Você está na página 1de 17

Journal of Organizational Behavior

J. Organiz. Behav. 21, 147162 (2000)

Anger in the workplace: an emotion script


approach to anger episodes between
workers and their superiors, co-workers
and subordinates
JULIE FITNESS*
Department of Psychology, Macquarie University, Sydney 2109, Australia

Summary

The overall aim of this study was to investigate laypeople's accounts of the causes,
features, and consequences of workplace anger episodes and to examine similarities and
dierences amongst superiors', co-workers' and subordinates' anger experiences. One
hundred and seventy-ve respondents participated in structured interviews about a
work-related anger episode with a superior, co-worker, or subordinate. Various features
of the anger episodes diered according to the status of the respondent, with superiors
angered by morally reprehensible behaviors and job incompetence, co-workers angered
by morally reprehensible behaviors and public humiliation, and subordinates angered by
unjust treatment. Subordinates were less likely than superiors to confront the anger
target and more likely to consider the incident unresolved. Humiliating oences elicited
more intense hate than non-humiliating oences; hate was also negatively associated
with situational power and with a perceived successful resolution of the anger-eliciting
event. Theoretical implications of the results are discussed in relation to the role of power
in the experience and expression of anger in the workplace. Copyright # 2000 John
Wiley & Sons, Ltd.

Introduction
Understanding the causes, characteristics, and consequences of emotions in the workplace is one
of the most important, though neglected, areas of organizational psychology. Of course, the
neglect of emotions by psychologists is not peculiar to the organizational eld. As Lazarus (1991)
noted, despite thousands of years of conjecture by philosophers, theologians, and even dramatists, about the role of emotions in human experience, the scientic study of emotions within
academic psychology is a relatively recent phenomenon. Consequently it is not surprising that
emotions and emotional phenomena have, in turn, only recently begun to attract attention from
scholars with an interest in the social psychology of the workplace; a context that, with its
inherent power structures and relational complexities, has been described as `saturated' with
* Correspondence to: Julie Fitness, Department of Psychology, Macquarie University, Sydney 2109, Australia.
E-mail: Jtness@bunyip.psy.mq.edu.au
Thanks are due to Sally Mathews and Helen Lawson Williams for their assistance in data coding.

CCC 08869383/2000/02014716$17.50
Copyright # 2000 John Wiley & Sons, Ltd.

148

J. FITNESS

emotion (Ashforth and Humphrey, 1995). Moreover, given that the workplace has been
identied as one of the most interpersonally frustrating contexts that people have to deal with
(Allcorn, 1994; Bensimon, 1997), it might be expected that anger, in particular, will be a
frequently experienced (if not always expressed) workplace emotion.
Researchers have linked workplace anger to a number of negative outcomes, including violence
and hostility (Folger and Baron, 1996), antisocial behaviors such as theft (Chen and Spector,
1992), and organizational retaliatory behaviors, or revenge (Bies and Tripp, 1998; Skarlicki and
Folger, 1997). However, although there is a growing body of research within the sociological and
organizational literatures on the expression and regulation of emotions in dierent occupational
contexts, such as between customers and service providers (e.g., Hochschild, 1983; Rafaeli and
Sutton, 1989; 1991), there is still relatively little known about the antecedents, characteristics, and
outcomes of anger episodes between workers within organizations (Glomb and Hulin, 1997).
Thus, the overall aim of this study was to investigate the causes, features, and consequences of
anger at work and to examine the similarities and dierences amongst the reported anger
experiences of superiors, co-workers, and subordinates.

Theoretical approach to the study of anger at work


Several dierent approaches have been taken to the scientic study of emotion ( for a review, see
Strongman, 1996), with researchers tending to focus on dierent aspects of emotions, such as
their physiological substrates, cognitive antecedents, or behavioral outcomes. Recently, however,
a number of scholars have begun to work from a more integrative perspective (e.g., Oatley and
Jenkins, 1996). Specically, they argue that emotions like anger have biological underpinnings
and an evolutionary history; they have played, and continue to play, important roles in adaptive
human functioning. However, these theorists also argue that, evolutionary origins notwithstanding, from the moment of birth people learn from their families and communities how to think
about, talk about, express, and regulate dierent emotions. That is, people acquire sociallyshared, culturally-specic knowledge about emotions, including details of what typically causes
them, what they feel like, how people feeling them are likely to behave, and what their likely
outcomes are.
Such emotion knowledge structures are currently referred to as emotion scripts, in recognition
of the fact that emotion episodes tend to be played out over time between two or more interactants (Shaver et al., 1987; Fehr and Baldwin, 1996; Fitness, 1996, 2000). Of course, such `folk'
theories about how emotions work may or may not bear much resemblance to scientic theories
of emotions, though Oatley and Jenkins (1996) argue that the two share a number of similarities.
Nevertheless, understanding laypeople's emotion scripts is important because, as social cognitive
research has demonstrated, such scripts play a powerful role in shaping and inuencing
laypeople's perceptions, expectations, judgments, and memories, of emotional episodes in their
lives (Smith, 1995; Fitness, 1996).
One frequently used method of investigating emotion scripts is to extract and categorize
prototypical features from laypeople's accounts of recalled or imagined emotion episodes (e.g.,
Shaver et al., 1987; Fitness and Fletcher, 1993). Typically in this procedure, respondents are not
required to recall or imagine an emotional episode in relation to any specic context; thus, some
people may report on an emotion incident at work, whereas others may report on an emotion
incident with a spouse, or child, or friend. These context-free studies are useful in that they
delineate the kinds of general rules that apply to the eliciting conditions and behavioral features
of dierent emotions. However, it is becoming increasingly evident that both context and the
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

