supervisor, because research suggests that tenure with the
Interested participants picked up surveys from and returned sur- supervisor influences reactions (Bauer & Green, 1996; Wayne,
veys to the researcher. Over the course of 8 weeks, 427 individuals Shore, & Liden, 1997).
agreed to participate in the study (30.5% response rate). The Social desirability check. In order to assess for socially desir-
average age of the participants was 42.7 years old (SD ⫽ 11.95); able responses in the deviance, abusive supervision, negative rec-
average company tenure was 8.7 years (SD ⫽ 8.15), and tenure iprocity beliefs, and trait anger measures, we examined the corre-
with a supervisor was 3.7 years (SD ⫽ 4.08); 38.3% were currently lation between individual self-reported items to those of social
working in supervisory positions; and 56.9% of the sample was desirability. We assessed social desirability with an 18-item short
female. version of the Paulhus (1991) measure, which has been used in
previous research investigating workplace deviance (Tripp, Bies,
Measures & Aquino, 2002).
Consistent with previous research (Aquino et al., 1999), we
Abusive supervision. We used a shortened 5-item version of eliminated items that correlated greater than .30 with the Paulhus
Tepper’s (2000) Abusive Supervision measure. To develop this items. Thus, four negative reciprocity items were eliminated from
shortened measure, we performed exploratory and confirmatory the original set: “If a person wants to be your enemy, you should
factor analyses on two separate data sets that used the original
15-item measure (N ⫽ 741 from Tepper, 2000, and N ⫽ 338 from
Tepper et al., 2004).4 (See Appendixes A and B for results.) The 4
We thank Ben Tepper for sharing his data with us for these analyses.
analyses revealed two distinct factors: One reflected active inter- 5
Although research suggests distinguishing between deviant behaviors
personal abuse by the supervisor (e.g., “ridicules me” and “tells me targeted against supervisors and those targeted at other individuals
my thoughts and feelings are stupid”). The other reflected more (Hershcovis et al., 2007), the interpersonal deviance measures had not been
passive acts of abuse (e.g., “doesn’t give me credit for jobs adapted in this way previously. To ensure that these three measures of
requiring a lot of effort”). Because the active dimension is most deviance were distinct, we conducted a series of confirmatory factor
consistent with our interest, we used the 5 items representing this analyses. First, we examined a three-factor model that included the three
separate deviance measures (interpersonal deviance, organizational devi-
factor as our indicator of abusive supervision. Respondents indi-
ance, and supervisor-directed deviance). This model provided an accept-
cated their agreement with each item using a 7-point scale (1 ⫽ able fit to the data, 2(374) ⫽ 1,689.05, p ⬍ .001, RMSEA ⫽ .09, CFI ⫽
strongly agree, 7 ⫽ strongly disagree). .90, NFI ⫽ .88. We compared the three-factor model to (a) a two-factor
Deviance. We assessed interpersonal and organizational devi- model that combined supervisor-directed deviance and interpersonal devi-
ance with measures developed by Bennett and Robinson (2000). ance into a single interpersonal factor, 2(376) ⫽ 2,094.32, p ⬍ .001,
The measures used a 7-point scale (1 ⫽ never, 7 ⫽ daily) and RMSEA ⫽ .10, CFI ⫽ .88; NFI ⫽ .85, (b) a two-factor model that
asked respondents to indicate the number of times in the last year combined interpersonal deviance and organizational deviance into a single
that they had engaged in the behavior described. Twelve items displaced deviance factor, 2(376) ⫽ 2,301.37, p ⬍ .001, RMSEA ⫽ .11,
assessed perceptions of organizational deviance. Respondents CFI ⫽ .88, NFI ⫽ .85, and (c) a single-factor model that combined all
were asked to indicate behaviors targeted at the company for deviance items into one deviance factor, 2(377) ⫽ 2,522.09, p ⬍ .001,
RMSEA ⫽ .12, CFI ⫽ .86, NFI ⫽ .84. The results of our analysis revealed
which they were currently working. Seven items assessed percep-
the three-factor model produced a significant improvement in chi-squares
tions of interpersonal deviance. Respondents were asked to indi- over the two-factor interpersonal deviance model, ⌬2(2) ⫽ 405.27, p ⬍
cate behaviors targeted at coworkers. Further, we adapted lan- .001; the two-factor displaced deviance model, ⌬2(2) ⫽ 612.32, p ⬍ .001;
guage from the interpersonal deviance items from both the Bennett and the one-factor model, ⌬2(3) ⫽ 833.04, p ⬍ .001, suggesting the
and Robinson (2000) and Aquino, Lewis, and Bradfield (1999) three-factor model provided a better fit than the alternative models (Schu-
measures to generate a 10-item measure of supervisor-directed macker & Lomax, 1996).
