Attitudes & behaviors article reviews

Journal of Vocational Behavior 59, 291–309 (2001)
doi:10.1006/jvbe.2001.1803, available online at http://www.idealibrary.com on

Counterproductive Work Behavior (CWB) in Response to Job
Stressors and Organizational Justice: Some Mediator

and Moderator Tests for Autonomy and Emotions

Suzy Fox

Institute of Human Resources and Industrial Relations, Loyola University Chicago

and

Paul E. Spector and Don Miles

University of South Florida

Relations among job stressors, perceived justice, negative emotional reactions to work,
counterproductive work behavior (CWB), autonomy, and affective traits were investi-
gated. Participants representing a wide variety of jobs across many organizations were
surveyed both inside and outside a university setting. Results were consistent with a the-
oretical job stress framework in which organizational constraints, interpersonal confict,
and perceived injustice are job stressors, CWB is a behavioral strain response, and nega-
tive emotion mediates the stressor–strain relationship. Only very weak support was found
for the moderating role of affective disposition (trait anger and trait anxiety), and no
support was found for the expected moderating role of autonomy in the stressor–CWB
relationship. °C 2001 Academic Press

Recent years have seen an explosion of interest among organizational research-
ers in counterproductive work behaviors (CWB), such as aggression, interpersonal
confict, sabotage, and theft. Although most of this work has been directed toward
validating integrity tests with the objective of devising ways to identify counter-
productive employees so that companies can avoid hiring them, two streams of
research have focused on ascertaining the causes of these behaviors. Spector and
colleagues (Chen & Spector, 1992; Fox & Spector, 1999; Spector 1975, 1978;
Storms & Spector, 1987) have portrayed CWB as an emotion-based response to
stressful organizational conditions. Greenberg and colleagues (e.g., Greenberg,
1990) and Skarlicki, Folger and colleagues (Skarlicki & Folger, 1997; Skarlicki,

We thank Linda Mezydlo Subich for her most helpful substantive and editorial suggestions through-
out the process of revising this article.

Address correspondence and reprint requests to Suzy Fox, Institute of Human Resources and In-
dustrial Relations, Loyola University Chicago, 820 N. Michigan Ave., Chicago, IL 60611. E-mail:
[email protected]

291
0001-8791/01 $35.00

Copyright °C 2001 by Academic Press
All rights of reproduction in any form reserved.

292 FOX, SPECTOR, AND MILES

Folger, & Tesluk, 1999) have taken an organizational justice perspective, viewing
CWB as a cognition-based response to experienced injustice. These two perspec-
tives are not incompatible, and in fact Spector (1978) noted links with the equity
(justice) concept, and Greenberg (1990) noted links with frustration theory. The
current study integrates both perspectives, assessing relations among job stressors,
perceptions of injustice, and CWB within the framework of job stress theory. It is
further proposed that, consistent with this theoretical framework, emotional reac-
tions to job stressors and injustice perceptions, affective disposition, and perceived
control over work are key links in these relations.

The Nature of Counterproductive Work Behavior (CWB)

CWB is behavior that is intended to have a detrimental effect on organizations
and their members. It can include overt acts such as aggression and theft or more
passive acts, such as purposely failing to follow instructions or doing work incor-
rectly. CWB has been conceptualized in a number of ways, including organizational
aggression (Neuman & Baron, 1998; Fox & Spector, 1999), antisocial behavior
(Giacalone & Greenberg, 1997), delinquency (Hogan & Hogan, 1989), deviance
(Hollinger, 1986; Robinson & Bennett, 1995), retaliation (Skarlicki & Folger,
1997), revenge (Bies, Tripp, & Kramer, 1997), and mobbing/bullying (Knorz &
Zapf, 1996). The common theme is that these behaviors are harmful to the organi-
zation by directly affecting its functioning or property, or by hurting employees in
a way that will reduce their effectiveness. A number of researchers (Fox & Spector,
1999; Robinson & Bennett, 1995) have found evidence that perceptions of CWBs
and/or relations of CWBs to individual and organizational variables allow us to
distinguish two categories of behaviors: those targeting the organization and those
targeting other persons in the organization.

A Job Stress/Emotion/CWB Model

Integrating both the CWB and job stress literatures, Spector (1998) and Spector
and Fox (in press) developed a job stress/emotion/CWB model that suggests these
behaviors are responses to job stressors at work. According to this view, people
monitor and appraise events in the environment (Lazarus, 1991), and certain events
that are seen as threats to well-being are job stressors that induce negative emotional
reactions, such as anger or anxiety (Spector, 1998). Common examples of job
stressors are role confict and ambiguity (Kahn et al., 1964), interpersonal confict
(Spector, Dwyer, & Jex, 1998), and situational constraints (Peters & O’Connor,
1990). Strain is an outcome of the job stress process that can be psychological (e.g.,
job dissatisfaction or turnover intention), physical (e.g., somatic symptoms such
as headache, physiological changes such as increased blood pressure, and long-
term pathology), or behavioral (e.g., smoking or withdrawal from work). CWB is
a manifestation of behavioral strain.

Emotions play a central role in the job stress process. Because emotions represent
the immediate response to situations that are perceived as stressful (Lazarus, 1991;
Lovallo, 1997; Payne, 1999), and because they energize and motivate subsequent

293 COUNTERPRODUCTIVE WORK BEHAVIOR

behavior and physiological change (Cartwright & Cooper, 1997; Spector, 1998),
we can defne their role as a mediator between job stressors and strains, therefore
serving a mediator role for CWB. In fact, Fox and Spector (1999) found evidence of
this mediating role for emotions in the relation between organizational constraints
(a stressor) and CWB. The Spector and Fox (in press) model suggests a fow from
environment to perception, to negative emotion, to CWB, but there are several
mitigating factors. Individuals vary in their propensity to appraise situations as
stressful both among one another and across time, so perceptions of control and
personality are two important factors to consider.

Much has been written concerning the importance of control as both an additive
and moderator factor in the job stress process (e.g., Frankenhaeuser & Johannson,
1986; Karasek, 1979; Spector, 1986; Thompson, 1981). The core common to this
literature is that it is not the objective events or stimuli in the environment that
cause adverse outcomes (strains), but rather the perceptions of the individual of not
having the control (which Karasek further specifes to include decision latitude and
task autonomy) to cope with these threats or demands. In the CWB arena, Allen and
Greenberger (1980) suggested that perceptions of control are an important determi-
nant of counterproductive work behavior, as nonconstructive behavioral responses
are more likely when a person perceives low control of the situation. Support for
this idea in the work context was provided by Storms and Spector (1987). Yet, it
is not control in general that is important, but rather control over the job stressor
itself. Having control over specifc tasks (i.e., task autonomy) may be helpful in
reducing the stressfulness of task-related stressors such as workload, but it will
not affect the stressfulness of unrelated stressors such as interpersonal confict.

Personality traits are also relevant factors in CWB, and an entire integrity test
industry has grown around the idea that personality tests can predict these behav-
iors (Ones, Viswesvaran, & Schmidt, 1993). Affective dispositions, the tendency
to experience similar emotions across situations, seem particularly relevant. Neg-
ative affectivity (NA), a generalized dispositional tendency for an individual to
experience negative emotions across time and situations, has been studied widely
in relation to perceptions of job stressors, injustice, constraints, and strains (Chen
& Spector, 1991; Skarlicki, Folger, & Tesluk, 1999; Spector & O’Connell, 1994).
Trait anxiety (Spielberger & Sydeman, 1994), the relatively stable tendency to
perceive stressful situations as threatening, has been used frequently in stress re-
search in place of NA (Ganster & Schoebroeck, 1991). Fox and Spector (1999)
found trait anxiety to be associated with CWB.

Another potentially important personality characteristic is trait anger, the like-
lihood that individuals perceive a wide range of situations as anger-provoking
(Spielberger, Krasner, & Solomon, 1988). Individuals high in trait anger have re-
ported experiencing more frequent and intense day-to-day anger across a wide
variety of provocative situations, stronger tendencies to respond to provocations
with physical and verbal antagonism, and lower instances of constructive coping
(Deffenbacher, 1992). Fox and Spector (1999) found trait anger to be particularly
associated with CWBs targeting persons in the organization.

294 FOX, SPECTOR, AND MILES

Organizational Justice as a Stressor

Organizational justice is concerned with employee perceptions of fair or just
treatment on the job. It fts defnitions of job stressors as being situations that elicit
an adaptive response (Jex & Beehr, 1991) or situations that elicit negative emotional
reactions (Spector, 1998). For example, the seminal equity theory and empirical
work by Adams (1963) suggests that inequity (injustice) motivates people to make
adaptive responses in a variety of ways, both cognitive and behavioral. More
recent work on justice has linked perceptions of injustice to negative emotions
(e.g., Skarlicki & Folger, 1997). Zohar (1995) specifcally demonstrated the role
of organizational justice in the job stress process as a role stressor, in the elicitation
of both negative emotion and consequent strain responses, but did not link stress
to counterproductive behavior.

Two major forms of justice have been studied. Distributive justice relates to
people’s perceptions of the fairness of the outcomes they receive relative to their
contributions and to the outcomes and contributions of others. Procedural jus-
tice involves people’s perceptions of the fairness of procedures used to deter-
mine those distributions (Folger & Greenberg, 1985; Levanthal, Karusa, & Fry,
1980). Several studies have linked both forms of justice perceptions with coun-
terproductive organizational behaviors. Skarlicki and Folger (1997) summarized
research that indicates employees may respond to perceptions of unfair treatment
with negative emotions, such as anger, outrage, and resentment (Folger, 1993);
desire for retribution; and a range of direct and indirect behavioral responses
such as theft (Greenberg, 1990), vandalism, sabotage, reduction of citizenship
behaviors, withdrawal, and resistance (Jermier, Knights, & Nord, 1994). Skar-
licki, Folger, and Tesluk (1999) further demonstrated that the relation between
perceived injustice and what they call organizational retaliatory behavior (ORB)
is moderated by personality factors such as negative affectivity and agreeable-
ness. Cropanzano and Baron (1991) linked injustice to emotions and workplace
confict.

Indeed, many parallels have emerged between the job stress and organizational
justice explanations of counterproductive organizational behavior, including the
central roles of emotional responses and affective dispositions. By viewing per-
ceived injustice as a type of job stressor, we arrive at a unifying framework for
understanding CWB that incorporates constraints, confict, justice perceptions,
control (autonomy), emotional responses, and affective dispositions as antecedents
of distinct categories of behavioral responses.

The Current Study

The purpose of the current study was to explore this unifed job stress/injustice/
emotions approach to counterproductive work behavior. Our primary objective
was to investigate whether job stressors and organizational justice relate in a sim-
ilar fashion to CWB. The job stressor/emotion/CWB model posits a mediating
role for emotions. Thus we set out to investigate the extent to which relations of
job stressors (interpersonal confict and organizational constraints) and organiza-
tional justice with CWB are mediated by negative emotions. We also looked at the

295 COUNTERPRODUCTIVE WORK BEHAVIOR

possibility suggested by the job stress literature that employee control (autonomy)
and affective dispositions moderate the relations between job stressors/organi-
zational justice and CWB. Throughout, we looked for distinct patterns of relations
with CWB behaviors targeting the organization versus people. Specifcally, we
hypothesized the following:

Hypothesis 1: High levels of confict and organizational constraints and low levels of per-
ceived justice are associated with high levels of negative emotions and CWB.

Hypothesis 2: High levels of negative emotions are associated with high levels of CWB.
Hypothesis 3: Negative emotions mediate the relation between stressors/injustice and

CWB.
Hypothesis 4: Perceived task autonomy, trait anxiety, and trait anger moderate the rela-

tions between job stressors/injustice and CWB. Those individuals perceiving low autonomy
and those individuals who are high in these affective traits are more likely to respond to job
stressors/injustice with CWB.