ANGER AT WORK

149

nature of the relationship between emotion interactants play important roles in the form and
function of emotion scripts. For example, a number of researchers have found that anger is
elicited by the appraisal that one has been treated unfairly, or unjustly, by another (e.g., Averill,
1982; Smith and Ellsworth, 1985). However, in their study of anger in families, Carpenter and
Halberstadt (1996) found that whereas children were most angered by perceptions of unfair
parental treatment, along with parental inattention and discipline, parents were most angered by
their children's disobedience, `naughty' behaviors and personality deciencies (e.g., slowness,
carelessness). These ndings suggest that perceived injustice may constitute only one, albeit
important, type of anger-elicitor, and that the nature of potentially anger-eliciting events may
vary according to context, and the relationship between the interactants.
Given the nature of the workplace, with its structural power dierences between superiors and
subordinates, and implicit power dierences amongst co-workers, it seems likely that workplace
anger-elicitors might also dier, depending on who is angry with whom. To date, however,
empirical studies of anger-related events at work (e.g., Skarlicki and Folger, 1997) have focused
primarily on the perceptions of subordinates, rather than co-workers or superiors. The rst
specic aim of this study, then, was to investigate and compare the causes and perceived fairness
of anger-eliciting events from the perspectives of superiors, co-workers, and subordinates.

Prototypical behavioral features and consequences of


workplace anger
Along with examining the prototypical causes of anger, a number of emotion researchers have
explored the prototypical feelings and behaviors involved in anger episodes, both generally and
within specic contexts, such as marriage. Overall, the ndings suggest that anger is associated
with tense, energized feelings and a tendency to attack (or at least, want to attack), the anger
target, both verbally and non-verbally (Averill, 1982; Shaver et al., 1987). Similarly, in their study
of marital emotions, Fitness and Fletcher (1993) found that angry spouses tended to confront
their partners and to engage with them in an attempt to communicate their feelings and to give
their partners an opportunity to make amends (see also Fitness, in press). In the workplace,
however, confronting an anger-instigator may not be a viable option if the instigator occupies
higher status, or holds more power, than the oended party. For example, in a recent study of
conict resolution strategies, Drory and Ritov (1997) found that in a low-power opponent
condition, there was a higher preference for dominating, and a lower preference for avoiding,
obliging, and integrating. They concluded that in general, employees tend to adopt a rather
submissive stance towards those who control their future rewards.
These ndings suggest that workers angered by subordinates may be more likely than workers
angered by superiors to immediately confront oenders, raising the question of what lowerpower workers typically do during angry interactions. Moreover, the literature suggests that there
may be dierences in workers' later or ongoing behaviors, depending on their status in the
interaction. For example, research suggests that although workers who feel unjustly treated may
not take direct or confrontational action to remedy the situation, they may take covert retaliatory
action, such as theft or sabotage, in an attempt to `get even' or to balance an apparently
inequitable situation (Skarlicki and Folger, 1997; Bies and Tripp, 1998). Thus, the second,
specic aim of this study was to investigate respondents' reported behaviors during and after
angry episodes, and their beliefs about whether or not the episodes had been successfully
resolved, and to compare these behaviors and beliefs according to respondents' status in the
interaction.
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

150

J. FITNESS

Hate: a power-related emotion


The nal aspect of this study of workplace anger derived from a theoretical interest in a
fascinating, though scientically neglected, interpersonal emotion: hate. Hate is reliably rated by
laypeople as one of the best examples of an emotion (e.g., see Shaver et al., 1987), and hate is
often cited in the popular press as the underlying motive for malicious crimes, including arson,
rape and murder. Despite its relevance to the layperson, however, psychologists have found hate
a dicult construct to dene and there is some disagreement over its nature and constituents. For
example, Smith-Lovin (1995) described hate as a sentiment, or relatively enduring state of simply
`not liking' someone; Shaver et al (1987) classied hate as an anger-related emotion; and Oatley
and Jenkins (1996) proposed that hate is related to disgust, rather than anger, and is distinguished
by prototypical disgust features such as avoidance and being `sickened by'.
In their study of marital emotions, Fitness and Fletcher (1993) found that, compared with
anger-eliciting events, hate-eliciting events were more likely to have been triggered by humiliation
and demeaning treatment, and that reported hate intensity was positively associated with
appraisals of low power, or control, over the situation, relative to the oending spouse. Reported
hate intensity was also positively associated with more destructive long-term marital outcomes.
Overall, these results suggest a potentially important role for the emotion of hate in workplace
anger scripts. For example, they suggest that humiliating anger-eliciting events will elicit more
intense hate than non-humiliating events. In addition, and in line with Roseman's (1984)
speculation that one might `hate the boss' because it was not safe to be angry with someone who
had the power to punish a recalcitrant employee, the ndings suggest that compared with higherpower respondents, lower-power respondents should feel more intense hate for oenders.
Moreover, both the marital study ndings and lay theories about the potentially destructive
outcomes of hate suggest that hate intensity will be negatively associated with the perceived
successful resolution of a work related, anger-eliciting event.
In order to investigate these issues, the third, specic aim of this study was to compare
humiliating and non-humiliating anger-eliciting events with respect to respondents' reported hate
for the oender, and to examine the associations between reported hate intensity, power, and
beliefs about whether or not the anger-eliciting event had been satisfactorily resolved.

Summary of aims
In summary, the rst aim of this study was to investigate the similarities and dierences amongst
superior, co-worker, and subordinate-instigated anger events at work. The second aim was to
examine workers' immediate and long-term behaviors during angry episodes, along with their
beliefs about whether or not the episodes had been satisfactorily resolved, and to compare these
behaviors and beliefs according to workers' status. The third aim was to investigate the associations between reported hate intensity and humiliation, power, and episode resolution.

