The Relationship of Emotional Exhaustion to Work Attitudes, Job
Performance, and Organizational Citizenship Behaviors
Colorado State University
Deborah E. Rupp
University of Illinois at Urbana–Champaign
Zinta S. Byrne
Personnel Decisions International
The authors investigated the negative consequences of emotional exhaustion for individual employees
and their employers. On the basis of social exchange theory, the authors proposed that emotional
exhaustion would predict job performance, 2 classes of organizational citizenship behavior, and turnover
intentions. In addition, the authors posited that the relationship between emotional exhaustion and
effective work behaviors would be mediated by organizational commitment. With only a few exceptions,
the results of 2 field studies supported the authors’ expectations. In addition, emotional exhaustion
exerted an independent effect on these criterion variables beyond the impact of age, gender, and ethnicity.
Research has linked emotional exhaustion to a plethora of
ailments, including physiological problems, depression, family
difficulties, and a general breakdown in feelings of community
(Kahill, 1988; Maslach & Leiter, 1997). By themselves, these
human concerns provide ample reason to study emotional exhaus-
tion, but there are additional justifications. A growing body of
research has begun to demonstrate that emotional exhaustion can
have deleterious consequences for organizations as well. For ex-
ample, exhausted workers manifest lower levels of commitment
and a greater likelihood of seeking employment elsewhere (Lee &
Ashforth, 1996; Wright & Cropanzano, 1998). With these issues in
mind, the present studies were undertaken to investigate the rela-
tionship of emotional exhaustion to organizationally relevant cri-
teria. We had two principal objectives: First, we sought to
ascertain whether there is a relationship between emotional
exhaustion and effective work behaviors (job performance and
organizational citizenship behavior). Second, on the basis of
social exchange theory, we proposed that the impact of emo-
tional exhaustion on work behavior is mediated by organiza-
tional commitment. The two field studies reported here provide
an initial test of this model.
Understanding Emotional Exhaustion
The Components of Burnout
Historically, research on emotional exhaustion emerged from
Maslach’s (1982) influential model of burnout. In Maslach’s orig-
inal framework, burnout had three parts. The first component and
the topic of the present investigation, emotional exhaustion,isa
chronic state of emotional and physical depletion. As Demerouti,
Bakker, Nachreiner, and Schaufeli (2001, p. 499) suggested:
“Emotional exhaustion closely resembles traditional stress reac-
tions that are studied in occupational stress research, such as
fatigue, job-related depression, psychosomatic complaints, and
anxiety.” Given these observations, it is reasonable to conceptu-
alize emotional exhaustion as a type of strain that results from
workplace stressors. The second component of the model, deper-
sonalization, is a type of interpersonal distancing and lack of
connectedness with one’s coworkers and clients. The third com-
ponent, diminished personal accomplishment, refers to a negative
evaluation of the self.
Subsequent to the formulation of the original three-part model,
emotional exhaustion has emerged as a central variable for under-
standing the burnout process (Baba, Jamal, & Tourigny, 1998;
Cordes & Dougherty, 1993; Gaines & Jerimer, 1983; Zohar,
1997). The reasons for this are both empirical and conceptual.
Empirically, some work has suggested that emotional exhaustion
exhibits somewhat stronger relationships than do the other com-
ponents to important outcome variables (Lee & Ashforth, 1993,
1996; Wright & Bonett, 1997). Conceptually, Shirom (1989) ar-
gued that emotional exhaustion best captures the “core meaning”
of burnout (cf., Pines & Aronson, 1988). Moreover, emphasizing
emotional exhaustion has allowed scholars to more clearly distin-
guish burnout from related concepts, such as self-efficacy and
self-esteem (Shirom, 1989). In keeping with these empirical find-
ings and conceptual frameworks, we explored the relationship of
Russell Cropanzano, Psychology Department, Colorado State Univer-
sity; Deborah E. Rupp, Department of Psychology and the Institute for
Labor and Industrial Relations, University of Illinois at Urbana–Cham-
paign; Zinta S. Byrne, Personnel Decisions International, Fort Collins,
Correspondence concerning this article should be addressed to Russell
Cropanzano, who is now at Department of Management and Policy, Eller
College of Business and Public Administration, University of Arizona,
McClelland Hall, Room 450U, P.O. Box 210108, Tucson, Arizona 85721-
0108. E-mail: email@example.com
Journal of Applied Psychology Copyright 2003 by the American Psychological Association, Inc.