METHOD

Participants

Participants were 292 employees at a variety of organizations in southern and
central Florida. Of these, 214 (73%) were University of South Florida psychology
and management students who also were employed, and 78 (27%) were nonstudent
employees from manufacturing, fnancial, utility, entertainment, and academic
organizations in Tampa. The nonstudent employees were given the questionnaire
booklets on a strictly voluntary basis by their supervisors. The supervisors who
distributed the surveys were part-time graduate students in a Human Resource
Management class designed for practicing managers. Participants returned the
surveys directly to the researchers by U.S. mail or in sealed envelopes in sealed
containers at work. Of the 292 participants, 109 (37%) were men and 183 (63%)
were women.

Measures

The anonymous self-report survey included measures of job stressors (auton-
omy, constraints, confict, and justice), affect (positive emotions, negative emo-
tions, trait anger, and trait anxiety), and counterproductive work behaviors (CWB).
These measures were ordered following the model from left to right, that is, stres-
sors, affective responses, and behavioral responses.

Job stressors. Work constraints were measured by the Organizational Con-
straints Scale (OCS; Spector & Jex, 1998), an 11-item scale based on constraint
areas identifed by Peters and O’Connor (1980). Respondents indicated the fre-
quency with which their job performance was hindered by constraints such as
rules and procedures, availability of resources, co-workers, interruptions, and in-
adequate training. Spector and Jex (1998) reported a mean Cronbach’s ® of .85
across eight samples. Predictive validity was demonstrated with mean correlations
of .26 with physical symptoms across fve studies and ¡.38 with job satisfaction
across seven studies.

296 FOX, SPECTOR, AND MILES

Confict was assessed with Spector and Jex’s (1998) four-item Interpersonal
Confict at Work Scale (ICAWS), which measures how often the employee expe-
rienced arguments, yelling, and rudeness in interactions with co-workers. For the
confict scale (ICAWS), Spector and Jex (1998) reported a mean Cronbach’s ® of
.74 across 13 samples. Predictive validity was demonstrated with mean correla-
tions of .26 with physical symptoms across 7 studies and ¡.32 with job satisfaction
across 10 studies. For both the OCS and ICAWS measures, fve response choices
range from 1 D Less than once per month or never to 5 D Several times per day.
High scores represent high levels of constraints or confict.

Work autonomy was measured with the Factual Autonomy Scale (FAS; Fox,
Spector, & Van Katwyk, 1997), which was developed with the objective of provid-
ing items that are factual in nature and resistant to affective bias. Six items begin
with “Do you have to ask permission to : : :” and end with an aspect of scheduling,
such as “take a rest break” or “change the hours you work.” Three items begin with
“Does someone tell you : : :” and end with “what,” “when,” or “how” the person
was to work. The fve response choices range from 1 D never to 5 D always. Fox et
al. found a Cronbach’s ® of .81; they found the self-reported FAS to correlate sig-
nifcantly (r D .22) with supervisory ratings of job performance (which the more
commonly used autonomy subscale of the Job Diagnostic Survey failed to do).

Perceptions of organizational justice were assessed with distributive and proce-
dural justice scales reported in Moorman (1991). The Distributive Justice Index,
consisting of 6 items tapping the extent to which the employee is fairly rewarded
considering work inputs, was developed by Price and Mueller (1986). Five response
choices range from 1 D very unfairly to 5 D very fairly, with a high score repre-
senting a high level of distributive justice. The 12-item procedural justice scale,
which Moorman (1991) based on Levanthal’s (1980) rules of procedural justice,
was designed initially to assess two factors, fairness of formal procedures and
interactional justice. The present data confrm Moorman’s (personal e-mail corre-
spondence, February 3, 1999) fndings that these factors are highly intercorrelated
and lack an interpretable factor structure. Therefore we combined the 12 items into
a single Procedural Justice scale (as did Moorman et al., 1998). Response choices
range from 1 D strongly disagree to 5 D strongly agree, with high scores represent-
ing high levels of procedural justice. Moorman (1991) found Cronbach’s ®s of .94,
.94, and .93 for the distributive, procedural, and interactive justice scales. He re-
ported that distributive justice predicted the altruism, courtesy, sportsmanship, and
conscintiousness components of organizational citizenship behavior (r D .15, .18,
.22, and .23, respectively); procedural justice predicted courtesy, sportsmanship,
and conscientiousness (r D .17, .16, and .22, respectively); and interactive justice
predicted altruism, courtesy, sportsmanship, and conscientiousness (r D .16, .32,
.29, and .32, respectively).

Affect. A wide range of emotions experienced in response to the job was mea-
sured with the Job-Related Affective Well-Being Scale (JAWS), developed by Van
Katwyk, Fox, Spector, and Kelloway (2000). Items on the JAWS ask employees
to indicate how often any part of the job has made them feel each of 30 emotional

297 COUNTERPRODUCTIVE WORK BEHAVIOR

states. The fve response choices range from 1 D almost never to 5 D extremely
often or always. High scores represent high levels of each emotion. A positive
emotions score was obtained by summing the scores on the 13 positive affect
items; a negative emotions score was obtained by summing scores on the 17 neg-
ative affect items. Only the negative emotions score was used in the current study.
Van Katwyk et al. (2000) reported a Cronbach’s ® of .95 for the overall JAWS scale;
this score was related to organizational constraints (r D¡.39), confict (r D¡.34),
turnover intentions (r D¡.60), and physical symptoms (r D¡.33). Negative item
subscales (Cronbach’s ® D .80) predicted these variables as well, with r ’s ranging
from .34 to .58.

Spielberger’s (1979) State-Trait Personality Inventory was used to measure af-
fective disposition. The 10-item Trait Anxiety scale measures a generalized ten-
dency to experience anxiety across time and situations. The 10-item Trait Anger
scale assesses the likelihood of perceiving a wide range of situations as anger-
provoking. Four response choices range from 1 D almost never to 4 D almost al-
ways, with high scores representing high levels of trait anxiety or trait anger.
Spielberger (1979) reported ® coeffcients ranging from .80 to .92 for Trait Anx-
iety and .82 to .92 for Trait Anger across different ages and genders. In a prior
study, Fox and Spector (1999) found trait anxiety and trait anger to be among
the strongest predictors of counterproductive work behavior (r D .36 and .59,
respectively).

Counterproductive work behavior (CWB). Counterproductive work behaviors
were assessed with a behavioral checklist based on a master list compiled from a
number of existing measures (Fox & Spector, 1999; Hollinger, 1986; Knorz & Zapf,
1996; Neuman & Baron, 1998; Robinson & Bennett, 1995, Skarlicki & Folger,
1997; Spector, 1975). Our goal was to avoid duplication while including as many
distinct behaviors as possible. We put the resulting 64-item list into a scale format,
asking participants to indicate how often they had done each of the behaviors on
their present job. The fve response choices range from 1 D never to 5 D every day,
with high scores representing high incidence of counterproductive behaviors. In
addition to an overall counterproductive behavior score, subscores were provided
for behaviors targeting the organization or productivity (e.g., “Tried to look busy
while doing nothing,” “Put in to be paid for more hours than you worked,” “Came
to work late without permission,” and “Told people outside the job what a lousy
place you work for”) and those targeting other persons in the organization (e.g.,
“Insulted someone about their job performance,” “Made fun of someone’s personal
life,” “Refused to help a coworker,” and “Started an argument with a coworker”).
This classifcation was consistent with Robinson and Bennett’s (1995) taxonomy of
organizational deviance and Fox and Spector’s (1999) earlier measures of personal
and organizational CWB. Fox and Spector (1999) reported a Cronbach’s ® of .86
for their earlier version of the CWB measure. Organizational CWB was related to
constraints, locus of control, job satisfaction and frustration (r D .37, .32, ¡.45, and
.36, respectively), and Personal CWB was related to constraints, locus of control,
and frustration (r D .26, .19, and .23, respectively).

298 FOX, SPECTOR, AND MILES

RESULTS

To determine if we were justifed in combining the student and nonstudent
samples, zero-order correlations were run separately for the two samples. The z
tests comparing the resulting correlations for the two samples found differences in
only 3 of the 55 correlations among the study variables, perhaps the level expected
by chance. Differences were found for trait anger and procedural justice, CWB
targeting the organization and procedural justice, and constraints and positive
emotion. In all three cases, the relations were stronger for the nonstudent samples.
The overall consistency between the results for the two samples was perhaps due
to the nature of the University of South Florida, a primarily commuter university
with an unusually high proportion of nontraditionally aged, employed students.
Therefore for this study, the student and nonstudent samples were combined for
further analysis.

Means, standard deviations, Cronbach’s alphas, and zero-order correlations for
all study variables are presented in Table 1.

To test Hypothesis 1, we looked at the relations between the four job stres-
sor variables (confict, constraints, distributive justice, and procedural justice) and
negative emotion and between the four stressor variables and the two categories
of CWB. Confict was related signifcantly to negative emotion (r D .49), organi-
zational CWB (r D .32), and personal CWB (r D .40). Organizational constraints
were related signifcantly to negative emotion (r D .47), organizational CWB (r D
.32), and personal CWB (r D .25). Distributive justice was related signifcantly to
negative emotion (r D¡.38) and organizational CWB (r D¡.17) but not personal
CWB. Procedural justice was related signifcantly to negative emotion (r D ¡.44),
organizational CWB (r D ¡.26), and personal CWB (r D ¡.15). To rule out the
possibility that these results capitalize on chance effects, we adjusted the alpha
for the number of signifcance tests. Based on a corrected ® of .004, all the above
relations remained signifcant with the exception of that between procedural jus-
tice and personal CWB (paralleling the nonsignifcance of distributive justice to
personal CWB).

To test Hypothesis 2, we looked at correlations between negative emotion and
the two categories of CWB (organizational and personal). Both were signifcant
(r D .45 and .30, respectively).

Hypothesis 3 predicted that negative emotion mediates the relations between
job stressors and CWB. In each case mediation was tested following the procedure
recommended by Baron and Kenny (1986) in which three regression models are
investigated: the CWB on the stressor, the proposed mediator (negative emotion)
on the stressor, and the CWB on the stressor and negative emotion together. If
the beta of the stressor variable is signifcant in the frst model but nonsignifcant
or substantially reduced in the combined model, we have a pattern consistent
with mediation. Results of this analysis are presented in Table 2. Except for the
nonsignifcant personal CWB–distributive justice relation (for which a mediation
test is meaningless because it fails Kenny’s frst test of a signifcant predictor–
criterion relation), mediation of negative emotion in all CWB–justice relations
was indicated, as the beta of the stressor became nonsignifcant when emotion was

299

TA
B

LE
1

M

ea
ns

, S
ta

nd
ar

d
D

ev
ia

tio
ns

, R
el

ia
bi

lit
ie

s,
a

nd
C

or
re

la
tio

ns

Va
ria

bl
e

M
ea

n
S

D

1
2

3
4

5
6

7
8

9
10

11

1.

C
on

fi
ct

6.

0
2.

8
(.

76
)

2.

C
on

st
ra

in
ts

22

.0

7.
4

.3
4

(.
84

)
3.

D

is
tr

ib
ut

iv
e

19
.7

6.

3
¡.

23

¡.
42

(.

93
)

Ju
st

ic
e

4.
P

ro
ce

du
ra

l
56

.7

15
.9

¡.

35

¡.
40

.5

4
(.

96
)

Ju
st

ic
e

5.
P

os
iti

ve

36
.7

11

.8

¡.

20

¡.
31

.5

1
.5

0
(.

95
)

E
m

ot
io

n
6.

N
eg

at
iv

e
38

.1

11
.9

.4

9
.4

7
¡.

38

¡.

44

¡.
43

(.

91
)

E
m

ot
io

n
7.

T
ra

it
17

.3

5.
3

.3
0

.2
1
¡.

14

¡.

28

¡.
26

.5

0
(.

87
)

A
nx

ie
ty

8.

T
ra

it
17

.9

5.
7

.2
9

.3
4
¡.

13

¡.

18

¡.
18

.4

1
.4

3
(.

87
)

A
ng

er

9.
A

ut
on

om
y

28
.8

8.

5
¡.

20

¡.
22

.1

8
.2

1
.2

3
¡.

24

¡.
19

¡.

08

(.
87

)
10

.
C

W
B

32

.2

7.
6

.3
2

.3
2
¡.