Method
Participants
One-hundred-and-seventy-ve Australian community residents (79 men and 96 women) participated in the study. Eighty respondents had been angered by a superior, 57 had been angered by a
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

ANGER AT WORK

151

co-worker, and 38 had been angered by a subordinate. The mean age for both male and female
interviewees was 31.4 years (SD 11.3) and ranged from 17 to 60 years. Respondents reporting
on subordinate-instigated anger incidents were a little older than respondents reporting on coworker and superior-instigated incidents, but not signicantly so. Overall, six dierent occupational groups were represented: 35 per cent of respondents were clerical or oce workers, 26 per
cent worked in retail or sales, 15 per cent worked in hospitality (including chefs, waiters and bar
sta), 11 per cent worked in management, 7 per cent were nurses, social workers, or teachers, and
6 per cent were tradespeople, including carpenters and mechanics.

Materials
Anger interview schedule
The design and format of this semi-structured interview schedule was adapted from questionnaires used previously in research on emotions in marriage (Fitness and Fletcher, 1993). First,
respondents were asked to `remember a time when you felt really angry with someone at work'
and to describe what had happened. Respondents were then asked a detailed series of open-ended
questions about the incident, including the status of the interactants; the nature of the event that
had triggered their anger; what they had thought and felt at the time; how they had behaved, both
at the time and later; and whether or not they considered the incident had been successfully
resolved. Along with answering the open-ended questions, respondents also rated the recalled
intensity of their anger in response to the event, and their recalled hate for the oender, on 6 point
Likert scales (ranging from `not at all' to `extremely'). They also rated on 6 point scales their
remembered appraisals of the fairness of the anger-eliciting event, and of their perceived power in
the situation, relative to the oender.

Procedure
Two hundred students in a third year Organizational Psychology class at Macquarie University,
Sydney, Australia, interviewed a currently employed man or woman about an anger-eliciting
incident experienced at work. The overall purpose of the study was explained to students in
advance, and the importance of following the interview schedule and obtaining accurate, detailed
data was stressed. Students were required to obtain signed consent forms from their interviewees,
and to reassure them that their responses were condential and anonymous. Students were also
told that although they were required to conduct the interview and bring the completed questionnaire to class for a discussion of workplace anger, they were not required to hand in their data
for research purposes unless they chose to do so. Accordingly, completed interview schedules
were submitted anonymously and voluntarily. Students were permitted to tape-record the
interviews, but were required to faithfully transcribe the data for class discussion and submission.
Overall, the length and quality of submitted interview data suggested that students took their
interviewing assignment seriously, an impression conrmed by the students' enthusiastic
feedback about the whole exercise.
Of the 200 students who collected interview data, 21 elected not to submit their interview data;
and of the 179 data sheets submitted, 12 were either incomplete or illegible, while 9 described an
incident with a customer and were not used in the current study. This left a sample of 158
interviews for coding and analysis. Because the number of subordinate-related anger incidents
was initially rather low (n 21), an appeal was made to a large class of third year psychology
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

152

J. FITNESS

students to recruit potential interviewees who had experienced an anger episode with a subordinate at work from amongst their families and friends. Students who found willing interviewees were given an interview schedule and careful instructions for collecting the data. Their
eorts brought the number of subordinate-related anger incidents to 38.

Coding
Coding the interview data was carried out in two stages. In the rst stage, free responses for every
question category (i.e., types of anger-eliciting oences; anger-related behaviors) were transcribed in full by a research assistant. These responses were then sorted by the author into
categories, according to their thematic similarity. For example, in relation to the `type of angereliciting event' question, descriptions of having been unjustly accused of stealing, having been
allocated an unfair workload, or having been denied leave despite putting in extra hours, were
categorized as examples of `unjust treatment', whereas descriptions involving having being
teased, criticized, or ridiculed in front of others were categorized as examples of `public humiliation'. Similarly, with respect to `anger-related behaviors', shouting, protesting, or arguing were
categorized as examples of `immediate confrontation', whereas maintaining silence, walking
away, or ignoring the oender were categorized as examples of `immediate withdrawal'. The aim
was to devise a concise but informative coding system that would capture the breadth of
responses as economically as possible without being either under or over inclusive.
In the second stage of the coding, two research assistants went through the 175 interview
protocols, coding responses to each of the question categories (e.g., type of anger-eliciting event,
anger-related behaviors) according to the scheme developed by the author. The raters also coded
eliciting events according to whether or not they had involved humiliation (yes or no). Inter-rater
reliability across categories was calculated using Cohen's kappa, a statistic that adjusts
percentage of agreement rates to correct for chance agreements (Howell, 1992). Overall,
reliabilities were high, with the lowest Cohen's kappa being 0.88 for type of triggering event and
whether or not the oence could be described as humiliating. Disagreements were resolved in
discussion with the author.

Results and discussion


Anger-eliciting events
Table 1 depicts the numbers and percentage frequencies of anger-eliciting events, both overall
and broken down according to the status of the person who angered the respondent.
In line with more general studies on prototypical anger-eliciting events, the largest overall event
category (44 per cent) involved being directly and unjustly treated by another; e.g., being falsely
accused of lying, stealing, or poor performance; being unjustly criticized; having a reasonable
request denied, or being given an onerous workload. Typically, respondents reporting events in
this category emphasized the unfairness and injustice of the oence. For example, one young
retail assistant was told to `make up' a considerable amount of money missing from the till at the
end of a shift or be dismissed, despite the fact that several other assistants had also used the till;
she was adamant she had neither taken nor made a mistake with the money. She believed she had
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

ANGER AT WORK

153

Table 1. Numbers and percentage frequencies (in parentheses) of anger-eliciting events overall and
according to oender status

Anger-eliciting event:
Unjustly treated
Immoral behavior
Job incompetence
Disrespect
Public humiliation

Overall
(n 175)

Superior
(n 80)

Co-worker
(n 57)

Subordinate
(n 38)

77 (44%)
40 (23%)
26 (15%)
20 (11%)
12 (7%)

55 (69%)
7 (8%)
9 (11%)
8 (10%)
1 (2%)

16 (28%)
20 (35%)
4 (7%)
8 (14%)
9 (16%)