2003, Vol. 88, No. 1, 160–169 0021-9010/03/$12.00 DOI: 10.1037/0021-9010.88.1.160
emotional exhaustion to important work behaviors, attitudes, and
The Effects of Emotional Exhaustion
A variety of conceptual frameworks have been used to under-
stand the effects of burnout (e.g., Freudenberger, 1983; Lee &
Ashforth, 1990, 1993, 1996; Shirom, 1989). These theoretical
models have tended to focus on the consequences of emotional
exhaustion for individual workers and their families (Kahill, 1988;
Lee & Ashforth, 1996) and have paid relatively less attention to
emotional exhaustion’s effects on organizationally relevant crite-
ria, such as job performance and citizenship behavior (Wright &
Bonett, 1997). Because an organizationally focused model of
burnout may be quite relevant for a more complete understanding
of effective work behaviors and attitudes, as well as for the
promotion of more humane work organizations, we considered the
implications of social exchange theory for emotional exhaustion.
Understanding Social Exchange Theory
Social exchange theory states that employees form relationships
at work. Although there are many variants of social exchange
theory (for a review see Cropanzano, Rupp, Mohler, & Schminke,
2001), modern versions of this framework tend to describe two
types of interpersonal relationships (cf. Blau, 1964; Organ, 1988,
1990). Economic exchange relationships are more short term.
They involve the exchange of relatively concrete, often economic
benefits that are exchanged in a quid pro quo fashion. These types
of relationships are quite different from social exchange relation-
ships, which are more important to our present purposes. Social
exchange relationships tend to involve the exchange of socioemo-
tional benefits. They are associated with close personal attach-
ments and open-ended obligations. When individuals form social
exchange relationships with organizations, they tend to have
higher job performance, more organizational citizenship behaviors
(OCB), and weaker turnover intentions (e.g., Hendrix, Robbins,
Miller, & Summers, 1998; Wayne, Shore, & Liden, 1997).
Because social exchange relationships emphasize the obliga-
tions, attachments, and identification that employees feel toward
their employers, past research has proposed using organizational
commitment to operationalize an employee’s social exchange re-
lationship with his or her employing organization (Bishop & Scott,
2000; Bishop, Scott, & Burroughs, 2000; Cropanzano & Byrne,
2000; Cropanzano, Howes, Grandey, & Toth, 1997; Deckop, Man-
gel, & Cirka, 1999; Randall, Cropanzano, Bormann, & Birjulin,
1999). The model tested in our studies follows this logic.
Generally speaking, research suggests that individuals form
social exchange relationships to the extent that they receive worth-
while benefits and that these benefits are assigned in a fair manner
(Cropanzano et al., 2001). Jobs that produce emotional exhaustion
are likely to violate both of these conditions. First, emotional
exhaustion can be seen as a cost that qualifies the value of any
benefits received through employment. Second, employees are apt
to resent an organization that overworks them to the point of
emotional exhaustion, causing them to perceive the organization’s
actions as unfair. Emotional exhaustion, because it is personally
costly and often seen as unjustified, should impede the develop-
ment of high quality, social exchange relationships (which would
be manifested through lowered organizational commitment). Evi-
dence presented by social exchange theorists suggests that the
absence of a social exchange relationship should engender higher
turnover intentions, lower job performance, less OCB benefical to
organizations (OCBO), and less OCB beneficial to one’s supervi-
sor (OCBS; for evidence, see Konovsky & Pugh, 1994; Moorman,
Blakely, & Niehoff, 1998; Rupp & Cropanzano, 2002, Settoon,
Bennett, & Liden, 1996). This possibility is illustrated in Figure 1.
Although the empirical literature is not complete, as we discuss in
the next section, it is generally consistent with the model displayed
in Figure 1.
Figure 1. Theoretical model linking emotional exhaustion to effective work behaviors by means of organiza-
tional commitment. OCB ⫽ organizational citizenship behavior.
Emotional Exhaustion and Organizational Commitment
Several studies have found that commitment is (negatively)
associated with emotional exhaustion (e.g., Jackson, Turner, &
Brief, 1987; Leiter & Maslach, 1988). Indeed, a meta-analytic
examination of seven studies by Lee and Ashforth (1996) obtained
a corrected r ⫽⫺.43 between organizational commitment and
burnout. These findings are consistent with predictions of social
exchange theorists who propose commitment as a useful measure
of social exchange. Thus, we proposed the following hypothesis:
Hypothesis 1: Emotional exhaustion will be negatively re-
lated to organizational commitment.
Emotional Exhaustion and Turnover
Other researchers have found that emotionally exhausted work-
ers are likely to withdraw from the work environment (Westman &
Eden, 1997). In work settings, this withdrawal can manifest itself
by turnover. This effect has been amply demonstrated, and the
present study seeks only to replicate it. For example, in a meta-
analytic review, Lee and Ashforth (1996) found that emotional
exhaustion and turnover intentions were correlated at .44. Other
researchers have found that emotional exhaustion is related to
actual turnover as well (Wright & Cropanzano, 1998). Practical
considerations prevented us from collecting actual turnover data;
thus, we limited ourselves to turnover intentions as follows:
Hypothesis 2: Emotional exhaustion will be positively related
to turnover intentions.