17

¡.

26

¡.
16

.4

5
.2

8
.3

5
¡.

25

(.
88

)
O

rg
an

iz
at

io
n

11
.

C
W

B

49
.6

11

.2

.4
0

.2
5
¡.

09

¡.

15

¡.
02

.3

0
.2

5
.3

7
¡.

05

.6
3

(.
96

)
P

er
so

n

N
o

te
.

N
D

29
2;

C
oe

ff
ci

en
t a

lp
ha

s
ar

e
sh

ow
n

in
p

ar
en

th
es

es
o

n
th

e
di

ag
on

al
. C

or
re

la
tio

ns
a

bo
ve

j:1
3j

ar
e

si
gn

if
ca

nt
a

t t
he

p
<
:0

5
le

ve
l.

C
or

re
la

tio
ns

a
bo

ve
j:1

5j
ar

e
si

gn
if

ca
nt

a
t t

he
p
<
:0

1
le

ve
l.

C
or

re
la

tio
ns

a
bo

ve
j:1

9j ar
e

si
gn

if
ca

nt
a

t t
he

p
<
:0

01
le

ve
l.

COUNTERPRODUCTIVE WORK BEHAVIOR

300 FOX, SPECTOR, AND MILES

TABLE 2
Analysis of Mediating Role of Negative Emotion

Organizational CWB Personal CWB

Independent variables Step 1 Step 2 Step 1 Step 2

Constraints
negative emotion
R2 at each step
R2 n
F

.32¤¤¤

.10¤¤¤

.14¤

.39¤¤¤

.22¤¤¤

.12
40.5¤¤¤

.25¤¤¤

.06¤¤¤

.14¤

.23¤¤¤

.10¤¤¤

.04
16.5¤¤¤

Confict
negative emotion
R2 at each step
R2 n
F

.32¤¤¤

.11¤¤¤

.13¤

.39¤¤¤

.22¤¤¤

.11
40.3¤¤¤

.40¤¤¤

.16¤¤¤

.33¤¤¤

.13¤

.17¤¤¤

.01
29.81¤¤¤

Procedural justice
negative emotion
R2 at each step
R2 n
F

¡.26¤¤¤

.07¤¤¤

¡.08 ns
.42¤¤¤
.21¤¤¤
.14

38.44¤¤¤

¡.15¤

.02¤

¡.03 ns
.29¤¤¤
.09¤¤¤
.07

14.11¤¤¤

Distributive justice
negative emotion
R2 at each step
R2 n
F

¡.17¤¤

.03¤¤

¡.01 ns
.46¤¤¤
.20¤¤¤
.17

37.17¤¤¤

¡.09 ns

.01 ns

.02 ns

.31¤¤¤

.09¤¤¤

.08
14.09¤¤¤

Note. Standardized parameter estimates are shown. The Distributive Justice–Personal CWB media-
tion test is not meaningful, as the original relationship is nonsignifcant. In all cases, the beta of negative
emotion regressed on the stressor variable (constraints, confict, procedural justice, and distributive
justice) was signifcant at the p < :0001 level.

¤ p < .05.
¤¤ p < .01.
¤¤¤ p < .001.

added to the regression model. A similar pattern of reduced but still signifcant
stressor betas is shown for each of the remaining four CWB–stressor relations.
Thus, a pattern emerged in all seven signifcant stressor/CWB relations, suggesting
a mediating role of negative emotion.

Hypothesis 4 predicted that perceived control over work tasks (autonomy), trait
anxiety, and trait anger moderate the relations between stressors (constraints, con-
fict, and justice) and CWB (personal and organizational). Tables 3, 4, and 5 present
results of moderated regression analyses. Alphas for these hypothesis tests were
corrected for multiple tests to p < .01.

Contrary to expectations, the only signifcant autonomy–stressor interaction was
for personal CWB. Even more puzzling, the direction of the moderating effect of
autonomy was contrary to prediction, in that a higher level of interpersonal confict
is associated with personal CWB when autonomy is high. In contrast, a signifcant
and expected trait anxiety–stressor interaction was found for personal CWB. Trait

301 COUNTERPRODUCTIVE WORK BEHAVIOR

TABLE 3
Results of Moderated Regression Analysis for Autonomy as Moderator

CWB organizational CWB personal

Step Independent variable Total R2 mR2 Total R2 mR2

1 Org. Constraints .10¤¤¤ .10¤¤¤ .06¤¤¤ .06¤¤¤
2 Autonomy .14 .03¤¤¤ .06 .06
3 Org. Con. £ Autonomy .14 .00 .07 .01

1 Confict .11¤¤¤ .11¤¤¤ .16¤¤¤
2 Autonomy .14 .03¤¤¤ .16 .00
3 Confict £ Autonomy .14 .00 .18 .02¤¤

1 Procedural Justice .07¤¤¤ .07¤¤¤ .02¤ .02¤
2 Autonomy .11 .04¤¤¤ .02 .00
3 P. Just. £ Autonomy .11 .00 .03 .01

1 Distributive justice .03¤¤ .03¤¤ .01 .01
2 Autonomy .08 .05¤¤¤ .01 .00
3 D. Just. £ Autonomy .08 .00 .03 .02¤

Note. N D 292.
¤ p < .05.
¤¤ p < .01.
¤¤¤ p < .001.

TABLE 4
Results of Moderated Regression Analysis for Trait Anxiety as Moderator

CWB organizational CWB personal

Step Independent variable Total R2 mR2 Total R2 mR2

1 Org. Constraints .10¤¤¤ .10¤¤¤ .06¤¤¤ .06¤¤¤
2 Trait Anxiety .15 .05¤¤¤ .10 .04¤¤¤
3 Org. Con. £ T. Anxiety .15 .00 .13 .03¤¤

1 Confict .11¤¤¤ .11¤¤¤ .16¤¤¤ .16¤¤¤
2 Trait Anxiety .15 .04¤¤¤ .18 .02¤
3 Confict £ T. Anxiety .15 .00 .19 .01¤

1 Procedural Justice .07¤¤¤ .07¤¤¤ .02¤ .02¤
2 Trait Anxiety .12 .05¤¤¤ .07 .05¤¤¤
3 P. Just. £ T. Anxiety .12 .00 .07 .00

1 Distributive justice .03¤¤ .03¤¤ .01 .01
2 Trait anxiety .10 .07¤¤¤ .07 .05¤¤¤
3 D. Just. £ T. Anxiety .10 .00 .07 .00

Note. N D 292.
¤ p < .05.
¤¤ p < .01.
¤¤¤ p < .001.

302 FOX, SPECTOR, AND MILES

TABLE 5
Results of Moderated Regression Analysis for Trait Anger as Moderator

CWB organizational CWB personal

Step Independent variable Total R2 mR2 Total R2 mR2

1 Org. Constraints .10¤¤¤ .10¤¤¤ .06¤¤¤ .06¤¤¤
2 Trait Anger .17 .07¤¤¤ .15 .09¤¤¤
3 Org. Con. £ T. Anger .17 .00 .16 .01

1 Confict .11¤¤¤ .11¤¤¤ .16¤¤¤ .16¤¤¤
2 Trait Anger .18 .07¤¤¤ .23 .07¤¤¤
3 Confict £ T. Anger .18 .00 .26 .03¤¤¤

1 Procedural Justice .07¤¤¤ .07¤¤¤ .02¤ .02¤
2 Trait Anger .16 .09¤¤¤ .14 .12¤¤¤
3 P. Just. £ T. Anger .17 .01 .14 .00

1 Distributive justice .03¤¤ .03¤¤ .01 .01
2 Trait anger .14 .11¤¤¤ .14 .13¤¤¤
3 D. Just. £ T. Anger .14 .00 .14 .00

Note. N D 292.
¤ p < .05.
¤¤ p < .01.
¤¤¤ p < .001.

anxiety did not moderate relations between stressors and organizational CWB.
Also, a signifcant trait anger–stressor interaction was found for personal CWB, but
not organizational CWB. In each of these last two cases, the pattern of signifcant
moderation revealed that when the trait was high, there was a steeper slope between
stressor and CWB than when the trait was low. That is, for individuals high in trait
anxiety, but not for individuals low in trait anxiety, higher levels of constraints were
associated with higher levels of personal CWB. Similarly, for individuals high in
trait anger, but not for those low in trait anger, higher levels of confict were
associated with higher levels of personal CWB. Thus, only very limited support
was found for Hypothesis 4. Personality characteristics (trait anger and anxiety)
interacted as expected with stressors for personal, but not organizational, CWB,
and weak evidence suggested that autonomy moderated the confict–personal CWB
relation in the opposite direction as hypothesized.

DISCUSSION

Results of the current study provide some support for predictions derived from
the stress/emotion/CWB model (Spector et al., 1998). Specifcally, it was found
that job stressors, including perceived injustice related to both negative emotions
and CWB (Hypothesis 1); that negative emotions related to CWB (Hypothesis 2);
and that in most cases there was at least partial mediation of emotions in the
relations between job stressors and CWB (Hypothesis 3). This last test is critical
to the model and suggests emotions play a central role in the process from stressor
to strain in general and CWB in particular.

303 COUNTERPRODUCTIVE WORK BEHAVIOR

Indeed, inspection of the mediator analyses in Table 2 shows patterns of relations
that are generally supportive. In seven of the eight cases, the stressor variable was
correlated signifcantly with CWB. Negative emotion was correlated signifcantly
with all stressor and CWB variables. The regression results for all seven cases
in which the stressor was related signifcantly to CWB showed a reduction in the
regression coeffcient when emotion was added to the equation containing only the
stressor as a predictor. In three cases the coeffcient for the stressor lost signifcance,
and in fve of seven cases, the magnitude of the coeffcient was reduced by more
than half.

In general, organizational stressors (such as constraints and injustice) were more
closely associated with organizational than personal types of CWB, and interper-
sonal confict was more closely associated with personal than organizational CWB.
Yet, the results of post hoc z tests showed these differences to be signifcant only in
the cases of autonomy and procedural justice being more strongly associated with
organizational CWB and confict being more strongly associated with personal
CWB. Similarly, support for mediation was somewhat stronger for organization-
targeted than person-targeted CWB. Altogether, the tighter linkage between con-
straints inhibiting job performance and CWB directed at the organization rather
than other people, suggests the two types of CWB may follow somewhat separate
paths.

The failure to fnd more evidence for complete mediation is also not unexpected.
Spector and Fox (in press) discussed how causality is likely multidirectional, with
background mood or emotional state predisposing people to perceive or not per-
ceive job stressors. Thus, people experiencing high levels of negative mood may
be hyperreactive to job stressors. This can result in a cycle in which background
mood predisposes a person to perceive the work situation as a job stressor, which
further induces negative mood and heightened strain.

In general, our results for Hypotheses 1 and 2 were consistent with prior re-
search in showing a relation of justice with negative emotions (Skarlicki & Folger,
1997; Zohar, 1995), and with CWB (Skarlicki & Folger, 1997). Justice also was
correlated signifcantly with both of the other job stressors (interpersonal confict
and organizational constraints). These fndings lend further support to the idea that
perceptions of injustice can be conceptualized as a form of perceived job stress.
Situations seen by people as unfair are stressors that may lead to negative emo-
tions and presumably to subsequent strains beyond CWB. For example, justice
has been shown to relate signifcantly with job (dis-)satisfaction (e.g., Moorman,
1991). Subsequent work should expand this area of inquiry to look at other types
of strains (e.g., physical symptoms and ill health) that have been linked to stressors
at work (Jex & Beehr, 1991).