6 (16%)
12 (31%)
14 (38%)
4 (10%)
2 (5%)

w2 (8, n 175) 59.82, p 5 0.000

been targeted because she was a recently appointed casual worker, and was incensed at the
perceived unfairness of her supervisor's behavior.
Perceived unfairness was not, however, the only type of anger-eliciting event reported by
workers. The second largest event category (23 per cent) involved cases where the respondent
believed the target of his or her anger had behaved in a morally reprehensible way, such as being
lazy or dishonest; telling lies, stealing, cheating (e.g., on expenses), `bludging' or taking advantage of others, or conducting a sexual relationship with a supervisor or subordinate. Sometimes
these behaviors were reportedly anger-provoking because of their negative consequences for the
respondent, e.g., when a co-worker's apparent laziness meant the respondent had to work twice
as hard to cope with customers (a common complaint from waiters and bar-sta). However,
there were several instances where behaviors that did not appear to have had a directly negative
impact on the respondent evoked anger nonetheless because they violated important workplace
rules. For example, a social worker described overhearing a colleague being rude to a client and
feeling angry because `it's wrong to take our frustrations out on clients'; another learned that a
work colleague was having an aair with the boss and felt outraged because `supervisors and
subordinates ought to keep a professional distance'; another was angered by a co-worker who
was constantly late because `he shouldn't get the same pay if he doesn't work the same hours'.
The third largest event category (15 per cent) involved job incompetence; events that hampered
the respondent's ability to get the job done, or that cost the company money or goodwill;
mistakes that may have been unintentional, but that still evoked anger in the respondent
(especially when the oender failed to realize the seriousness of the situation). For example,
several respondents described incidents where their colleagues' slowness or apparent stupidity
had resulted in long lines of angry customers waiting to be served, while others described
expensive computer crashes and various human-caused mechanical mishaps.
The fourth largest event category (11 per cent) involved perceptions that an oender was acting
in an arrogant or disrespectful manner. For example, in a long, passionate account an intensive
care nurse described her rage at having her instructions to junior nursing sta overruled by the
attending Registrar who told her over the telephone in an amused tone to be a `good little girl'
and stop `making a fuss'. Arrogance or rudeness did not have to directly impact the respondent,
however, to be anger-provoking; some respondents reported anger on behalf of others (especially
subordinates) whom they observed to have been treated in a dismissive, demeaning manner by
the oender.
The smallest event category (7 per cent) involved incidents where the respondent had been
directly and publicly humiliated by the oender. For example, in a twist on the more stereotypical
sexual harassment scenario, a young male clerk assigned to a new department described his
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

154

J. FITNESS

Table 2. Fairness and power appraisals, and anger and hate intensity means, according to oender status
Superior

Co-worker

Subordinate

F(2,174)

Appraisal:
Fairness of event
Power in situation

1.56a
2.61a

2.29b
3.30b

4.03c
3.71b

40.96{
7.85*

Emotion:
Anger intensity
Hate intensity

4.82a
4.14a

4.60a
4.21a

4.24b
2.58b

4.70*
14.00{

Note. The F ratio is derived from a one-way analysis of variance across the three status conditions. Means with dierent
superscripts were signicantly dierent according to NeumanKeuls post hoc comparisons. Because of the number of
tests conducted on the data, a conservative signicance level was set at p 5 0.01.
* p 5 0.01; { p 5 0.001.

humiliation when his female supervisor publicly joked about his physical appearance and
speculated on his sexual prowess to other female oce sta. Other examples included public
criticism, especially in front of customers or subordinates.
A chi square analysis comparing the frequencies of the dierent types of anger-eliciting events
across the three status conditions obtained a signicant result, w2 (8, n 175) 59.82,
p 5 0.000. In line with the proposal that types of anger-eliciting oences might dier depending
on the nature of the relationship between interactants, 69 per cent of superior-instigated incidents
involved directly unjust treatment, compared with only 28 per cent and 16 per cent of co-worker
and subordinate-instigated oences respectively; 65 per cent of co-worker instigated incidents
involved morally reprehensible behaviors, such as laziness or dishonesty, public humiliation, and
disrespect, and 69 per cent of subordinate-instigated oences involved job incompetence or
morally reprehensible behaviors.
Respondents' explicit appraisals of the fairness of anger-eliciting events closely mirrored these
results, with respondents angered by superiors appraising oences as signicantly less fair than
respondents angered by co-workers or subordinates, and respondents angered by co-workers
appraising oences as signicantly less fair than respondents angered by subordinates,
F(2,174) 40.96, p 5 0.001 (see Table 2). These results were also somewhat in line with respondents' ratings of recalled anger intensity, with respondents angered by superiors and co-workers
reporting marginally more intense anger than respondents angered by subordinates,
F(2,174) 4.70, p 0.01. Finally, and as expected, respondents angered by subordinates
appraised their power in the anger-eliciting situation as signicantly higher than respondents
angered by superiors or co-workers, F(2,174) 7.85, p 5 0.01.

Prototypical behavioral features and consequences of


workplace anger
The second, specic aim of this study was to investigate superiors', co-workers' and subordinates'
behaviors during and after angry episodes, and to examine their beliefs about the successful
resolution of the anger episode (see Table 3). In line with theoretical speculations that subordinates would be less likely than superiors to confront the targets of their anger, only 45 per
cent of respondents angered by superiors immediately confronted them during the course of
anger-eliciting events, compared with 58 per cent of respondents angered by co-workers, and
71 per cent of respondents angered by subordinates, w2 (2, n 175) 7.38, p 0.02.
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

ANGER AT WORK

155

Table 3. Numbers and percentage frequencies (in parentheses) of anger-eliciting behaviors and
consequences overall and according to oender status
Overall
(n 175)

Superior
(n 80)

Co-worker
(n 57)

Subordinate
(n 38)

Immediate behavior
Confront
96 (55%)
Withdraw
79 (45%)
w2 (2, n 175) 7.38, p 0.02