RESEARCH REPORTS 1163
treat them like an enemy” (r ⫽ .37), “If someone treats you badly, suggest the data are not normally distributed. Although ordinary
you should treat that person badly in return” (r ⫽ .36), “When least squares is robust to nonnormality (Mertler & Vannatta,
someone hurts you, you should find a way they won’t know about 2002), we conducted a transformation for negatively skewed data
to get even” (r ⫽ .37), and “If someone treats me badly, I feel I (Tabachnick & Fidell, 1996) and also analyzed the transformed
should treat them even worse” (r ⫽ .39). The Cronbach alpha data. The results of these analyses are consistent with the results
coefficient for the negative reciprocity beliefs measure without presented below.
these four items was .86. All other items showed low correlations Regression results are provided in Table 2. Hypothesis 1 pro-
with social desirability (i.e., r ⬍ .30) and were retained in our poses that abusive supervision will be positively related to (a)
analyses. supervisor-directed deviance, (b) organizational deviance, and (c)
interpersonal deviance. The results support this hypothesis. Abu-
Results sive supervision is positively and significantly related to each type
of deviance. Further, trait anger was significantly and positively
Measurement Model Results related to supervisor-directed deviance and interpersonal deviance.
Additionally, negative reciprocity beliefs were significantly and
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Table 1
Summary Statistics and Zero-Order Correlations
Variable M SD 1 2 3 4 5 6 7 8
Note. Internal reliability coefficients (alphas) appear in parentheses along the main diagonal. Correlations greater than .14 are significant at p ⬍ .01 and
those greater than .10, at p ⬍ .05, two-tailed.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
supervisor-directed deviance than individuals who did not endorse theories of frustration-aggression (Dollard et al., 1939) and devi-
negative reciprocity. The results also show that negative reciproc- ance. (See Perrowe & Spector, 2002 for a review.) The results of
ity beliefs did not significantly influence the relationship between this study suggest that trait anger is differentially related to differ-
abusive supervision and other types of deviant behavior (neither ent types of deviance. These results are consistent with Douglas
organizational nor interpersonal deviance). Thus, as expected, and Martinko (2001), who have argued that high trait-anger indi-
negative reciprocity is significantly related only to the quid pro viduals believe others purposefully and unnecessarily caused them
quo behaviors targeted against the abuser. harm. They found that individuals with high trait-anger were more
Our results also reveal some unexpected findings. First, there inclined to feel vengeful and make hostile attributions about other
was a significant main effect for negative reciprocity beliefs on all persons. Additionally, a recent meta-analysis on workplace aggres-
types of deviance. Whereas the interaction findings support the sion found that trait anger was more strongly related to interper-
belief that negative reciprocity promotes retribution, the main sonal than organizational forms of aggression (Hershcovis et al.,
effect indicates that individuals who hold strong negative reciproc- 2007). Similarly, we found that trait anger is related only to
ity beliefs also reported greater levels of all types of deviance. This deviance directed at individuals. Thus, it seems that interpersonal
finding is consistent with Eisenberger et al. (2004), who have interaction and how it is interpreted by individuals with high trait
argued that individuals with strong negative reciprocity beliefs anger is important to understanding trait anger and behavior.