Our proposed model suggested a moderating role for trait anger and anxiety. In
general, the data were nonsupportive of moderation. Using a conservative (.01) ®

(corrected for multiple signifcance tests), only the interactive effects of autonomy
and confict, trait anxiety and constraints, and trait anger and confict for per-
sonal CWB remained signifcant. Using a more liberal ® of .05, however, patterns

304 FOX, SPECTOR, AND MILES

emerged which suggest the usefulness of further investigation of moderating roles
for trait anger and anxiety in stressor-personal CWB relations. In these cases, as
expected, individuals higher in the affective trait reported being more sensitive
to stressors in their personal CWB reactions. There was no evidence of modera-
tion for organizational CWB or with the justice variables. Perhaps organizational
CWB is more affected by environmental variables than personality, whereas per-
sonal CWB is affected by both. Indirect support for this idea comes from work
with organizational citizenship behavior (OCB), which in some ways is the op-
posite of CWB. McNeely and Meglino (1994) found that the personality trait of
empathy predicted OCB directed toward other people at work but not toward the
organization. Of course, it should be kept in mind that we looked at only two
personality variables and four stressors, and it is conceivable that results would
differ for other variables. Furthermore, moderated regression is known to suffer
from low statistical power, so future tests should have larger samples.

One of the most surprising fndings was the failure of the data to support the
predicted moderating role of job control (autonomy), particularly in the relations
between task-related stressors (constraints and injustice) and task-related (organi-
zational) CWB. It may be that the category of organizational CWBs is too broad,
and contains many behaviors that are not task-specifc, such as destroying com-
pany property. To explore this possibility, the moderator analysis was performed
using a subset of the organizational CWB items, which appear to directly in-
volve work sabotage. This post hoc test revealed a signifcant interaction effect
of distributive justice by autonomy on work sabotage behaviors, in the expected
direction. That is, for individuals who perceived low autonomy, but not those who
perceived high autonomy, low distributive justice was associated with high lev-
els of work sabotage behaviors. Future studies might systematically divide the
CWB items into subscales according to their relevance to task performance, inter-
personal relations, withdrawal, and so forth in order to determine more specifc,
differential patterns of relations to organizational antecedents such as job charac-
teristics.

Autonomy was found to moderate the relation between interpersonal confict
and personal CWB, but closer examination revealed the direction of interaction
to be opposite the hypothesized relation. For individuals who perceived high but
not low autonomy, higher confict was associated with higher personal CWB.
A possible explanation is that autonomy gave individuals latitude to respond to
confict with personal retaliatory behaviors. That is, those individuals who had
high autonomy were in powerful enough positions that they had the latitude to
engage in personal CWB without having to fear retribution. This may explain the
lack of overall moderation by autonomy. In a previous study, Fox and Spector
(1999) found the belief that one has the ability to harm the organization without
being punished to be one of the strongest predictors of CWB. This may con-
found the theorized relationship, that individuals with greater job autonomy are
less likely to experience stress and thereby less likely to respond counterproduc-
tively.

305 COUNTERPRODUCTIVE WORK BEHAVIOR

Limitations of the Study
Reliance on cross-sectional, self-report methodology is always problematic. It

is an ongoing concern in organizational behavior research that the use of a single
source of data, such as self-report questionnaires, may result in an overstatement
of relationships among the variables. However, given our focus on affective and
behavioral responses to the perceived rather than “objective” environment, the
diffculty of obtaining uncontaminated measures of counterproductive behavior,
and ethical concerns with the possibility of putting research participants at risk
in the accumulation of evidence of CWB, we believe that anonymous self-reports
provide the closest available approximation of these relations (for a more extensive
discussion of these issues, see Fox & Spector, 1999).

At the same time, we recognize the need for creative research designs using
nonincumbent or more objective sources of data. Good examples of such studies
include Perlow and Latham’s (1993) longitudinal study using recorded instances
of abusive behavior as the dependent variable; and Robinson and O’Leary-Kelly’s
(1998) use of peer ratings of workgroup climate as the measure of environmental
factors. Yet, each of these designs is able to measure only a subset of the model
we are considering. Future research might incorporate peer, subordinate, and su-
pervisor reports of the incumbent’s counterproductive work behavior (see Heacox,
1996), such as survey dyads consisting of an employee’s self-report of perceptions
of stressors and a co-worker’s reports of behaviors by that employee. An alterna-
tive might be self-reports of being the target of CWB, as in the mobbing research
of Knorz and Zapf (1996). As the models become more complex, tests of the data
certainly will require more rigorous analytic tools, such as Structural Equation
Modeling. Finally, insights might be gained from a policy-capturing investigation
of the circumstances under which employees would fnd various forms of CWB
to be justifed or acceptable.

A further limitation to the generalizability of the results of the study was the
use of a convenience sample of nonstudent employees and the combination of that
sample with a sample of employed students. Yet, there was little evidence that the
student sample responded differently than did the nonstudent sample. Furthermore,
generalizability was enhanced in that the populations from which the participants
were drawn were highly heterogeneous, covering quite a few organizations and a
wide range of organizational conditions.

Finally, at the current time the model underlying this research is rather general.
We chose specifc job stressors and other variables that prior research suggested
might be important. We recognize that this was a piecemeal test of parts of the
model. As evidence for linkages among specifc variables is uncovered, it might
prove useful to conduct a more complete test rather than investigating individual
linkages. Structural equation modeling could accomplish this objective.

Nevertheless, the current study provides evidence to support some of the link-
ages suggested by our job stress/emotion/CWB model. Perhaps most importantly,
results for all signifcant stressor/CWB relations supported the hypothesized me-
diator role for emotions, either fully or partially. The results, however, do suggest

306 FOX, SPECTOR, AND MILES

the need to refne the model; for example, personality seemed more important as a
moderator of personal CWB than organizational CWB, and did not seem relevant
for justice. The Spector and Fox (in press) model may provide a general frame-
work, but additional work is needed to uncover specifc linkages among variables.
Above all, this cross-sectional study cannot provide a causal test.

Still, our fndings suggest that organizations may be able to reduce the levels
of work behaviors that undermine their effectiveness by developing human re-
source policies and practices that take into consideration their possible emotional
effects on employees. This is not to say that all negative situations can or should be
avoided, but rather that attempts to manage actively the emotional effects of human
resources, operations and resource allocation systems are likely to result in lower
CWB. The implications of this study suggest an alternative to the predominant “se-
lection” solution to CWB, in which individuals with certain personality tendencies
that may predict CWBs are screened out of the organization during the selection
process. Consideration of justice, autonomy, and employees’ feelings needs to be
included in the design of jobs and human resource systems—not because it is “the
nice thing to do,” but because of its ramifcations in the kinds of behaviors that,
even in covert or subtle ways, may do serious harm to the organization and its
members.

REFERENCES

Adams, J. S. (1963). Toward an understanding of inequity. Journal of Abnormal and Social Psychology,
67, 422–436.

Allen, V. L., & Greenberger, D. B. (1980). Destruction and perceived control. In A. Baum & J. E.
Singer (Eds.), Applications of personal control. Hillsdale, NJ: Erlbaum.

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psycho-
logical research: Conceptual, strategic and statistical considerations. Journal of Personality and
Social Psychology, 51, 1173–1182.

Bies, R. J., Tripp, T. M., & Kramer, R. M. (1997). At the breaking point: Cognitive and social dynamics
of revenge in organizations. In R. A. Giacalone & J. Greenberg (Eds.), Antisocial behavior in
organizations (pp. 18–36). Thousand Oaks, CA: Sage.

Cartwright, S., & Cooper, C. L. (1997). Managing workplace stress. Thousand Oaks, CA: Sage. Chen,
P. Y., & Spector, P. E. (1991). Negative affectivity as the underlying cause of correlations between
stressors and strains. Journal of Applied Psychology, 76, 398–407.

Chen, P. Y., & Spector, P. E. (1992). Relationships of work stressors with aggression, withdrawal, theft
and substance use: An exploratory study. Journal of Occupational and Organizational Psychology,
65, 177–184.

Cropanzano, R., & Baron, R. (1991). Injustice and organizational confict: The moderating role of
power restoration. International Journal of Confict Management, 2, 5–26.

Deffenbacher, J. L. (1992). Trait anger: Theory, fndings, and implications. In C. D. Spielberger &
J. N. Butcher (Eds.), Advances in personality assessessment (Vol. 9, pp. 177–201). Hillsdale, NJ:
Erlbaum.

Folger, R. (1993). Reactions to mistreatment at work. In J. K. Murmighan (Ed.), Social psychology in
organizations: Advances in theory and research (pp. 161–183). Englewood Cliffs, NJ: Prentice
Hall.

Folger, R., & Greenberg, J. (1985). Procedural justice: An interpretive analysis of personnel systems.
In K. M. Rowland & G. R. Ferris (Eds.), Research in personnel and human resource management
(Vol. 3, pp. 141–183). Greenwich, CT: JAI Press.

307 COUNTERPRODUCTIVE WORK BEHAVIOR

Fox, S., & Spector, P. E. (1999). A model of work frustration-aggression. Journal of Organizational
Behavior, 20, 915–931.

Fox, S., Spector, P. E., & Van Katwyk, P. (1997). Objectivity in the assessment of control at work. Paper
presented at the annual meeting of the Society for Industrial and Organizational Psychology, St.
Louis.

Frankenhaeuser, M., & Johansson, G. (1986). Stress at work: Psychobiological and psychosocial as-
pects. International Review of Applied Psychology, 35, 287–299.

Ganster, D. C., & Schaubroeck, J. (1991). Work stress and employee health. Journal of Management,
17, 235–271.

Giacalone, R. A., & Greenberg, J. (1997). Antisocial behavior in organizations. Thousand Oaks, CA:
Sage.

Greenberg, J. (1990). Employee theft as a reaction to underpayment inequity: The hidden cost of pay
cuts. Journal of Applied Psychology, 75, 561–568.

Heacox, N. J. (1996). The relationship between organizational frustration and aggressive behaviors in
the workplace. Unpublished doctoral dissertation, California School of Professional Psychology,
San Diego.

Hogan, J., & Hogan, R. (1989). How to measure employee reliability. Journal of Applied Psychology,
74, 273–279.

Hollinger, R. C. (1986). Acts against the workplace: Social bonding and employee deviance. Deviant
Behavior, 7, 53–75.

Jermier, J. M., Knights, D., & Nord, W. (1994). Resistance and power in organizations. London:
Routledge.

Jex, S. M., & Beehr, T. A. (1991). Emerging theoretical and methodological issues in the study
of work-related stress. Research in Personnel and Human Resource Management, 9, 311–
365.

Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snoek, D., & Rosenthal, R. A. (1964). Organizational stress:
Studies in role confict and ambiguity. New York: Wiley.

Karasek, R. A. (1979) Job demands, job decision latitude, and mental strain: Implications for job
redesign, Administrative Science Quarterly, 24, 285–308.

Knorz, C., & Zapf, D. (1996). Mobbing—Eine extreme Form sozialer Stressoren am Arbeitsplatz,
[Mobbing—An extreme form of social stressors in the workplace]. Zeitschrift fuer Arbeits- und
Organisationspsychologie, 40, 12–21.

Lazarus, R. S. (1991). Psychological stress in the workplace. In P. L. Perrewe (Ed.), Journal of Social
Behavior and Personality: Vol. 6. Handbook on job stress(special issue) (pp. 1–13). San Rafael,
CA: Select Press.

Levanthal, G. S. (1980). What should be done with equity theory? New approaches to the study of
fairness in social relationships. In K. Gergen, M. S. Greenberg, & R. H. Willis (Eds.), Social
exchange: Advances in theory and research (pp. 27–55). New York: Plenum.

Levanthal, G. S., Karuza, J., & Fry, W. R. (1980). Beyond fairness: A theory of allocation preferences.
In G. Mikula (Ed.), Justice and social interaction (pp. 167–218). New York: Springer-Verlag.

Lovallo, W. R. (1997). Stress & Health: Biological and psychological interactions. Thousand Oaks,
CA: Sage.

McNeely, B. L., & Meglino, B. M. (1994). The role of dispositional and situational antecedents in
prosocial organizational behavior: An examination of the intended benefciaries of prosocial
behavior. Journal of Applied Psychology, 79, 836–844.

Moorman, R. H. (1991). Relationship between organizational justice and organizational citizen-
ship behaviors: Do fairness perceptions infuence employee citizenship? Journal of Applied
Psychology, 76, 845–855.

Moorman, R. H., Blakely, G. L., & Niehoff, B. P. (1998). Does perceived organizational support
mediate the relationship between procedural justice and organizational citizenship behavior?
Academy of Management Journal, 41, 351–357.