36 (45%)
44 (55%)

33 (58%)
24 (42%)

27 (71%)
11 (29%)

Later behavior
Withdraw
63 (36%)
Revenge
46 (26%)
Punishment
28 (16%)
Constructive
24 (14%)
Quit
14 (8%)
w2 (8, n 175) 36.85, p 5 0.000

31 (39%)
24 (30%)
6 (7%)
7 (9%)
12 (15%)

26 (45%)
14 (25%)
10 (17%)
5 (9%)
2 (4%)

6 (15%)
8 (21%)
12 (31%)
12 (33%)

Successful resolution
Yes
75 (43%)
No
100 (57%)
w2 (2, n 175) 18.21, p 5 0.000

27 (34%)
53 (66%)

21 (37%)
36 (63%)

28 (74%)
10 (26%)

Furthermore, the majority of respondents angered by superiors reported that they avoided an
immediate confrontation because they feared the consequences of expressing their feelings to a
more powerful oender. Respondents angered by subordinates, on the other hand, reported
confronting behaviors in order to `sort the situation out' or even, in some cases, to intimidate
subordinates and let them know `who's in charge'.
Behaviors occurring after the anger-eliciting incident was over were coded into ve categories:
constructive behaviors (e.g., accepting an apology, actively negotiating a resolution); emotional
withdrawal (staying cold, ignoring, giving the `silent treatment'); quitting as a direct result of the
oence, so-called `legitimate' punishment (including appealing to higher authorities, arranging to
have an oender dismissed or demoted, or getting the Union involved); and revenge, or
`illegitimate' punishment (including arranging for oenders to be assigned undesirable jobs;
spreading gossip or lies; and sabotage, such as hiding important documents, going slowly on
urgent jobs, and adding salt, or chilli sauce to mealsreported by waiters as revenge on chefs).
A chi square analysis revealed an overall signicant dierence in behavior type frequencies,
depending on oender status, w2 (8, n 175) 36.85, p 5 0.000. With respect to the individual
behavior categories, a number of signicant dierences were obtained. For example, only 15 per
cent of respondents angered by subordinates later withdrew and `went cold', compared with 39
per cent and 45 per cent of respondents angered by superiors and co-workers respectively, w2 (2,
n 175) 11.11, p 5 0.003; on the other hand, 33 per cent of respondents angered by
subordinates reported they took constructive action to resolve the conict, compared with only 9
per cent of respondents angered by co-workers, and 9 per cent angered by superiors, w2(2,
n 175) 15.74, p 5 0.000. There was also a signicant status dierence in punishment
behaviors, with 31 per cent of respondents angered by subordinates reporting that they took
legitimate action to demonstrate to oenders the error of their ways, compared with 17 per cent of
respondents angered by co-workers and 7 per cent angered by superiors, w2 (2, n 175) 11.26,
p 5 0.004. There were no signicant dierences, however, in the reported frequencies of revenge
behaviors across the status categories. Finally, with respect to quitting as a direct result of the
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

156

J. FITNESS

Table 4. Numbers and percentage frequencies (in parentheses) of humiliating anger-eliciting events overall
and according to oender status

Humiliating:
Yes
No

Overall
(n 175)

Superior
(n 80)

Co-worker
(n 57)

Subordinate
(n 38)

67 (38%)
108 (62%)

38 (46%)
43 (54%)

23 (40%)
34 (60%)

6 (16%)
31 (84%)

w2(2, n 175) 10.42, p 5 0.005

oence, overall numbers were rather low, and restricted to workers angered by superiors (15 per
cent) or co-workers (4 per cent).
It should be noted that one other behavior, telling others about the anger-eliciting event, was
reported by 83 per cent of respondents, regardless of status. The majority of these respondents
reported telling others in order to obtain emotional support and to bolster their versions of
events. As one respondent explained, `I knew my version of events was right, but I needed other
people to agree with me, to tell me what a creep my boss was, so I wouldn't start thinking I'd
imagined things'. These unexpected ndings support research by Rime (1995) on the importance
of the social sharing of emotion and suggest that workers' friends, families, and colleagues play a
potentially signicant role in the ongoing progress (and eventual successful or unsuccessful
resolution) of workplace anger episodes.
With respect to whether respondents considered anger episodes had been successfully resolved,
74 per cent of respondents angered by subordinates agreed that they had, compared with only
34 per cent of respondents angered by superiors and 37 per cent of respondents angered by
co-workers, w2(2, n 175) 18.21, p 5 0.000. Several respondents who had been angered by
superiors spoke at length of their ongoing distrust of, and hatred for, their supervisors or
co-workers; however, the majority of respondents who had been angered by subordinates
cheerfully reported that their relationships were `back to normal'.

Humiliation, hate, power, and successful resolution of


the anger-eliciting incident
To be categorized as humiliating, respondents' accounts had to specically mention features such
as feeling ashamed, humiliated, or embarrassed; feeling put down, diminished or ridiculed;
wanting to hide; feeling discomfort at `everyone knowing' about the incident (see Table 4).
Overall, 38 per cent of oences were classied as humiliating, and a signicant status  humiliation eect was obtained, with superior and co-worker oences more likely than subordinate
oences to be categorized as humiliating, w2 (2, n 175) 10.42, p 5 0.005.
In line with marital studies that have found a positive association between humiliation and
hate, it was suggested that humiliating workplace oences would elicit more hate for the oender
than non-humiliating oences. This expectation was conrmed, with respondents reporting more
hate in response to humiliating (m 4.39) than non-humiliating (m 3.48) oences,
F(1,174) 11.88, p 5 0.000. No eect was obtained for reported anger intensity, however,
with non-humiliating oences eliciting as nearly as much anger (m 4.54) as humiliating
oences (m 4.80).
Along with its potential relationship with humiliation, it was also expected that reported hate
intensity would be negatively associated with perceived self-power, relative to the oender, and
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

ANGER AT WORK

157

Table 5. Correlations amongst humiliation, anger, hate, perceived self-power, and successful resolution of
the anger incident

Humiliation
Anger
Hate
Perceived self-power
Successfully resolved

0.17*

0.26{
0.27{

0.27{
0.07
0.20*

0.11
0.15
0.33{
0.14

Note: Alpha levels set at 0.01.