Emotional Exhaustion and Job Performance
Although there is some evidence linking emotional exhaustion
to performance, it is limited. For example, in a 2 year longitudinal
study, Wright and Bonett (1997) found that emotional exhaustion
at Time 1 predicted job performance at Time 2. Moreover, this
effect remained significant even after Time 1 job performance was
taken into account. Jones and Best (1995); Leiter, Harvie, and
Frizzell (1998); Nowack and Hanson (1983); Quattrochi-Turbin,
Jones, and Breedlove (1983); and Wright and Cropanzano (1998)
also presented data attesting to a negative relationship between
emotional exhaustion and job performance. From this we hypoth-
esized the following:
Hypothesis 3: Emotional exhaustion will be negatively re-
lated to job performance.
Emotional Exhaustion and Organizational Citizenship
We were unable to locate any studies that examined the rela-
tionship between OCB and emotional exhaustion. Nevertheless,
social exchange theory specifies that disrupted working relation-
ships should engender less OCB directed toward the organization
and the supervisor (e.g., Cropanzano et al., 2001; Organ, 1990).
Given these theoretical considerations, we advanced the following
Hypothesis 4: Emotional exhaustion will be (a) negatively
related to OCBO and (b) negatively related to OCBS.
Organizational Commitment as a Mediator
As articulated in previous work (e.g., Organ, 1988, 1990), social
exchange theory suggests that one’s relationship with an employer
provides a proximal cause for work behaviors and turnover inten-
tions. As shown in Figure 1, emotional exhaustion tends to dimin-
ish the quality of these relationships. Hence, it follows that orga-
nizational commitment should mediate the association between
emotional exhaustion and our criterion variables. Unfortunately,
we could locate no evidence bearing on this possibility. However,
keeping in mind the considerable work that supports social ex-
change theory in other contexts (e.g., Cropanzano et al., 2001), we
thought it appropriate to test this model in the two studies reported
herein by hypothesizing the following:
Hypothesis 5: Organizational commitment will at least par-
tially mediate the relationship between emotional exhaustion
and (a) turnover intentions, (b) job performance ratings, (c)
OCBO, and (d) OCBS.
Demographic Control Variables
Finally, Wright and Bonett (1997) underscored the fact that
different demographic groups may be differentially impacted by
variables such as emotional exhaustion. For this reason, we con-
trolled for age, gender, and ethnicity in all of our hypotheses tests.
Study 1: An Initial Test of the Model
Study 1 provides the first examination of the model diagrammed
in Figure 1. To our knowledge, the overall model has never been
tested. Fortunately, there is evidence for some of the proposed
paths. As described above, the link between emotional exhaustion
and commitment seems well established. There is only limited
evidence supporting the emotional exhaustion–job performance
path, although the available work does seem to favor our model.
Finally, we could locate no tests of the emotional exhaustion–OCB
path, so we assessed two forms of OCB in both studies.
Participants and procedures. Participants were subordinates and su-
pervisors working in a large hospital in the western United States. Nine
hundred fifteen surveys were distributed to hospital employees during the
administration of another hospital-wide survey. Two hundred four surveys
were returned by using this method (obtaining an overall response rate of
22%). Of the 204 employee surveys returned, supervisors supplied OCB
and performance ratings for 150 subordinates.
One hundred sixty-seven participants were female. The average age of
respondents was about 42 years. About 89.7% (n ⫽ 183) of respondents
reported their ethnicity as non-Hispanic White, 4.5% (n ⫽ 9) reported
Asian American, 1% (n ⫽ 2) reported African American, 1% (n ⫽ 2)
reported Latino(a), and 2.5% (n ⫽ 5) reported Other. Approximately 1.5%
(n ⫽ 3) did not complete this item on the questionnaire. Given the small
number of individuals endorsing most of the ethnic groups, we followed
the strategy used by Cropanzano, Prehar, and Chen (2002) and coded
ethnicity into two categories: (a) Individuals other than non-Hispanic
Whites and (b) non-Hispanic Whites.
To ensure the employees’ confidentiality, surveys were returned directly
to the researchers, who established collection times in the hospital’s caf-
eteria. In addition, a coding scheme was developed by which the research-
ers would know who completed a survey so that (a) the employee’s
supervisor could be given a questionnaire on that employee, and (b) the
employee-reported and supervisor-reported data could be linked in the
database. Once the supervisors were given their questionnaires, the coding
sheet of names was destroyed.