hold a catharsis approach to aggression and get satisfaction from The results of this study have important implications for re-
aggressing against any target, whether guilty or innocent. Thus, searchers and managers. Most notable, the results suggest that
negative reciprocity beliefs may also indicate a general propensity abusive supervision is positively associated with workplace devi-
toward workplace deviance. ance. Below, we consider two ways in which abusive supervision
The findings on trait anger are also interesting. Trait anger was might affect the overall level of workplace deviance.
positively related to supervisor-directed and interpersonal devi- First, abusive supervision is associated with higher levels of
ance, but not organizational deviance. Anger has been linked to deviance directed at others. In our study, we explicitly examined
Table 2
Multiple Regressions of Hypothesized Relationships
Variable Step 1 Step 2 Step 3 Step 1 Step 2 Step 3 Step 1 Step 2 Step 3
Control
Trait anger .26*** .18*** .18*** .11** .00 ⫺.01 .30*** .22*** .22***
Age ⫺.15*** ⫺.17*** ⫺.16*** ⫺.27*** ⫺.25*** ⫺.24*** ⫺.16*** ⫺.15*** ⫺.15***
Tenure with supervisor .08 .09* .09** .03 .03 .04 .09 .10** .10**
Predictor
Abusive supervision .39*** .37*** .15*** .14*** .16*** .15***
Negative reciprocity .11** .10** .28*** .27*** .18*** .18***
Moderator
Abusive Supervision ⫻
Negative Reciprocity .08* .05 .07
⌬ R2 .17*** .01* .10*** .00 .06*** .01
2
R .09 .26 .27 .08 .18 .19 .12 .19 .19
Adjusted R2 .09 .25 .26 .08 .17 .17 .12 .18 .18
F 13.47*** 27.31*** 23.40*** 11.73*** 17.06*** 14.26*** 18.02*** 17.60*** 15.12***
Note. Standardized beta coefficients are reported. Statistical tests are based on one-tailed tests.
*
p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
RESEARCH REPORTS 1165
Andersson, L. M., & Pearson, C. M. (1999). Tit for tat? The spiraling effect the bad, and the ugly. In R. W. Griffin, A. O’Leary-Kelly, & J. M.
of incivility in the workplace. Academy of Management Journal, 24, Collins (Eds.), Dysfunctional behavior in organizations: Non-violent
452– 471. dysfunctional behavior (pp. 49 – 67). Stamford, CT: JAI Press.
Aquino, K., & Douglas, S. (2003). Identity threat and antisocial behavior Bies, R. J., & Tripp, T. M. (1998c). Two faces of the powerless: Coping
in organizations: The moderating effects of individual differences, ag- with tyranny in organizations. In R. M. Kramer & M. A. Neale (Eds.),
gressive modeling, and hierarchical status. Organizational Behavior and Power and influence in organizations (pp. 203–220). Thousand Oaks,
Human Decision Processes, 90, 195–208. CA: Sage.
Aquino, K., Lewis, M. U., & Bradfield, M. (1999). Justice constructs, Bies, R. J., Tripp, T. M., & Kramer, R. M. (1997). At the breaking point:
negative affectivity, and employee deviance: A proposed model and Cognitive and social dynamics of revenge in organizations. In R. A.
empirical test. Journal of Organizational Behavior, 20, 1073–1091. Giacalone & J. Greenberg (Eds.), Antisocial behavior in organizations
Aquino, K., Tripp, T. M., & Bies, R. J. (2001). How employees respond to (pp. 18 –36). Thousand Oaks, CA: Sage.
personal offense: The effects of blame attribution, victim status, and Bollen, K. A. (1989). Structural equations with latent variables. New
offender status on revenge and reconciliation in the workplace. Journal York: Wiley.
of Applied Psychology, 86, 52–59. Brehm, J., & Brehm, S. (1981). Psychological resistance: A theory of
Aquino, K., Tripp, T. M., & Bies, R. J. (2006). Getting even or moving on?