* p 5 0.01; { p 5 0.001.

negatively associated with a perceived successful resolution to the angry interaction. These
expectations were also conrmed. As depicted in Table 2, respondents interacting with superiors
and co-workers reported signicantly more intense hate for the oender than respondents
interacting with subordinates, F(2,174) 14.00, p 5 0.000. Moreover, moderately signicant
negative correlations were obtained between reported hate intensity and perceived self-power in
the situation, r 0.20, p 5 0.01, and perceived successful resolution, r 0.33, p 5 0.001 (see
Table 5). Again, no such associations were obtained between anger intensity, self-power, and the
perceived successful resolution of anger eliciting events.
These results conrm the ndings of other studies (e.g., Fitness and Fletcher, 1993) that hate
may be experienced by the layperson within the context of an anger incident in response to
certain aspects of that incident; in the current context, humiliation, and/or perceiving that one
has less power in the situation than the oender. Unfortunately, humiliation, including public
ridicule, bullying, and emotional abuse, was a relatively common feature of workers' accounts,
with sometimes devastating eects. For example, one respondent reported she needed psychiatric
assistance to `get over' the trauma of being publicly informed that she was both ugly and stupid,
and another talked about his suicidal feelings as a consequence of his humiliating ordeal with a
sadistic superior. It is also interesting to note that, although the sample size was very small, of the
14 respondents who quit as a result of the anger-eliciting event, 13 of these events were classied
as humiliating. Respondents' accounts suggested that many superiors appeared not to notice the
impact of their humiliating and bullying behaviors, or even to regard them benignly as `teasing',
again perhaps because subordinates are more likely to withdraw than to protest about their
treatment. These ndings and the results of other studies on workplace bullying (e.g., Randell,
1997) demonstrate the need for managerial awareness of, and sensitivity to, this important issue.
Finally, the results suggest that, in line with the marital ndings, feeling hate for an oender is
a bad omen for the future of the relationship between the two parties to an angry interaction.
Clearly, more research is needed to esh out the causes, features, and consequences of workplace
hatred in order to better understand and manage this potentially destructive emotion.

General discussion
The aim of this study was to use an emotion-script approach to investigate the causes, features,
and consequences of anger at work and to examine the similarities and dierences amongst the
reported anger experiences of superiors, co-workers, and subordinates. With respect to the causal
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

158

J. FITNESS

features of prototypically anger-eliciting workplace events, the results of the study suggest a
common, underlying theme; specically, the perception that one's needs, wants, expectations, or
even assumptions about how other people `should' behave, have been violated or thwarted in
some way (see also Lazarus, 1991, and Shaver et al., 1987). However, the results also suggest that
context plays an important role in the form and function of anger scripts. For example, the
oences that most angered superiors and co-workers (incompetence, disrespect, and morally
reprehensible behaviors such as laziness, or untruthfulness) are remarkably similar to those
identied by Carpenter and Halberstadt (1996) in their study of what makes parents angry with
children (disobedience, naughty behaviors, and personality deciencies); in addition, the oence
that most angered subordinates was the same as the oence that most angered children (unfair
treatment). Respondents' accounts also made it clear that co-workers closely monitor one
another's behaviors and again, much like family siblings, are vigilant for signs that co-workers
are obtaining undeserved benets through rule violations, such as cheating, stealing, or currying
favor with superiors. Certainly, there are elements of unfairness involved in all these behaviors;
however, in many cases, the focus of respondents' anger was moral outrage, rather than perceived
injustice.
These results suggest a structural and functional similarity between family and workplace
anger scripts, particularly given the nding that respondents interacting with subordinates felt
less intense anger than respondents interacting with co-workers or superiors. Other studies have
found that anger is often less intense toward children than toward adults (e.g., Savitsky et al.,
1975), presumably because children are considered less responsible for their behavior. This
suggests that some superiors may regard lower status workers in much the same way as parents
regard children with respect to their abilities and maturity levels, although it is also important to
note that higher power people have the capacity to take more decisive action to resolve angereliciting events (to their own satisfaction, at least), which may also account for their more
moderate feelings of anger. It is interesting to note, however, that a number of respondents in the
current study explained how they deliberately feigned anger at work in order to intimidate
subordinates; again, very much as parents might do with recalcitrant children. These ndings are
in line with Ostell's (1996) review of the tactical use of anger to intimidate others at work, and
with a series of studies reviewed by Clark et al. (1996) on the strategic presentation of anger. In
particular, they found that angry people are perceived by others as dominant and that, in an
experimental situation at least, people will strategically present anger for instrumental purposes.
They also cited a pertinent comment by Parrott (1993) who noted in regard to work supervisors
that `if one's workers have become complacent, a display of irritability may make them more
anxious about their situations and induce more motivation and concentration ( p. 290)'. Clearly,
there is scope here for more research on subordinates' appraisals, emotions, and responses to
such blatant abuses of power. However, it would also be interesting to examine the extent to
which workers strategically present anger in order to threaten or intimidate workplace colleagues
and supervisors.