Neuman, J. H., & Baron, R. A. (1998). Workplace violence and workplace aggression: Evidence
concerning specifc forms, potential causes, and preferred targets. Journal of Managements, 24,
391–419.

308 FOX, SPECTOR, AND MILES

Ones, D. S., Viswesvaran, C., & Schmidt, F. L. (1993). Comprehensive meta-analysis of integrity test
validities: Findings and implications for personnel selection and theories of job performance.
Journal of Applied Psychology, 78, 679–703.

Payne, R. (1999). Stress at work: A conceptual framework. In J. Firth-Cozens & R. Payne (Eds.),
Stress in health professionals: Psychological and organisational causes and interventions (pp.
3–16). Chichester, UK: Wiley.

Perlow, R., & Latham, L. L. (1993). Relationship of client abuse with locus of control and gender: A
longitudinal study. Journal of Applied Psychology, 78, 831–834.

Peters, L. H., & O’Connor, E. J. (1980). Situational constraints and work outcomes: The infuences
of a frequently overlooked construct. Academy of Management Review, 5, 391–397.

Price, J. L., & Mueller, C. W. (1986). Handbook of organizational measurement. Marshfeld, MA:
Pittman.

Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace behaviors: A multidimen-
sional scaling study. Academy of Management Journal, 38, 555–572.

Robinson, S. L., & O’Leary-Kelly, A. M. (1998). Monkey see, monkey do: The infuence of work
groups on the antisocial behavior of employees. Academy of Management Journal, 41, 658–672.

Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: The roles of distributive, procedural,
and interactional justice. Journal of Applied Psychology, 82, 434–443.

Skarlicki, D. P., Folger, R., & Tesluk, P. (1999). Personality as a moderator in the relationship between
fairness and retaliation. Academy of Management Journal, 42, 100–108.

Spector, P. E. (1975). Relationships of organizational frustration with reported behavioral reactions of
employees. Journal of Applied Psychology, 60, 635–637.

Spector, P. E. (1978). Organizational frustration: A model and review of the literature. Personnel
Psychology, 31, 815–829.

Spector, P. E. (1986). Perceived control by employees: A meta-analysis of studies concerning
autonomy and participation at work. Human Relations, 39, 1005–1016.

Spector, P. E. (1998). A control theory of the job stress process. In C. L. Cooper (Ed.), Theories of
organizational stress (pp. 153–169). Oxford, UK: Oxford Univ. Press.

Spector, P. E., Dwyer, D. J., & Jex, S. M. (1988). The relationship of job stressors to affective,
health, and performance outcomes: A comparison of multiple data sources. Journal of Applied
Psychology, 73, 11–19.

Spector, P. E., & Fox, S. (in press). An emotion-centered model of voluntary work behavior: Some
parallels between Counterproductive Work Behavior (CWB) and Organizational Citizenship
Behavior (OCB). Human Resource Management Review.

Spector, P. E., & Jex, S. M. (1998). Development of four self-report measures of job stressors and strain:
Interpersonal Confict at Work Scale, Organizational Constraints Scale, Quantitative Workload
Inventory, and Physical Symptoms Inventory. Journal of Occupational Health Psychology, 3,
356–367.

Spector, P. E., & O’Connell, B. J. (1994). The contribution of personality traits, negative affectivity,
locus of control and Type A to the subsequent reports of job stressors and job strains. Journal of
Occupational and Organizational Psychology, 67, 1–11.

Spielberger, C. D. (1979). Preliminary manual for the State-Trait Personality Inventory (STPI).
Unpublished manuscript, University of South Florida, Tampa.

Spielberger, C. D., Krasner, S. S., & Solomon, E. P. (1988). The experience, expression, and control
of anger. In M. P. Janisse (Ed.), Health psychology: Individual differences and stress. New York:
Springer-Verlag.

Spielberger, C. D., & Sydeman, S. J. (1994). State-Trait anxiety inventory and State-Trait anger
expression inventory. In M. E. Maruish (Ed.), The use of psychological tests for treatment
planning and outcome assessment. Hillsdale, NJ: Erlbaum.

Storms, P. L., & Spector, P. E. (1987). Relationships of organizational frustration with reported
behavioral reactions: The moderating effect of perceived control. Journal of Occupational
Psychology, 60, 227–234.

309 COUNTERPRODUCTIVE WORK BEHAVIOR

Thompson, S. C. (1981). Will it hurt less if I can control it? A complex answer to a simple question,
Psychological Bulletin, 90, 89–101.

Van Katwyk, P. T., Fox, S., Spector, P. E., & Kelloway, E. K. (2000). Using the Job-related Affective
Well-being Scale (JAWS) to investigate affective responses to work stressors. Journal of
Occupational Health Psychology, 5, 219–230.

Zohar, D. (1995). The justice perspective on job stress. Journal of Organzational Behavior, 16, 487–495.

Received January 18, 2000; published online June 1, 2001

Work & Stress

Vol. 22, No. 3, July�September 2008, 187�200

POSITION PAPER

Work engagement: An emerging concept in occupational health psychology

Arnold B. Bakkera, Wilmar B. Schaufelib, Michael P. Leiterc and Toon W. Tarisd

aInstitute of Psychology, Erasmus University Rotterdam, The Netherlands; bDepartment of Social and

Organizational Psychology, Utrecht University, The Netherlands; cDepartment of Psychology, Acadia

University, Canada; dBehavioral Science Institute, Department of Work and Organizational Psychology,

Radboud University Nijmegen, The Netherlands

This position paper introduces the emerging concept of work engagement: a positive, fulfilling, affective-

motivational state of work-related well-being that is characterized by vigour, dedication, and

absorption. Although there are different views of work engagement, most scholars agree that engaged

employees have high levels of energy and identify strongly with their work. The most often used

instrument to measure engagement is the Utrecht Work Engagement Scale, a self-report instrument that

has been validated in many countries across the world. Research on engagement has investigated how

engagement differs from related concepts (e.g., workaholism, organizational commitment), and has

focused on the most important predictors of work engagement. These studies have revealed that

engagement is a unique concept that is best predicted by job resources (e.g., autonomy, supervisory

coaching, performance feedback) and personal resources (e.g., optimism, self-efficacy, self-esteem).

Moreover, the first studies have shown that work engagement is predictive of job performance and client

satisfaction. The paper closes with an account of what we do not know about work engagement, and

offers a brief research agenda for future work.

Keywords: burnout; job resources; performance; workaholism; work engagement

Introduction

Recently, psychology has been criticized as primarily dedicated to addressing mental illness

rather than mental ‘‘wellness.’’ This prevailing negative bias of psychology is illustrated by the

fact that the number of publications on negative states exceeds that on positive states by a

ratio of 14:1 (Myers, 2000). The purpose of Positive Psychology is ‘‘ . . . to begin to catalyze a

change in the focus of psychology from pre-occupation only with repairing the worst things in

life to also building positive qualities’’ (Seligman & Csikszentmihalyi, 2000, p. 5). This

advocated positive turn is also relevant for occupational health psychology. Failing to

recognize the positive aspects of work is inappropriate and, as Turner, Barling, and

Zacharatos (2002, p. 715) have argued, ‘‘ . . . it is time to extend our research focus and

explore more fully the positive sides, so as to gain full understanding of the meaning and

effects of working.’’

This special issue responds to the call for more research into positive psychology by

focusing on work engagement: a positive, fulfilling, affective-motivational state of work-related

*Corresponding author. Email: [email protected]

ISSN 0267-8373 print/ISSN 1464-5335 online

# 2008 Taylor & Francis

DOI: 10.1080/02678370802393649

http://www.informaworld.com

188 A.B. Bakker et al.

well-being that can be seen as the antipode of job burnout. Engaged employees have high levels

of energy, are enthusiastic about their work, and they are often fully immersed in their job so

that time flies (Macey & Schneider, 2008; May, Gilson, & Harter, 2004; Schaufeli & Bakker, in

press). Whereas research on burnout has produced thousands of articles during the past three

decades, research on work engagement has just begun to emerge. This is curious, in that modern

organizations expect their employees to be proactive and show initiative, take responsibility for

their own professional development, and to be committed to high quality performance
standards. Thus, they need employees who feel energetic and dedicated, and who are absorbed

by their work, i.e., who are engaged with their work (Bakker & Schaufeli, 2008). As we will see

in this article, work engagement can make a true difference for employees and may offer

organizations a competitive advantage (see also Bakker, in press; Demerouti & Cropanzano,

in press).

Different views on work engagement

Interestingly, it is research on burnout that has stimulated most contemporary research on

work engagement. Contrary to those who suffer from burnout, engaged employees have a

sense of energetic and effective connection with their work, and instead of stressful and

demanding they look upon their work as challenging. Two different but related schools of

thought exist that consider work engagement as a positive, work-related state of well-being or

fulfilment. According to Maslach and Leiter (1997), engagement is characterized by energy,
involvement, and efficacy, the direct opposites of the three burnout dimensions. They argue

that, in the case of burnout, energy turns into exhaustion, involvement into cynicism, and

efficacy into ineffectiveness. By implication, engagement is assessed by the opposite pattern of

scores on the three dimensions of the Maslach Burnout Inventory (MBI; Maslach, Jackson,

& Leiter, 1996): low scores on exhaustion and cynicism, and high scores on professional

efficacy.

The alternative view considers work engagement as an independent, distinct concept that

is related negatively to burnout. Consequently, work engagement is defined and operatio-

nalized in its own right as ‘‘ . . . a positive, fulfilling, work-related state of mind that is

characterized by vigour, dedication, and absorption’’ (Schaufeli, Salanova, González-Romá,

& Bakker, 2002, p. 74). That is, in engagement, fulfilment exists in contrast to the voids of life

that leave people feeling empty as in burnout. Vigour is characterized by high levels of energy

and mental resilience while working, the willingness to invest effort in one’s work, and

persistence even in the face of difficulties. Dedication refers to being strongly involved in one’s

work, and experiencing a sense of significance, enthusiasm, inspiration, pride, and challenge.

Absorption is characterized by being fully concentrated and happily engrossed in one’s work,
whereby time passes quickly and one has difficulties with detaching oneself from work.

Accordingly, vigour and dedication are considered direct opposites of exhaustion and

cynicism, respectively, the two core symptoms of burnout. The continuum that is spanned by

exhaustion and vigour has been labelled ‘‘energy,’’ whereas the continuum that is spanned by

cynicism and dedication has been labelled ‘‘identification’’ (González-Roma, Schaufeli,

Bakker, & Lloret, 2006). Hence, work engagement is characterized by a high level of energy

and strong identification with one’s work, whereas burnout is characterized by the opposite: a

low level of energy and poor identification with one’s work (see also Demerouti & Bakker,

2008). In addition, based on in-depth interviews (Schaufeli, Taris, Le Blanc, Peeters, Bakker,

& De Jonge, 2001), absorption was included as the third constituting aspect of work

engagement.

189 Work & Stress

Kahn (1990) took a different approach when he conceptualized engagement as

the ‘‘ . . . harnessing of organization member’s selves to their work roles: in engagement, people

employ and express themselves physically, cognitively, emotionally and mentally during role

performances’’ (p. 694). Thus, engaged employees put much effort into their work because they

identify with it. According to Kahn (1990, 1990) a dynamic, dialectical relationship exists

between the person who drives personal energies (physical, cognitive, emotional, and mental)

into his or her work role on the one hand, and the work role that allows this person to express
him or herself on the other hand. Kahn (1992) differentiated the concept of engagement from

psychological presence or the experience of ‘‘being fully there,’’ namely when ‘‘ . . . people feel

and are attentive, connected, integrated, and focused in their role performance’’ (p. 322). Put

differently, here engagement as behaviour (driving energy in one’s work role) is considered as the

manifestation of psychological presence, a particular mental state. In its turn, engagement is

assumed to produce positive outcomes, both at the individual level (personal growth and

development) as well as at the organizational level (performance quality).