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Inness, M., et al. (2007). Predicting workplace aggression: A meta- Rudolph, U., Roesch, S. C., Greitemeyer, T., & Weiner, B. (2004). A
analysis. Journal of Applied Psychology, 92, 228 –238. meta-analytic review of help giving and aggression from an attributional
Hoyle, R. H., & Panter, A. T. (1995). Writing about structural equation perspective: Contributions to a general theory of motivation. Cognition
models. In R. H. Hoyle (Ed.), Structural equation modeling: Concepts, & Emotion, 18, 815– 848.
issues, and applications (pp. 158 –176). Thousand Oaks, CA: Sage. Ryan, T. P. (1997). Modern regression methods. New York: Wiley.
Jones, D. A. (2003). Predicting retaliation in the workplace: The theory of Salin, D. (2001). Prevalence and forms of bullying among business pro-
planned behavior and organizational justice. In D. H. Nagao (Ed.), Best fessionals: A comparison of two different strategies for measuring
paper proceedings of the 63rd Annual Meeting of the Academy of bullying. European Journal of Work and Organizational Psychology,
Management [CD], L1– 6. Briarcliff Manor, NY: Academy of Manage- 10, 425– 441.
ment. Schumacker, R. E., & Lomax, R. G. (1996). A beginner’s guide to struc-
Keashly, L. (1998). Emotional abuse in the workplace: Conceptual and tural equation modeling. Mahwah, NJ: Erlbaum.
empirical issues. Journal of Emotional Abuse, 1, 85–117. Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: The role
Lee, R. M. (1993). Doing research on sensitive topics. London: Sage. of distributive, procedural, and interactional justice. Journal of Applied
McLean Parks, J. (1998). The fourth arm of justice: The art and science of Psychology, 82, 434 – 443.
revenge. Research on Negotiation in Organizations, 6, 113–144. Skarlicki, D. P., & Folger, R. (2004). Broadening our understanding of
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
(Appendixes follow)
1168 RESEARCH REPORTS
Appendix A
Exploratory Factor Analysis Factor Loadings for Tepper’s (2000) Abusive Supervision
Measure
Factor
Item 1 2
a
1. Ridicules me. .23 .71
2. Tells me my thoughts or feelings are stupid.a .28 .61
3. Gives me the silent treatment. .45 .49
4. Puts me down in front of others.a .30 .79
5. Invades my privacy. .58 .29
6. Reminds me of my past mistakes and failures. .46 .50
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
7. Doesn’t give me credit for jobs requiring a lot of effort. .60 .29
This document is copyrighted by the American Psychological Association or one of its allied publishers.
Note. Dominant factor loadings are presented in boldface. Factor 1 ⫽ passive-aggressive abusive supervision; Factor
2 ⫽ active-aggressive abusive supervision. Factors accounted for 48.1% and 9.0% of the total variance, respectively
(N ⫽ 741). Data are from “Consequences of Abusive Supervision,” by B. J. Tepper, 2000, Academy of Management
Journal, 43, pp. 189 –190. Copyright 2000 by the Academy of Management. Reprinted with permission.
a
Items retained for the shortened measure that we used in the current study.
Appendix B Appendix C
Confirmatory Factor Analysis Results for Tepper’s (2000) Supervisor-Directed Deviance Measure
Abusive Supervision Measure
Item
Model 2 df ⌬2 CFI NFI RMSEA
1. Made fun of my supervisor at work.
2-factor 195.40 34 .96 .96 .11 2. Played a mean prank on my supervisor.
1-factor 336.92 35 141.52*** .95 .94 .16 3. Made an obscene comment or gesture toward my supervisor.
4. Acted rudely toward my supervisor.
Note. N ⫽ 338. All chi-square values are significant at p ⬍ .001. CFI ⫽ 5. Gossiped about my supervisor.
comparative fit index; NFI ⫽ normed fit index; RMSEA ⫽ root-mean- 6. Made an ethnic, religious, or racial remark against my supervisor.
square error of approximation. Data are from Tepper et al. (2004). 7. Publicly embarrassed my supervisor.
***
p ⬍ .001, two-tailed. 8. Swore at my supervisor.
9. Refused to talk to my supervisor.
10. Said something hurtful to my supervisor at work.