Prototypical features of workplace anger scripts:


the role of power
Overall, the results of this study suggest that the power relationship between participants in an
angry workplace interaction plays a critically important role in the prototypical behaviors and
outcomes associated with the progress of the anger script. Two, distinct kinds of anger script
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

ANGER AT WORK

159

emerged, depending on the nature of this power relationship. First, a summary of the anger script
for higher-power workers: They are likely to become moderately angry over the perceived
incompetence or reprehensible behavior of lower-power workers; however, the oences that make
them angry rarely involve personal humiliation and they feel low levels of hate for oenders. They
do not appraise the events as unfair, and their immediate reactions are to confront oenders
rather than to withdraw and say nothing. After the anger episodes, superiors either take constructive action to resolve the situation, or they punish subordinates in order to teach them a
lesson; most are satised that the episodes have been successfully resolved.
Now, a summary of the anger script for lower power workers: They are likely to become angry
over unjust treatment by higher power workers, and to appraise the events as highly unfair. They
experience moderate to high levels of hate for oenders, especially if the oences have involved
humiliation, and their immediate reactions involve withdrawal. After the anger episodes, lower
power workers either maintain their distance from oenders, or they take revenge in the interests
of `getting even'. Few are satised that the episodes have been successfully resolved.
The emergence of these two, distinctive kinds of scripts conrm that power is an important
factor in laypeople's experiences, expectations, and memories, of workplace anger episodes,
inuencing emotions, behaviors, and outcomes in various (and not always predictable) ways. Of
course, it is not possible to tell from the current study whether the same results would have
emerged had the perspectives of both parties to the same angry interaction been obtained.
Nevertheless, the ndings suggest that the impact of angry workplace interactions may be
potentially greater or longer lasting for subordinates than for superiors. In part, this may be
because, as the results of the current study makes clear, subordinates are less likely than superiors
to directly confront oenders and/or try to constructively resolve the situation; thus, superiors
may be unaware that subordinates are experiencing ongoing feelings of distress and distrust in the
aftermath of an angry workplace interaction. There is also the fact that, as noted by Fiske (1993),
higher power people generally pay less attention to lower power people, and so may miss cues
about subordinates' ongoing thoughts and feelings.
Overall, these ndings underscore the importance of considering the nature of the power
relationship between interactants when attempting to understand and resolve anger-eliciting
events in the workplace. Moreover, the fact that no signicant gender dierences were found in
the current study suggests that in the workplace, power may be a more important factor than
gender when it comes to managing workplace conict and the experience and expression of
anger.

Limitations of the current study and directions


for future research
There are a number of limitations to the current study, which was restricted to self-report and
which represents only an exploratory and preliminary sketch of a potentially wide-ranging area of
research. In particular, the data in the current study describe only one partner's perspective on
the anger-eliciting incident; researchers would obtain richer results if they were to gather data on
the same anger-eliciting event from both perspectives. Future research could also more
systematically examine anger scripts by having larger samples of workers at dierent status levels
generate recalled and hypothetical anger-eliciting scenarios, and by gathering more nelydetailed data about their constituent emotions (e.g., hurt, anxiety, depression, etc.). As a reviewer
of this paper noted, emotions come in blends and sequences, and each has its distinctive
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

160

J. FITNESS

motivations and consequences. No doubt, role-playing, analogue, and laboratory-based video


studies would be particularly useful for examining such ongoing emotion sequences and their
consequences.
It would also be useful to conduct more detailed studies of the nature and function of
workplace `rules' in order to better understand and predict the impact of dierent kinds of rule
violations on workers (e.g., see Argyle and Henderson, 1985, and Morrison and Robinson,
1997), especially given the propensity of workers to describe co-worker and subordinateinstigated oences with respect to their moral reprehensibility. In addition, respondents' accounts
suggested that ongoing conicts with co-workers can be the source of much misery, in part
because it is seen as inappropriate to complain about, or blow the whistle on, bullying, dishonest,
or otherwise obnoxious workplace colleagues. To this extent, workers in theoretically equalstatus relationships may perceive themselves to have as little power to escape or eectively deal
with the situation as workers in legitimately subordinate relationships. Given how little is known
about the features and outcomes of co-workers' emotional interactions, this is clearly an
interesting and important area for future research.
Researchers might also fruitfully consider the eects of individual dierences on the various
features of workplace anger scripts; e.g., emotion-related dispositional variables, such as
`emotional intelligence' (Mayer and Salovey, 1997), negative aectivity (Clark and Watson, 1991)
and even so-called chronic anger, or the propensity to interpret the world and others' behaviors
as deliberately frustrating and maliciously intentional (Isen and Diamond, 1987; Weiss and
Cropanzano, 1996). Such individual dierence factors are likely to play an important role in
shaping people's perceptions, experience, and memories of anger-eliciting interactions at work,
along with inuencing their behaviors and the long-term consequences of such interactions.
In conclusion, this study has demonstrated that laypeople have richly elaborated knowledge
structures about the causes and prototypical features of anger at work, depending on the nature
of the relationship among interactants. No doubt, such anger scripts play an important role in
people's expectations of how anger episodes will unfold at work, and in their interpretations,
justications, and memories of such episodes. Although this work is just beginning, the results of
this study have demonstrated the utility of the emotion script approach in enriching our
understanding of how people think about and make sense of their emotions in social contexts like
the workplace. Further research in this area will equip us to better understand emotional
interactions in that context.

References
Allcorn S. 1994. Anger in the Workplace: Understanding the Causes of Aggression and Violence. Quorum
Books: Westport.
Argyle M, Henderson M. 1985. The rules of relationships. In Understanding Personal Relationships: An
Interdisciplinary Approach, Duck S, Perlman D (eds). Sage: Beverly Hills; 6384.
Ashforth B, Humphrey R. 1995. Emotion in the workplace: a reappraisal. Human Relations 48: 97125.
Averill J. 1982. Anger and Aggression: An Essay on Emotion. Springer-Verlag: New York.
Bensimon H. 1997. What to do about anger in the workplace. Training and Development 51: 2832.
Bies R, Tripp T. 1998. Revenge in organizations: The good, the bad, and the ugly. In Dysfunctional
Behavior in Organizations: Non-violent Dysfunctional Behavior, Grin R, O'Leary-Kelly A, Collins J
(eds). JAI Press: Stanford, Connecticut; 4967.
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