Inspired by the work of Kahn (1990, 1992), Rothbard (2001) took a slightly different

perspective and defined engagement as a two-dimensional motivational construct that

includes attention (‘‘ . . . the cognitive availability and the amount of time one spends thinking

about a role’’; p. 656) and absorption (‘‘ . . . the intensity of one’s focus on a role’’; p. 656). It is

important to note that the key reference of engagement for Kahn (1990, 1992) is the work

role, whereas for those who consider engagement as the positive antithesis of burnout it is the

employee’s work activity, or the work itself.

Most scholars agree that engagement includes an energy dimension and an identification

dimension. Work engagement is characterized by a high level of energy and strong
identification with one’s work. The perspective of this special issue is that the field is served

best by a consistent construct for work engagement, one that focuses on employees’

experience of work activity. Unfortunately, the broad exploration of constructs over the past

decade has not produced consensus about its meaning. In contrast, a recent review of Macey

and Schneider (2008) documented the proliferation of various definitions of engagement,

many of them being old wine in new bottles. These authors try to ‘‘solve’’ the conceptual

confusion by proposing employee engagement as an all-inclusive umbrella term that contains

different types of engagement (i.e., trait engagement, state engagement, and behavioural

engagement), each of which entails various conceptualizations; e.g., proactive personality

(trait engagement), involvement (state engagement), and organizational citizenship behaviour

(behavioural engagement). In contrast, we advocate the use of engagement as a specific, well-

defined, and properly operationalized psychological state that is open to empirical research

and practical application. This special issue documents the fruitfulness of this approach.

Measurement

There are several instruments that can be used to assess work engagement (see Schaufeli &

Bakker, in press), but we will concentrate on the instruments that have been validated more

extensively. Those who follow Maslach and Leiter’s (1997, 2008) approach can use the

MBI (Maslach et al., 1996) to assess energy (low score on exhaustion), involvement (low

score on cynicism), and professional efficacy (high score on efficacy). An alternative

instrument for the assessment of work engagement is the Oldenburg Burnout Inventory

(OLBI; Demerouti & Bakker, 2008; Demerouti, Bakker, Nachreiner, & Ebbinghaus, 2002).

This instrument was developed originally to assess burnout, but includes both positively

and negatively phrased items, and hence it can be used to assess work engagement as well

190 A.B. Bakker et al.

(see González-Roma et al., 2006). Researchers interested in assessing work engagement with

the OLBI may recode the negatively framed items. The OLBI includes two dimensions: one

ranging from exhaustion to vigour and a second ranging from cynicism (disengagement) to

dedication. The reliability and factorial validity of the OLBI has been confirmed in studies

conducted in Germany, Greece, the Netherlands, the USA, and South Africa (Demerouti &

Bakker, 2008). Results of these studies clearly showed that a two-factor structure with vigour

and dedication (referred to as exhaustion and disengagement in several of these studies) as

the underlying factors fitted better to the data of several occupational groups than alternative

factor structures.

The most often used instrument to measure engagement is the Utrecht Work Engagement

Scale (UWES; Schaufeli & Bakker, in press; Schaufeli et al., 2002) that includes three

subscales: vigour, dedication, and absorption. The UWES has been validated in several

countries, including China (Yi-Wen & Yi-Qun, 2005), Finland (Hakanen, 2002), Greece

(Xanthopoulou, Bakker, Demerouti, & Kantas, in press), Japan (Shimazu et al., 2008), South

Africa (Storm & Rothmann, 2003), Spain (Schaufeli et al., 2002), and the Netherlands

(Schaufeli & Bakker, 2003; Schaufeli et al., 2002). All investigations used confirmatory factor

analyses and showed that the fit of the hypothesized three-factor structure to the data was

superior to that of alternative factor models. In addition, the internal consistencies of the

three subscales proved to be sufficient in each study. It should be noted, however, that some

studies failed to replicate the three-factor structure of work engagement (e.g., Shimazu et al.,

2008; Sonnentag, 2003). This may be attributed partly to translation problems when it comes

to items that contain metaphors (e.g., ‘‘Time flies when I am working’’). Furthermore,

Schaufeli and Bakker (in press) have argued that the overall score for work engagement may

sometimes be more useful in empirical research than the scores on the three separate

dimensions of the UWES. Schaufeli, Bakker, and Salanova (2006) developed a 9-item version

of the UWES, and provided evidence for its cross-national validity. They showed that the

three engagement dimensions are moderately strongly related.

What we know about work engagement

Previous research on work engagement has primarily used the well-validated UWES, and

focused on the predictors of work engagement (job and personal resources), outcomes

(performance), and differences from related concepts (e.g., workaholism, and organizational

commitment). In this section, we briefly review the available evidence, and then we turn to a

research agenda for work engagement.

Work engagement is not the same as workaholism

Workaholics spend a great deal of time in work activities when given the discretion to choose

whether to do so; they are excessively hard workers. In addition, workaholics are reluctant to

disengage from work and they persistently and frequently think about work when they are

not at work. This suggests that workaholics are obsessed with their work; they are compulsive

workers (Schaufeli, Taris, & Bakker, 2006; Scott, Moore, & Miceli, 1997). Engaged employees

work hard (vigour), are involved (dedicated), and feel happily engrossed (absorbed) in their

work. In this sense, they seem similar to workaholics. However, in contrast to workaholics,

engaged workers lack the typical compulsive drive. For them work is fun, not an addiction, as

was concluded from a qualitative study among 15 engaged workers (Schaufeli et al., 2001).

Engaged employees work hard because they like it and not because they are driven by a

191 Work & Stress

strong inner urge they cannot resist. For workaholics, their need to work is so exaggerated

that it endangers their health, reduces their happiness, and deteriorates their interpersonal

relations and social functioning (e.g., Bakker, Demerouti, & Burke, in press). In short, work

engagement can be discriminated from workaholism (Taris, Schaufeli, & Shimazu, in press).

Previous studies have also shown that work engagement can be discriminated from Type-A

behaviour (Hallberg, Johansson, & Schaufeli, 2007), and from job involvement and

organizational commitment (Hallberg & Schaufeli, 2006). In addition, Halbesleben and
Wheeler (2008) have shown that work engagement can be distinguished from job

embeddedness.

Job resources facilitate engagement

Previous studies have consistently shown that job resources such as social support from

colleagues and supervisors, performance feedback, skill variety, autonomy, and learning

opportunities are positively associated with work engagement (Bakker & Demerouti, 2008;

Halbesleben, in press; Schaufeli & Salanova, 2007). Job resources refer to those physical,

social, or organizational aspects of the job that may: (1) reduce job demands and the

associated physiological and psychological costs; (2) be functional in achieving work goals; or

(3) stimulate personal growth, learning, and development (Bakker & Demerouti, 2007;

Schaufeli & Bakker, 2004).
Job resources either play an intrinsic motivational role because they foster employees’

growth, learning, and development, or they play an extrinsic motivational role because they

are instrumental in achieving work goals. In the former case, job resources fulfil basic human

needs, such as the needs for autonomy, relatedness, and competence (Deci & Ryan, 1985;

Ryan & Frederick, 1997; Van den Broeck, Vansteenkiste, De Witte, & Lens, 2008). For

instance, proper feedback fosters learning, thereby increasing job competence, whereas

decision latitude and social support satisfy the need for autonomy and the need to belong,

respectively. Job resources may also play an extrinsic motivational role, because work

environments that offer many resources foster the willingness to dedicate one’s efforts and

abilities to the work task (Meijman & Mulder, 1998). In such environments it is likely that the

task will be completed successfully and that the work goal will be attained. For instance,

supportive colleagues and performance feedback increase the likelihood of being successful in

achieving one’s work goals. In either case, be it through the satisfaction of basic needs or

through the achievement of work goals, the outcome is positive and engagement is likely to

occur (Schaufeli & Bakker, 2004; Schaufeli & Salanova, 2007).

Consistent with these notions about the motivational role of job resources, several studies
have shown a positive relationship between job resources and work engagement (for a meta-

analysis, see Halbesleben, in press). For example, Schaufeli and Bakker (2004) found evidence

for a positive relationship between three job resources (performance feedback, social support,

and supervisory coaching) and work engagement (vigour, dedication, and absorption) among

four samples of Dutch employees. This study was replicated in a sample of over 2000 Finnish

teachers (Hakanen, Bakker, & Schaufeli, 2006). Results showed that job control, information,

supervisory support, innovative climate, and social climate were all related positively to work

engagement. In addition, Koyuncu, Burke, and Fiksenbaum (2006) examined potential

antecedents and consequences of work engagement in a sample of women managers and

professionals employed by a large Turkish bank. Results showed that of the six areas of work

life (Maslach & Leiter, 1997), particularly job control, rewards and recognition, and value fit

were significant predictors of all three engagement measures.

192 A.B. Bakker et al.

Recent longitudinal research has generally confirmed the positive relationship between

job resources and work engagement. Mauno, Kinnunen, and Ruokolainen (2007) utilized a 2-

year longitudinal design to investigate work engagement and its antecedents among Finnish

health care personnel. Job resources predicted work engagement better than job demands.

Job control and organization-based self-esteem proved to be the best lagged predictors of the

three dimensions of work engagement, after controlling for Time 1 scores on the dimensions

of engagement. Further, in their study among managers and executives of a Dutch telecom

company, Schaufeli, Bakker, and Van Rhenen (2008) found that changes in job resources
were predictive of engagement over a 1-year time period. Specifically, results showed that

increases in social support, autonomy, opportunities to learn and to develop, and

performance feedback were positive predictors of Time 2 work engagement after controlling

for baseline engagement. The two longitudinal studies included in the special issue (de Lange,

De Witte, & Notelaers, 2008; Hakanen, Schaufeli, & Ahola, 2008) offer additional evidence

for a causal effect of job resources on engagement.

Job resources become salient in the face of high job demands

Hobfoll (2002) has argued that resource gain acquires its saliency in the context of resource

loss. This implies that job resources become more salient and gain their motivational

potential when employees are confronted with high job demands (Bakker & Demerouti,

2007). Hakanen, Bakker, and Demerouti (2005) tested this interaction hypothesis in a sample

of Finnish dentists employed in the public sector. It was hypothesized that job resources (e.g.,
variability in the required professional skills and peer contacts) are most beneficial in

maintaining work engagement under conditions of high job demands (e.g., workload,

unfavourable physical environment). The dentists were split in two random groups in order to

cross-validate the findings. A set of hierarchical regression analyses showed that 17 out of 40

possible interactions were statistically significant (43%), showing that variability in

professional skills boosted work engagement when qualitative workload was high, and

mitigated the negative effect of high qualitative workload on work engagement.

Conceptually similar findings have been reported by Bakker, Hakanen, Demerouti, and

Xanthopoulou (2007). In their study among Finnish teachers working in elementary,

secondary, and vocational schools, they found that job resources act as buffers and diminish

the negative relationship between pupil misbehaviour and work engagement. In addition, they

found that job resources particularly influenced work engagement when teachers were

confronted with high levels of pupil misconduct. A series of moderated structural equation

analyses showed that 14 out of 18 possible two-way interaction effects were statistically

significant (78%). Particularly, supervisor support, innovativeness, appreciation, and

organizational climate were important job resources for teachers that helped them cope

with demanding interactions with students. Taken together, these findings clearly show that

job resources may become more salient and gain their motivational potential when employees

are confronted with high job demands.