Measures. Employees completed scales measuring emotional exhaus-
tion, organizational commitment, and turnover intentions. Emotional ex-
haustion was assessed with nine items of Maslach and Jackson’s (1981)
Emotional Exhaustion Scale. Responses were measured on the same
7-point Likert scale ranging from 1 (strongly disagree)to7(strongly
agree). Organizational commitment was measured by using the eight items
that compose Allen and Meyer’s (1990) Affective Commitment Scale.
These items were measured by using the same 1–7 response scale as the
emotional exhaustion measure. Turnover intentions were examined by
using the 3-item scale developed by Konovsky and Cropanzano (1991).
Further validation evidence for this scale can be found in Grandey and
Cropanzano (1999) and Randall et al. (1999).
Supervisors rated their subordinates on OCBS, OCBO, and job perfor-
mance. Ratings of OCBO were assessed by using Williams and Anderson’s
(1991) OCBO scale. OCBS was assessed by use of a 5-item scale devised
and validated by Malatesta (1995). These items were derived from those
presented in Williams and Anderson’s measures and were modified to
reflect behaviors that are specifically beneficial to the supervisor by in-
serting “you” as the item focus. For example, one question asked “Helps
you when you have a heavy work load” instead of “Helps others who have
heavy work loads.” Finally, to assess job performance each supervisor
completed Williams and Anderson’s measure of in-role behaviors. Wil-
liams and Anderson have shown that their In-Role Behavior scale measures
a construct distinct from OCB.
Descriptive statistics and correlations. The descriptive statis-
tics and a correlation matrix are displayed in Table 1. As the reader
will observe, all of the measures have acceptable levels of reli-
ability. Additionally, the zero-order correlations are generally con-
sistent with our hypotheses. However, the relationship between
emotional exhaustion and OCBS is quite small.
Mediation analyses. We predicted that organizational commit-
ment would mediate the relationship between emotional exhaus-
tion and the criterion variables. As Figure 1 illustrates, this implies
a significant indirect relationship between emotional exhaustion
and the outcome measures (James & Brett, 1984). Baron and
Kenny (1986) maintained that mediation can be demonstrated by
three regression tests. We will describe these three tests by using
the language of our particular studies. First, emotional exhaustion
(our predictor) must be related to organizational commitment (our
mediator). Second, emotional exhaustion must be related to each
outcome measure. Third, when both emotional exhaustion and
commitment are simultaneously included in a regression equation,
then the relationship between emotional exhaustion and the crite-
rion variables must be appreciably smaller than it is when emo-
tional exhaustion is the sole predictor. Findings from these tests are
Although these three tests are essential, Holmbeck (2002) ar-
gued that they are insufficient. What is needed is a method of
ascertaining whether the indirect path between the predictor (i.e.,
emotional exhaustion by means of commitment) and the criterion
is significant. Because the total variance accounted for by emo-
tional exhaustion is the sum of the direct and indirect relationships,
a significant indirect path necessitates a drop in the total predictor–
criterion relationship when the mediator is included. Conse-
quently, Holmbeck recommended a direct test of the indirect path
(i.e., the impact of emotional exhaustion by means of commit-
ment), removing the variance as a result of the direct effect.
Formulas for conducting this significance test have been presented
by Holmbeck (2002), Baron and Kenny (1986), and Sobel (1982,
1988). These formulas yield a z score, which can then be compared
with the a priori critical value (z ⫽ 1.645 for a one-tailed test when
p ⬍ .05, and z ⫽ 2.326 when p ⬍ .01).
Emotional exhaustion and organizational commitment. Given
these considerations, the first step in testing the model depicted in
Figure 1 is to demonstrate that emotional exhaustion is a useful
predictor of organizational commitment. Hypothesis 1 was tested
by first entering the control variables into the equation in Step 1.
Emotional exhaustion followed in Step 2. As displayed in Table 2,
emotional exhaustion significantly predicted organizational com-
Study 1: Means, Standard Deviations, Internal Consistency Reliabilities, and Intercorrelations
Variable MSD 123456789
1. Emotional exhaustion 2.97 1.43 .92
2. Age 41.67 11.51 ⫺.02 —
3. Gender ⫺.02 .04 —
4. Ethnicity ⫺.09 .06 .18** —
5. Commitment 4.58 1.12 ⫺.46** .12 .19** .14* .79
6. Turnover intentions 2.81 1.69 .57** ⫺.15* ⫺.02 ⫺.07 ⫺.58** .82
7. Job performance 4.40 0.56 ⫺.26** .08 .14* .26** .35** ⫺.25** .88
8. OCBO 3.81 0.62 ⫺.25** .09 .03 .20** .32** ⫺.28** .59** .79
9. OCBS 2.87 1.12 ⫺.14* .13 ⫺.03 .03 .23** ⫺.18* .24** .59** .89
Note. Gender was coded 1 ⫽ male, 2 ⫽ female. Ethnicity was coded 1 ⫽ non-White, 2 ⫽ White. Internal consistency reliabilities appear along the
diagonal. OCBO ⫽ organizational citizenship behaviors beneficial to organizations; OCBS ⫽ organizational citizenship behaviors beneficial to one’s
* p ⬍ .05, one-tailed. ** p ⬍ .01, one-tailed.