ANGER AT WORK

161

Carpenter S, Halberstadt A. 1996. What makes people angry? Laypersons' and psychologists'
categorizations of anger in the family. Cognition and Emotion 10: 627656.
Chen P, Spector P. 1992. Relationships of work stressors with aggression, withdrawal, theft and substance
use: an exploratory study. Journal of Occupational and Organizational Psychology 65: 177184.
Clark LA, Watson D. 1991. Tripartite model of anxiety and depression: psychometric evidence and
taxonomic implications. Journal of Abnormal Psychology 100: 316336.
Clark M, Pataki S, Carver V. 1996. Some thoughts and ndings on self-presentation of emotions in
relationships. In Knowledge Structures in Close Relationships: A Social Psychological Approach, Fletcher
GJO, Fitness J (eds). Lawrence Erlbaum: Mahwah, New Jersey; 247274.
Drory A, Ritov I. 1997. Eects of work experience and opponent's power on conict management styles.
The International Journal of Conict Management 8: 148161.
Fehr B, Baldwin M. 1996. Prototype and script analyses of laypeople's knowledge of anger. In Knowledge
Structures in Close Relationships: A Social Psychological Approach, Fletcher GJO, Fitness J (eds).
Lawrence Erlbaum: Mahwah, New Jersey; 219246.
Fiske S. 1993. Controlling other people: the impact of power on stereotyping. American Psychologist 48:
621628.
Fitness J. 1996. Emotion knowledge structures in close relationships. In Knowledge Structures in Close
Relationships: A Social Psychological Approach, Fletcher GJO, Fitness J (eds). Lawrence Erlbaum:
Mahwah, New Jersey; 195218.
Fitness J. 2000. Betrayal, rejection, revenge and forgiveness: an interpersonal script approach. In
Interpersonal Rejection, Leary M (ed.). Oxford University Press: New York; in press.
Fitness J, Fletcher GJO. 1993. Love, hate, anger, and jealousy in close relationships: a prototype and
cognitive appraisal approach. Journal of Personality and Social Psychology 65: 942958.
Folger R, Baron R. 1996. Violence and hostility at work: a model of reactions to perceived injustice. In
Violence On The Job, VandenBos G, Bulateo E (eds). American Psychological Association: Washington;
5185.
Glomb T, Hulin C. 1997. Anger and gender eects in observed supervisor-subordinate dyadic interactions.
Organizational Behavior and Human Decision Processes 72: 281307.
Hochschild A. 1983. The Managed Heart. University of California Press: London.
Howell D. 1992. Statistical Methods for Psychology, 3rd edn. Duxbury Press: California.
Isen A, Diamond GA. Aect and automaticity. In Unintended Thought, Uleman J, Bargh J (eds). Guilford
Press: New York; 124154.
Lazarus R. 1991. Emotion and Adaptation. Oxford University Press: New York.
Mayer J, Salovey P. 1997. What is emotional intelligence? In Emotional Development and Emotional
Intelligence, Salovey P, Sluyter D (eds). Basic Books: New York; 331.
Morrison EW, Robinson S. 1997. When employees feel betrayed: a model of how psychological contract
violation develops. Academy of Management Review 22: 226256.
Oatley K, Jenkins J. 1996. Understanding Emotions. Blackwell Publishers: Cambridge.
Ostell A. 1996. Managing dysfunctional emotions in organizations. Journal of Management Studies 33:
525557.
Parrott WG. 1993. Beyond hedonism: motives for inhibiting good moods and for maintaining bad moods.
In Handbook of Mental Control, Wegner D, Pennebaker JW (eds). Prentice Hall: Englewood Clis, NJ;
278305.
Rafaeli A, Sutton RI. 1989. The expression of emotion in organizational life. Research in Organizational
Behavior 11: 142.
Rafaeli A, Sutton RI. 1991. Emotional contrast strategies as means of social inuence: lessons from
criminal interrogations and bill collectors. Academy of Management Journal 34: 749775.
Randell P. 1997. Adult Bullying: Perpetrators and Victims. Routledge: New York.
Rime B. 1995. The social sharing of emotion as a source for the social knowledge of emotion. In Everyday
Conceptions of Emotions, Russell J et al. (eds). Academic Press: Kluwer; 475489.
Roseman I. 1984. Cognitive determinants of emotion: a structural theory. In Review of Personality and
Social Psychology: Emotions, Relationships, and Health, Vol. 5, Shaver P (ed.). Sage: Beverly Hills, CA;
1116.
Savitsky JC, Czyzewski D, Dubord D, Kaminsky S. 1975. Age and emotion of an oender as determinants
of adult punitive reactions. Journal of Personality 44: 311320.
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: 10611086.
Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

162

J. FITNESS

Skarlicki D, Folger R. 1997. Retaliation in the workplace: the roles of distributive, procedural, and
interactional justice. Journal of Applied Psychology 82: 434443.
Smith C, Ellsworth, P. 1985. Patterns of cognitive appraisal in emotion. Journal of Personality and Social
Psychology 48: 813838.
Smith K. 1995. Social psychological perspectives on laypersons' theories of emotion. In Everyday
Conceptions of Emotions, Russell J et al. (eds). Academic Press: Kluwer; 399414.
Smith-Lovin L. 1995. The sociology of aect and emotion. In Sociological Perspectives on Social
Psychology, Cook K, Fine G, House J (eds). Allyn & Bacon: Boston, MA; 118148.
Strongman KT. 1996. The Psychology of Emotion: Theories of Emotion in Perspective, 4th edn. Wiley:
Chichester.
Weiss H, Cropanzano R. 1996. Aective events theory: a theoretical discussion of the structure, causes and
consequences of aective experiences at work. Research in Organizational Behavior 18: 174.

Copyright # 2000 John Wiley & Sons, Ltd.

J. Organiz. Behav. 21, 147162 (2000)

Você também pode gostar