Personal resources facilitate engagement

In addition to job characteristics, several studies have focused on state-like personal resources

as predictors of work engagement. Personal resources are positive self-evaluations that are

linked to resiliency and refer to individuals’ sense of their ability to control and impact upon

their environment successfully (Hobfoll, Johnson, Ennis, & Jackson, 2003). It has been shown

193 Work & Stress

that such positive self-evaluations predict goal-setting, motivation, performance, job and life

satisfaction, and other desirable outcomes (for a review, see Judge, Van Vianen, & De Pater,

2004). The reason for this is that the larger an individual’s personal resources, the more

positive their self-regard and the more goal self-concordance is expected to be experienced

(Judge, Bono, Erez, & Locke, 2005). Individuals with goal self-concordance are intrinsically

motivated to pursue their goals, and as a result they trigger higher performance and

satisfaction (see also Luthans & Youssef, 2007).
Several authors have investigated the relationships between personal resources and work

engagement. For example, Rothmann and Storm (2003) conducted a cross-sectional study

among 1910 South African police officers, and found that engaged police officers had an

active coping style. They were problem-focused, taking active steps to attempt to remove or

rearrange stressors. Further, in their study among highly skilled Dutch technicians,

Xanthopoulou, Bakker, Demerouti, and Schaufeli (2007) examined the role of three personal

resources (self-efficacy, organizational-based self-esteem, and optimism) in predicting work

engagement. Results showed that engaged employees are highly self-efficacious; they believe

they are able to meet the demands they face in a broad array of contexts. In addition, engaged

workers believe that they will generally experience good outcomes in life (optimistic), and

believe they can satisfy their needs by participating in roles within the organization

(organizational-based self-esteem; see also Mauno et al., 2007).

These findings were replicated and expanded in a 2-year follow-up study (Xanthopoulou,

Bakker, Demerouti, & Schaufeli, 2008). The findings indicated that self-efficacy, organiza-

tional-based self-esteem, and optimism make a unique contribution to explaining variance in

work engagement over time, over and above the impact of job resources and previous levels of
engagement. As a final example, Bakker, Gierveld, and Van Rijswijk (2006) in their study

among female school principals found that those with most personal resources scored highest

on work engagement. Particularly resilience, self-efficacy, and optimism contributed to work

engagement, and were able to explain unique variance in engagement scores (in addition to

social support from team members and colleague principals, opportunities for development,

and social support from the intimate partner). Thus, resilience is another personal resource

that may facilitate work engagement.

In conclusion, job and personal resources are important antecedents of work engagement.

Job resources reduce the impact of job demands on strain, are functional in achieving work

goals, and stimulate personal growth, learning, and development. These resources particu-

larly have motivational potential in the face of high job demands. Further, engaged employees

seem to differ from other employees in terms of their personal resources, including optimism,

self-efficacy, self-esteem, resilience, and an active coping style. These resources seem to help

engaged workers to control and impact upon their work environment successfully (see also

Luthans, Norman, Avolio, & Avey, 2008).

Relationship with performance

Bakker (in press) mentions four reasons why engaged workers perform better than non-

engaged workers. Engaged employees: (1) often experience positive emotions, including

happiness, joy, and enthusiasm; (2) experience better psychological and physical health; (3)

create their own job and personal resources (e.g., support from others); and (4) transfer their

engagement to others. Whereas positive emotions broaden people’s thought-action repertoire

(Fredrickson, 2003), good health facilitates performance because individuals can use all their

mental and physical resources (skills, abilities, knowledge, etc.). Further, employees who

194 A.B. Bakker et al.

create their own resources are better able to deal with their job demands and to achieve their

work goals (Bakker & Demerouti, 2007). Finally, in most organizations performance is the

result of the combined effort of individual employees. It is therefore conceivable that the

crossover of engagement among members of the same work team increases performance.

Only a few studies have examined the relationship between work engagement and job

performance (see Bakker & Demerouti, 2008). Nevertheless, the results obtained so far look

promising. Bakker, Demerouti, and Verbeke (2004) showed that engaged employees received

higher ratings from their colleagues on in-role and extra-role performance, indicating that

engaged employees perform well and are willing to go the extra mile. Further, in their survey

among Dutch employees from a wide range of occupations, Schaufeli, Taris, and Bakker

(2006) found that work engagement is related positively to in-role performance. These

findings were expanded in another study among secretaries; Gierveld and Bakker (2005)

found that engaged secretaries scored higher on in-role and extra-role performance. In

addition, results suggested that engaged secretaries had more influence on daily business.

They were more often asked to carry out additional tasks, including personnel pre-selection,

the organization of trade exhibitions and conventions, and website maintenance.
Salanova, Agut, and Peiró (2005) conducted an important study among personnel

working in Spanish restaurants and hotels. Contact employees from over 100 service units

(hotel front desks and restaurants) provided information about organizational resources,

engagement, and service climate. Furthermore, customers from these units provided

information on employee performance and customer loyalty. Structural equation modelling

analyses were consistent with a full mediation model in which organizational resources and

work engagement predicted service climate, which in turn predicted employee performance

and then customer loyalty. Finally, Xanthopoulou, Bakker, Demerouti, and Schaufeli (in

press) conducted a diary study among employees working in a Greek fast-food restaurant,

and found that daily levels of work engagement were predictive of objective daily financial

returns.

We can conclude that research supports the link between work engagement and

performance. Employees who feel vital and strong, and who are enthusiastic about their

work, show better in-role and extra-role performance. As a consequence, engaged workers

realize better financial results, and have more satisfied clients and customers.

What we don’t know about work engagement: a brief research agenda

Since research on work engagement has just started, there are many questions that still need

to be answered. Below we discuss five topics that seem highly relevant for further progress in

the emerging field of work engagement.

Daily work engagement

Most previous studies on work engagement have used a between-person design and cannot

explain why engaged employees sometimes show below average or poor performance. Even

engaged employees may have their off-days, and researchers have therefore begun to examine

daily changes in work engagement. An important advantage of diary research is that it relies

less on retrospective recall than regular surveys, since the questions relate to individuals’

perceptions and feelings on a certain day. Additionally, daily changes in work engagement

within persons can be related causally to daily changes in performance. Diary research may

195 Work & Stress

also reveal what the day-to-day triggers are of state engagement (Sonnentag, Dormann, &

Demerouti, in press).

Short- vs. long-term consequences of engagement

The available research evidence suggests that work engagement has positive effects in the

short- (Sonnentag, 2003; Xanthopoulou et al., in press) and the long-term (Mauno et al.,

2007; Schaufeli et al., 2008). However, a relevant question is whether there is also a dark side

of engagement. Can the level of engagement be too high if employees are in a continuous

state of high engagement (see Sonnentag, Mojza, Binnewies, & Scholl, 2008)? Engagement

may give new energy, but does it take energy in the end? Can engaged employees finally burn

out? What is the role of recovery in this process? Future studies may use multiple waves with

short and longer periods between the waves of data collection to examine the short- and long-

term consequences of work engagement. There is particularly a dearth of research on the

relationship between engagement and health.

Interventions

Future research on work engagement would benefit from a resolute focus on interventions.

This research would make the most valuable contribution by not only focusing on something

positive, but also working directly on increasing the prevalence of positive relationships with

work. Moreover, from a theoretical perspective it would be interesting and important to test

the hypothesis that fostering engagement goes beyond preventing burnout. A disappointment

of the extensive research on job burnout is the dearth of research that explicitly tests

interventions to alleviate the syndrome (Maslach, Schaufeli, & Leiter, 2001). The research

literature is supplied abundantly with cross-sectional studies demonstrating burnout’s

correlates and too few demonstrating planned change. We urge researchers to go beyond

investigating work engagement’s causes and consequences. The greatest contribution will

come from systematic studies that evaluate the impact of new management procedures or

personal routines on work engagement. Interesting questions are whether engagement can be

trained, and whether the engagement frame facilitates interventions.

Self-regulation

How do people manage their own work engagement? According to self-regulation theories,
individuals use strategies that enable them to guide their goal-directed activities over time and

across changing circumstances. For instance, regulatory focus theory (Higgins, 2000)

discriminates between two regulatory foci among individuals. When promotion focused,

individuals are motivated by growth and development needs, have strong ideals, and prefer

gain to the avoidance of losses. When prevention focused, individuals are responsive to

security needs, the responsibility for safety and protection, have strong emphasis on

obligations, and prefer the avoidance of loss to gains. The regulatory focus that people

apply is a matter of individual differences but can also be influenced by the environment

(Higgins, 2000). It would be interesting to examine the impact of regulatory foci on work

engagement and vice versa. Do work environments that emphasize growth and resources

foster a promotion focus and work engagement? Promotion-focused employees may

successfully balance such environments with themselves using eagerness and approach

strategies (Brockner & Higgins, 2001). In contrast, environments that emphasize duties and

196 A.B. Bakker et al.

demands may foster a prevention focus. Prevention-focused employees may be engaged and

function successfully in such environments using vigilance and avoidance strategies. In short,

future studies could investigate whether engagement is highest when people encounter

regulatory fit between their chronic (preferred focus) and task-induced regulatory state.

Conceptual development and integration

Further progress in research would be more effective if there were broad agreement on the

meaning of work engagement. As noted above, there is broad consensus on two dimensions of

work engagement: energy and involvement/identification, both of which are included in the

OLBI (Demerouti & Bakker, 2008), the MBI (Maslach et al., 1996), and the UWES

(Schaufeli et al., 2002). As far as the dimensions of work engagement are concerned, further
work is needed to consider whether absorption is a core aspect of work engagement or an

outcome of energy and identification, and on the role of professional efficacy. Resolving these

questions requires further development in theory and measurement. Based on theoretical

analyses and research on the construct and concurrent validity, relationships between

engagement and other established constructs such as satisfaction, organizational commit-

ment, and organizational citizenship behaviour should be specified. In this way, the added

value of the recently emerged concept of work engagement should be demonstrated.

Conclusion

In this paper we have shown that research on work engagement may broaden our view of the
meaning and effects of working (see Turner et al., 2002). Employees with high levels of energy

and identification with their work have many resources available and seem to perform better.

It is even conceivable that engaged workers create their own job resources over time. Our

overview suggests that a focus on work engagement may not only benefit the individual but

also offer organizations a competitive advantage. We hope that the research agenda that we

have outlined above will be a useful resource for occupational health psychologists and will

stimulate future research on work engagement.

References

Bakker, A.B. (in press). Building engagement in the workplace. In C. Cooper & R. Burke (Eds.), The

peak performing organization. London: Routledge.

Bakker, A.B. & Demerouti, E. (2007). The job demands-resources model: State of the art. Journal of

Managerial Psychology, 22, 309�328.

Bakker, A.B. & Demerouti, E. (2008). Towards a model of work engagement. Career Development

International, 13, 209�223.

Bakker, A.B. & Schaufeli, W.B. (2008). Positive organizational behavior: Engaged employees in

flourishing organizations. Journal of Organizational Behavior, 29, 147�154.

Bakker, A.B., Demerouti, E., & Burke, R. (in press). Workaholism and relationship quality: A spillover-

crossover perspective. Journal of Occupational Health Psychology.

Bakker, A.B., Demerouti, E. & Verbeke, W. (2004). Using the Job Demands-Resources model to predict

burnout and performance. Human Resource Management, 43, 83�104.

Bakker, A.B., Gierveld, J.H., & Van Rijswijk, K. (2006). Succesfactoren bij vrouwelijke schoolleiders in

het primair onderwijs: Een onderzoek naar burnout, bevlogenheid en prestaties [Success factors among

female school principals in primary teaching: A study on burnout, work engagement, and

performance]. Diemen, The Netherlands: Right Management Consultants.

197 Work & Stress

Bakker, A.B., Hakanen, J.J., Demerouti, E. & Xanthopoulou, D. (2007). Job resources boost work

engagement particularly when job demands are high. Journal of Educational Psychology, 99, 274�284.

Brockner, J. & Higgins, E.T. (2001). Regulatory focus theory: Implications for the study of emotions at

work. Organizational Behavior and Human Decision Processes, 86, 35�66.

de Lange, A.H., De Witte, H., & Notelaers, G. (2008). Should I stay or should I go? Examining

longitudinal relations among job resources and work engagement for stayers versus movers. Work &

Stress, 22, 201�223.

Deci, W.L. & Ryan, R.M. (1985). Intrinsic motivation and self-determination in human behavior. New

York: Plenum.

Demerouti, E. & Bakker, A.B. (2008). The Oldenburg Burnout Inventory: A good alternative to

measure burnout and engagement. In J.R.B. Halbesleben (Ed.), Handbook of stress and burnout in

health care. Hauppauge, NY: Nova Science.

Demerouti, E., & Cropanzano, R. (in press). Work engagement and performance. In A.B. Bakker &

M.P. Leiter (Eds.), Work engagement: Recent developments in theory and research. New York:

Psychology Press.