mitment even beyond the effect of the demographic variables,
⫽ .20, p ⬍ .01. The overall model was significant as well,
F(4, 197) ⫽ 17.43, p ⬍ .01. Consequently, we achieved the first
criteria outlined by Baron and Kenny (1986).
Emotional exhaustion and turnover intentions. Hypothesis 2
predicted that emotional exhaustion would be related to turnover
intentions, beyond the effect of the demographic controls. As for
commitment, we analyzed the data in two steps—first entering the
control variables and then entering emotional exhaustion. In Ta-
ble 2 we present these findings. Emotional exhaustion accounted
for 31.7% of the variance beyond the effects of age, gender, and
ethnicity. This was, of course, significant. The full model was
significant as well, R
⫽ .36, F(4, 197) ⫽ 26.94, p ⬍ .01.
Table 3 displays the mediation tests. This test of Hypothesis 5a
was conducted in three steps. We excluded Step 1 from this table
because it duplicates Step 1 in Table 2. The demographic controls
were entered first, commitment was entered second, and emotional
exhaustion was entered in Step 3. The regression coefficient for
emotional exhaustion was significant, although smaller in size than
it was when commitment was excluded. Following from Holm-
beck (2002), we computed the significance of the indirect path
(i.e., the impact emotional exhaustion by means of commitment).
Study 1: Regression Results for Hypotheses 1–5
commitment Turnover intentions
ratings OCBO OCBS
Age 0.01 0.01 0.13* ⫺0.03 0.01 ⫺0.19** 0.00 0.01 0.05 0.01 0.01 0.07 0.02 0.01 0.14*
Gender 0.53 0.21 0.18** 0.01 0.32 0.00 0.09 0.13 0.06 ⫺0.07 0.14 ⫺0.04 ⫺0.13 0.26 ⫺0.04
Ethnicity 0.41 0.27 0.11 ⫺0.38 0.42 ⫺0.67 0.46 0.16 0.24** 0.46 0.18 0.22** 0.12 0.33 0.03
Age .01 0.01 0.10 0.02 0.01 ⫺0.16** 0.00 0.01 0.01 0.00 0.01 0.03 0.01 0.01 0.12
Gender 0.54 0.19 0.18** ⫺.01 0.26 ⫺0.00 0.12 0.13 0.08 ⫺0.04 0.14 ⫺0.02 ⫺0.10 0.26 ⫺0.03
Ethnicity 0.26 0.24 0.07 ⫺0.09 0.34 ⫺0.02 0.41 0.16 0.22** 0.41 0.18 0.20** 0.08 0.33 0.02
Emotional exhaustion ⫺0.35 0.05 ⫺0.45** 0.67 0.07 0.57** ⫺0.07 0.03 ⫺0.19** ⫺0.08 0.04 ⫺0.19** ⫺0.07 0.07 ⫺0.09
Note. OCBO ⫽ organizational citizenship behaviors beneficial to organizations; OCBS ⫽ organizational citizenship behaviors beneficial to one’s
* p ⬍ .05, one-tailed. ** p ⬍ .01, one-tailed.
Study 1: Mediation Tests
Turnover intentions Job performance ratings OCBO OCBS
Age ⫺0.02 0.01 ⫺0.12* 0.00 0.01 0.00 0.00 0.01 0.03 0.01 0.01 0.11
Gender 0.48 0.26 0.11* 0.04 0.12 0.03 ⫺0.12 0.14 ⫺0.07 ⫺0.20 0.26 ⫺0.07
Ethnicity 0.02 0.34 ⫺0.00 0.37 0.16 0.19** 0.36 0.17 0.18* 0.00 0.33 0.00
Organizational commitment ⫺0.88 0.09 ⫺0.59** 0.15 0.14 0.29** 0.15 0.05 0.28** 0.20 0.09 0.20**
Age ⫺0.02 0.01 ⫺0.12* ⫺0.00 0.01 ⫺0.00 0.00 0.01 0.02 0.01 0.01 0.11
Gender 0.48 0.26 0.11* 0.05 0.16 0.35 ⫺0.11 0.14 ⫺0.06 ⫺0.21 0.26 ⫺0.07
Ethnicity 0.67 0.31 0.01 0.36 0.16 0.19** 0.36 0.17 0.17* 0.00 0.33 0.00
Organizational commitment ⫺0.61 0.09 ⫺0.41** 0.13 0.05 0.26** 0.14 0.05 0.24** 0.21 0.10 0.21*
Emotional exhaustion 0.46 0.07 0.39** ⫺0.02 0.04 ⫺0.06 ⫺0.03 0.04 ⫺0.06 0.01 0.08 0.01
Note. The first step is not shown, as it is identical to Step 1 in Table 2. OCBO ⫽ organizational citizenship behaviors beneficial to the organization;
OCBS ⫽ organizational citizenship behaviors beneficial to one’s supervisor.
* p ⬍ .05, one-tailed. ** p ⬍ .01, one-tailed.
The obtained z score was 4.89, p ⬍ .01. Hence, despite the fact that
emotional exhaustion shows a significant regression coefficient,
our findings are consistent with the possibility that commitment is
an important mediator.
Emotional exhaustion and job performance ratings. Hypoth-
esis 3 posited that emotional exhaustion would predict job perfor-
mance ratings. As shown in Table 2, this anticipated relationship
was observed, ⌬R
⫽ .03, p ⬍ .01; the overall model was signif
icant as well, F(4, 145) ⫽ 4.31, R
⫽ .11, p ⬍ .01. We next
explored Hypothesis 5b by examining the impact of emotional
exhaustion when organizational commitment was entered into the
model. As shown in Table 3, commitment became a strong pre-
dictor of performance ratings, ⌬R
⫽ .08, p ⬍ .01. Emotional
exhaustion, on the other hand, showed only a nonsignificant (di-
rect) relationship. Although these results support our mediated
model, we conducted a test of the indirect path. This relationship
was significant, z ⫽ 2.61, p ⬍ .01.
Emotional exhaustion and organizational citizenship behaviors.
As proposed by Hypothesis 4a, emotional exhaustion predicted
OCBO even beyond the effect of the three demographic control
⫽ .03, p ⬍ .01. The overall model was also
significant, F(4, 145) ⫽ 3.33, R
⫽ .09, p ⬍ .01 (see Table 2).
Unfortunately, OCBS proved to be more difficult to predict. Emo-
tional exhaustion explained only 0.8% of the variance beyond the
controls, and this was nonsignificant. The overall model was
nonsignificant as well, F(4, 145) ⫽ 1.09, R
⫽ .03. Given these
findings, we concluded that Hypothesis 4b was not supported. As
shown in Table 3, when organizational commitment was added to
the equation the impact of emotional exhaustion on OCBO was
⫽ .00, and nonsignificant. Likewise, the indirect
path did exert a significant effect, t ⫽ 2.46, p ⬍ .01, thereby
providing support for Hypothesis 5a. For completeness, Table 3
also reports the mediation analyses for OCBS. As can readily
be seen, only commitment was a significant predictor in the final
step of the equation. Likewise, the indirect effect of emotional
exhaustion was not significant. Thus, there is no support for
Study 2: Replication on a Larger Sample
The goal of the second study was to follow the advice of
Murphy (1983). Murphy recommended that field studies be cross-
validated on independent groups of employees rather than on
hold-out samples. It was especially important to heed this advice in
the present study because we were testing a new theoretical model
of emotional exhaustion and because our findings for Hypothesis 4
in Study 1 were somewhat disappointing. Replication allowed us
to better separate substantive effects from chance occurrences.
Participants. Respondents worked at various organizations along the
Colorado front range. Our sample included both public and private sector
organizations in a diverse number of industries, including human services,
manufacturing, and fitness. Although the initial sample included 296
supervisor–subordinate dyads, only 232 provided complete and usable
data. Hence, the final response rate was 78%. Seventy-six percent of these
participants were women. In addition, 84.9% were White and non-
Hispanic. The remainder were Latino(a) (7.3%), African American (1.7%),
Asian American (2.2%), and Native American (0.9%). The remainder of
the respondents elected not to answer this item. The mean age of the
sample was 30 years.
Procedures. The majority of the data were collected on-site by the
researchers during the organizations’ monthly staff meetings. In these
meetings, subordinates and supervisors completed the survey instruments
in separate rooms so that employees would not be uncomfortable rating
their perceptions of the organization, their level of emotional exhaustion,
and their intentions to leave. This also allowed employees and supervisors
to complete correspondingly coded surveys, eliminating the need to have a
master list of names by which to match employees and supervisors, further
ensuring employee confidentiality.