Demerouti, E., Bakker, A.B., Nachreiner, F. & Ebbinghaus, M. (2002). From mental strain to burnout.

European Journal of Work and Organizational Psychology, 11, 423�441.

Fredrickson, B.L. (2003). Positive emotions and upward spirals in organizations. In K. Cameron, J.

Dutton & R. Quinn (Eds.), Positive organizational scholarship (pp. 163�175). San Francisco: Berrett-

Koehler.

Gierveld, J.H., & Bakker, A.B. (2005). De invloed van de secretaresse [The influence of the secretary].

Diemen, The Netherlands: Manpower.

González-Roma, V., Schaufeli, W.B., Bakker, A.B. & Lloret, S. (2006). Burnout and engagement:

Independent factors or opposite poles? Journal of Vocational Behavior, 68, 165�174.

Hakanen, J. (2002). Työuupumuksesta työn imuun*positiivisen työhyvinvointikäsitteen ja-menetelmän

suomalaisen version validointi opetusalan organisaatiossa [From burnout to job engagement*
validation of the Finnish version of an instrument for measuring job engagement (UWES) in an

educational organization]. Työ ja Ihminen, 16, 42�58.

Hakanen, J.J., Bakker, A.B. & Demerouti, E. (2005). How dentists cope with their job demands and stay

engaged: The moderating role of job resources. European Journal of Oral Sciences, 113, 479�487.

Hakanen, J., Bakker, A.B. & Schaufeli, W.B. (2006). Burnout and work engagement among teachers.

Journal of School Psychology, 43, 495�513.

Hakanen, J.J., Schaufeli, W.B., & Ahola, K. (2008). The Job Demands-Resources model: A three-year

cross-lagged study of burnout, depression, commitment, and work engagement. Work & Stress, 22,

224�241.

Halbesleben, J.R.B. (in press). A meta-analysis of work engagement: Relationships with burnout,

demands, resources and consequences. In A.B. Bakker & M.P. Leiter (Eds.), Work engagement:

Recent developments in theory and research. New York: Psychology Press.

Halbesleben, J.R.B., & Wheeler, A.R. (2008). The relative roles of engagement and embeddedness in

predicting job performance and intention to leave. Work & Stress, 22, 242�256.

Hallberg, U. & Schaufeli, W.B. (2006). ‘‘Same same’’ but different? Can work engagement be

discriminated from job involvement and organizational commitment? European Psychologist, 11,

119�127.

Hallberg, U.E., Johansson, G. & Schaufeli, W.B. (2007). Type A behavior and work situation:

Associations with burnout and work engagement. Scandinavian Journal of Psychology, 48, 135�142.

Higgins, E.T. (2000). Making a good decision: Value from fit. American Psychologist, 55, 1217�1230.

Hobfoll, S.E. (2002). Social and psychological resources and adaptation. Review of General Psychology,

6, 307�324.

Hobfoll, S.E., Johnson, R.J., Ennis, N. & Jackson, A.P. (2003). Resource loss, resource gain, and

emotional outcomes among inner city women. Journal of Personality and Social Psychology, 84,

632�643.

198 A.B. Bakker et al.

Judge, T.A., Bono, J.E., Erez, A. & Locke, E.A. (2005). Core self-evaluations and job and life

satisfaction: The role of self-concordance and goal attainment. Journal of Applied Psychology, 90,

257�268.

Judge, T.A., Van Vianen, A.E.M. & De Pater, I. (2004). Emotional stability, core self-evaluations, and

job outcomes: A review of the evidence and an agenda for future research. Human Performance, 17,

325�346.

Kahn, W.A. (1990). Psychological conditions of personal engagement and disengagement at work.

Academy of Management Journal, 33, 692�724.

Kahn, W.A. (1992). To be fully there: Psychological presence at work. Human Relations, 45, 321�349.

Koyuncu, M., Burke, R.J. & Fiksenbaum, L. (2006). Work engagement among women managers and

professionals in a Turkish bank: Potential antecedents and consequences. Equal Opportunities

International, 25, 299�310.

Luthans, F. & Youssef, C.M. (2007). Emerging positive organizational behavior. Journal of Manage-

ment, 33, 321�349.

Luthans, F., Norman, S.M., Avolio, B.J. & Avey, J.B. (2008). The mediating role of psychological capital

in the supportive organizational climate*employee performance relationship. Journal of Organiza-

tional Behavior, 29, 219�238.

Macey, W.H. & Schneider, B. (2008). The meaning of employee engagement. Industrial and

Organizational Psychology, 1, 3�30.

Maslach, C. & Leiter, M.P. (1997). The truth about burnout: How organizations cause personal stress and

what to do about it. San Francisco, CA: Jossey-Bass.

Maslach, C. & Leiter, M.P. (2008). Early predictors of job burnout and engagement. Journal of Applied

Psychology, 93, 498�512.

Maslach, C., Jackson, S.E. & Leiter, M. (1996). Maslach Burnout Inventory manual, 3rd ed. Palo Alto,

CA: Consulting Psychologists Press.

Maslach, C., Schaufeli, W.B. & Leiter, M.P. (2001). Job burnout. Annual Review of Psychology, 52,

397�422.

Mauno, S., Kinnunen, U. & Ruokolainen, M. (2007). Job demands and resources as antecedents of

work engagement: A longitudinal study. Journal of Vocational Behavior, 70, 149�171.

May, D.R., Gilson, R.L. & Harter, L.M. (2004). The psychological conditions of meaningfulness, safety

and availability and the engagement of the human spirit at work. Journal of Occupational and

Organizational Psychology, 77, 11�37.

Meijman, T.F., & Mulder, G. (1998). Psychological aspects of workload. In P.J.D. Drenth & H. Thierry

(Eds.), (Eds.), Handbook of work and organizational psychology, Vol. 2: Work psychology (pp. 5�33).

Hove: Psychology Press.

Myers, D.G. (2000). The funds, friends and faith of happy people. American Psychologist, 55, 56�67.

Rothbard, N.P. (2001). Enriching or depleting? The dynamics of engagement in work and family roles.

Administrative Science Quarterly, 46, 655�684.

Rothmann, S., & Storm, K. (2003, May). Work engagement in the South African Police Service. Paper

presented at the 11th European Congress of Work and Organizational Psychology, Lisbon, Portugal.

Ryan, R.M. & Frederick, C.M. (1997). On energy, personality, and health: Subjective vitality as a

dynamic reflection of well-being. Journal of Personality, 65, 529�565.

Salanova, M., Agut, S. & Peiró, J.M. (2005). Linking organizational resources and work engagement to

employee performance and customer loyalty: The mediation of service climate. Journal of Applied

Psychology, 90, 1217�1227.

Schaufeli, W.B., & Bakker, A.B. (2003). UWES*Utrecht Work Engagement Scale: test manual.

Unpublished manuscript, Department of Psychology, Utrecht University.

Schaufeli, W.B. & Bakker, A.B. (2004). Job demands, job resources, and their relationship with burnout

and engagement: A multi-sample study. Journal of Organizational Behavior, 25, 293�315.

Schaufeli, W.B., & Bakker, A.B. (in press). The conceptualization and measurement of work

engagement: A review. In A.B. Bakker & M.P. Leiter (Eds.), Work engagement: Recent developments

in theory and research. New York: Psychology Press.

199 Work & Stress

Schaufeli, W.B. & Salanova, M. (2007). Work engagement: An emerging psychological concept and its

implications for organizations. In S.W. Gilliland, D.D. Steiner & D.P. Skarlicki (Eds.), Research in

social issues in management: Vol. 5. Managing social and ethical issues in organizations. Greenwich,

CT: Information Age Publishers.

Schaufeli, W.B., Bakker, A.B. & Salanova, M. (2006). The measurement of work engagement with a

short questionnaire: A cross-national study. Educational and Psychological Measurement, 66,

701�716.

Schaufeli, W.B., Bakker, A.B., & Van Rhenen, W. (2008). How changes in job demands and resources

predict burnout, work engagement, and sickness absenteeism. Manuscript submitted for publication.

Schaufeli, W.B., Salanova, M., González-Romá, V. & Bakker, A.B. (2002). The measurement of

engagement and burnout: A two sample confirmatory factor analytic approach. Journal of Happiness

Studies, 3, 71�92.

Schaufeli, W.B., Taris, T.W. & Bakker, A.B. (2006). Dr. Jekyll and Mr. Hyde: On the differences between

work engagement and workaholism. In R.J. Burke (Ed.), Research companion to working time and

work addiction (pp. 193�217). Northampton: Edward Elgar.

Schaufeli, W.B., Taris, T.W., Le Blanc, P., Peeters, M., Bakker, A.B. & De Jonge, J. (2001). Maakt arbeid

gezond? Op zoek naar de bevlogen werknemer [Does work make happy? In search of the engaged

worker]. De Psycholoog, 36, 422�428.

Scott, K.S., Moore, K.S. & Miceli, M.P. (1997). An exploration of the meaning and consequences of

workaholism. Human Relations, 50, 287�314.

Seligman, M.E.P. & Csikszentmihalyi, M. (2000). Positive psychology: An introduction. American

Psychologist, 55, 5�14.

Shimazu, A., Schaufeli, W.B., Kosugi, S., Suzuki, A., Nashiwa, H., Kato, A., et al. (2008). Work

engagement in Japan: Development and validation of the Japanese version of the Utrecht Work

Engagement Scale. Applied Psychology: An International Review, 57, 510�523.

Sonnentag, S. (2003). Recovery, work engagement, and proactive behavior: A new look at the interface

between non-work and work. Journal of Applied Psychology, 88, 518�528.

Sonnentag, S., Dormann, C., & Demerouti, E. (in press). Not all days are created equal: The concept of

state work engagement. In A.B. Bakker & M.P. Leiter (Eds.), Work engagement: Recent developments

in theory and research. New York: Psychology Press.

Sonnentag, S., Mojza, E.J., Binnewies, C., & Scholl, A. (2008). Being engaged at work and detached at

home: A week-level study on work engagement, psychological detachment, and affect. Work & Stress,

22, 257�276.

Storm, K. & Rothmann, I. (2003). A psychometric analysis of the Utrecht Work Engagement Scale in

the South African police service. South African Journal of Industrial Psychology, 29, 62�70.

Taris, T.W., Schaufeli, W.B., & Shimazu, A. (in press). Workaholism versus engagement: What’s the

difference? In A.B. Bakker & M.P. Leiter (Eds.), Work engagement: Recent developments in theory and

research. New York: Psychology Press.

Turner, N., Barling, J., & Zacharatos, A. (2002). Positive psychology at work. In C. Snyder & S. Lopez

(Eds.), The handbook of positive psychology (pp. 715�730). Oxford: Oxford University Press.

Van den Broeck, A., Vansteenkiste, M., De Witte, H., & Lens, W., (2008). Explaining the relationships

between job characteristics, burnout, and engagement: The role of basic psychological need

satisfaction. Work & Stress, 22, 277�294.

Xanthopoulou, D., Bakker, A.B., Demerouti, E. & Schaufeli, W.B. (2007). The role of personal

resources in the job demands-resources model. International Journal of Stress Management, 14,

121�141.

Xanthopoulou, D., Bakker, A.B., Demerouti, E., & Schaufeli, W.B. (2008). Work engagement: A cycle of

job and personal resources. Manuscript submitted for publication.

200 A.B. Bakker et al.

Xanthopoulou, D., Bakker, A.B., Demerouti, E., & Kantas, A. (in press). The measurement of burnout

and engagement: A cross-cultural study comparing Greece and the Netherlands. New Review of

Social Psychology.

Xanthopoulou, D., Bakker, A.B., Demerouti, E., & Schaufeli, W.B. (in press). Work engagement and

financial returns: A diary study on the role of job and personal resources. Journal of Occupational and

Organizational Psychology.

Yi-Wen, Z. & Yi-Qun, C. (2005). The Chinese version of the Utrecht Work Engagement Scale: An

examination of reliability and validity. Chinese Journal of Clinical Psychology, 13, 268�270.