At the beginning of each data collection session, the researchers gave a
brief introduction explaining the purpose of the survey, confidentiality
procedures, the process for obtaining supervisory ratings, and the benefits
to employees for participating in the study. It was explained that partici-
pating in the study allowed a confidential method for providing their
perceptions of their organization that would be aggregated with other
employees’ perceptions in a general report to the organization. Employees
volunteering to participate were then given an informed consent form to
complete along with the questionnaire packet. After the general introduc-
tion, all supervisors went into a separate room and completed a rating form
on each of their consenting employees. Employees remained in the same
room. Participants then completed identical instruments to those in
Study 1, except that in Study 2, we omitted the turnover intention measure.
This was done to keep the survey as short as possible.
Absent employees were sent surveys with a confidential return envelope
that was mailed directly to the researchers. This packet included a memo
explaining the purpose of the study, assuring them their confidentiality, and
explaining the informed consent process. These employees were tracked by
name so that their supervisor’s could subsequently be sent a questionnaire;
however, the name list was destroyed prior to sending out the supervisor
surveys (further ensuring employee confidentiality). This procedure was
also explained in the employee memo.
Descriptive statistics and correlations. Means, standard devi-
ations, and variable intercorrelations are presented in Table 4. All
of the variables show acceptable levels of reliability except for
⫽ .67), which is somewhat below the .70 convention
recommended by Nunnally and Bernstein (1994). Regardless,
emotional exhaustion is associated with all of the criterion mea-
sures and organizational commitment, although the correlations
are of modest magnitude.
Emotional exhaustion and organizational commitment. Re-
sults for organizational commitment are shown in Table 5. Emo-
tional exhaustion proved to be a solid predictor of organizational
commitment, even beyond the effects of the demographic control
⫽ .12). Furthermore, the overall model was also
significant, F(4, 231) ⫽ 11.44, p ⬍ .01.
Emotional exhaustion and job performance ratings. For job
performance, emotional exhaustion was significant beyond the
effect of the controls (see Table 5). However, the overall regres-
sion model was not, F(4, 231) ⫽ 1.76. The null results for the
overall model are due to the demographic variables, none of which
explained an appreciable amount of variance in job performance
ratings. Steps 2 and 3 of the mediation tests are reported in Table 6.
We excluded Step 1 from this table because it duplicates Step 1 in
Table 5. Contrary to our expectations, the addition of organiza-
tional commitment did little to impact the effect of emotional
exhaustion because the relationship remained significant and of
comparable magnitude. As one might expect given these findings,
the t value for the indirect effect was low (t ⫽ .96) and
Emotional exhaustion and organizational citizenship behaviors.
Replicating the results of Study 1, Table 5 illustrates that emo-
tional exhaustion was a significant predictor of OCBO, ⌬R
p ⬍ .05. Here again, the generally weak effects for the control
variables (gender being something of an exception) drove the
overall equation to nonsignificance, F(4, 231) ⫽ 2.17. The results
suggest that organizational commitment did mediate the relation-
ship between emotional exhaustion and OCBO (see Table 6). After
including commitment, the effect of exhaustion was no longer
significant. More important, the z test suggested by Holmbeck
(2002) was consistent with the possibility of an indirect relation-
ship (z ⫽ 1.99, p ⬍ .05). Table 6 also shows supportive results for
OCBS. Although the overall equation was not significant, F(4,
231) ⫽ 1.81, emotional exhaustion did emerge as a significant
predictor; however; the effect size was small. Finally, there does
seem to be an indirect (i.e., mediated) relationship between emo-
tional exhaustion and OCBS. Inclusion of organizational commit-
ment causes the direct effect of exhaustion to drop to below
conventional levels of significance. Likewise, the indirect effect of
emotional exhaustion was significant, z ⫽ 1.89, p ⬍ .05.
The results of these two studies need to be examined with
respect to our two objectives. First, we examined whether emo-
tional exhaustion predicted effective work attitudes, behaviors, and
intentions. In this regard, our results are supportive. Emotional
exhaustion was found to predict organizational commitment, turn-
over intentions (Study 1), job performance, OCBO, and OCBS
(Study 2). Moreover, most of these relationships remained signif-
icant even after taking into account the effects of age, gender, and
ethnicity. Interestingly, when only the simple correlations were
considered, emotional exhaustion was related to most all criterion
variables in both studies. When taken together, the present studies
Study 2: Regression Results
Organizational commitment Job performance ratings OCBO OCBS
Age 0.03 0.01 0.22** 0.00 0.00 0.03 ⫺0.00 0.00 ⫺0.04 ⫺0.00 0.01 ⫺0.06
Gender 0.00 0.19 0.00 0.09 0.10 0.06 0.20 0.11 0.12* ⫺0.16 0.16 ⫺0.07
Ethnicity ⫺0.03 0.23 ⫺0.01 0.00 0.12 0.00 ⫺0.03 0.13 ⫺0.02 0.24 0.19 0.08
Age 0.02 0.01 0.18** 0.00 0.00 0.01 ⫺0.00 0.00 ⫺0.06 ⫺0.01 0.01 ⫺0.07
Gender ⫺0.06 0.18 ⫺0.02 0.07 0.10 0.05 0.18 0.11 0.11* ⫺0.18 0.16 ⫺0.07
Ethnicity 0.00 0.21 0.00 0.01 0.12 0.01 ⫺0.03 0.12 ⫺0.02 0.25 0.19 0.09
Emotional exhaustion ⫺0.32 0.06 ⫺0.35** 0.08 0.03 ⫺0.16** ⫺0.07 0.03 ⫺0.15* ⫺0.09 0.15 ⫺0.12*
Note. OCBO ⫽ organizational citizenship behaviors beneficial to organizations; OCBS ⫽ organizational citizenship behaviors beneficial to one’s
* p ⬍ .05, one-tailed. ** p ⬍ .01, one-tailed.
Study 2: Means, Standard Deviations, Internal Consistency Reliabilities, and Intercorrelations
Variable MSD 12345678
1. Emotional exhaustion 2.60 1.36 .90
2. Age 30.34 11.05 ⫺.13* —
3. Gender ⫺.08 .16** —
4. Ethnicity .02 .01 ⫺.08 —
5. Commitment 4.17 1.26 ⫺.37** .22** .04 ⫺.01 .82
6. Job performance 4.33 0.65 ⫺.17** .08 .06 .00 .12* .89
8. OCBO 4.11 0.68 ⫺.15** ⫺.02 .11* ⫺.02 .17* .60** .67
9. OCBS 3.61 1.05 ⫺.11* ⫺.07 ⫺.08 ⫺.08 .14* .59** .43** .88
Note. Gender was coded 1 ⫽ male, 2 ⫽ female. Ethnicity was coded 1 ⫽ non-White, 2 ⫽ White. Internal
consistency reliabilities appear along the diagonal. OCBO ⫽ organizational citizenship behaviors beneficial to
organizations; OCBS ⫽ organizational citizenship behaviors beneficial to one’s supervisor.
* p ⬍ .05, one-tailed. ** p ⬍ .01, one-tailed.
replicate earlier work on three criteria (organizational commit-
ment, turnover intentions, and job performance) and provide new
evidence for two others (OCBO and OCBS).
Our second objective was to examine the implications of social
exchange theory for emotional exhaustion. Thus, we explored
whether organizational commitment mediated the relationships
between emotional exhaustion and work outcomes. Using the
analytic techniques suggested by Baron and Kenny (1986) and
Holmbeck (2002), we found that emotional exhaustion exerted a
significant indirect path in the prediction of turnover intentions
(Study 1), job performance (Study 1), OCBO (Studies 1 and 2),
and OCBS (Study 2) by means of organizational commitment. One
unsupportive result concerned performance in Study 2. Although
emotional exhaustion predicted performance, commitment did not
mediate this association.
Another disappointing finding involved the relationship of emo-
tional exhaustion to OCBS. We did not find a significant relation-
ship in our first study. One possibility is that there was something
peculiar to our Study 1 sample. A close look at the size of the
relationships allows us to rule this possibility out. In fact, the size
of the emotional exhaustion–OCBS relationship was higher in
Study 1 (r ⫽⫺.14) and slightly lower in Study 2 (r ⫽⫺.11). The
smaller relationship was significant only because of the larger
sample size in Study 2. The more obvious finding is that the
relationship was small in both studies. In fact, emotional exhaus-
tion simply was not a strong predictor of OCB directed toward the
Given these considerations, we will venture a tentative interpre-
tation of the (weak) relationship between emotional exhaustion
and OCBS. Our interpretation centers on a key distinction between
these two types of citizenship behaviors. OCBO refers to behaviors
beneficial to a larger, and perhaps more impersonal, organization.
OCBS refers to behaviors beneficial to one’s supervisor. Individ-
uals often form close interpersonal relationships with their super-
visors (Rupp & Cropanzano, 2002) and supervisors can, of course,
directly control many of the rewards that employees seek. Given
this, it may be easier for emotionally exhausted individuals to
withhold citizenship behaviors beneficial to the organization as a
whole rather than to a single person with whom they are likely to
One ubiquitous concern in research of this kind involves the
inference of causality. As Bobko and Stone-Romero (1998)
warned, it is difficult to infer causality from cross-sectional data.
Given the cross-sectional nature of our study, the possibility of
alternative causal paths clearly exists. Likewise, even if the causal
paths shown in Figure 1 are in part correct, the possibility of
feedback loops could further complicate our model. For instance,
emotional exhaustion might lower commitment, but this low com-
mitment could further elevate emotional exhaustion. Clearly, our
present data cannot test a recursive model of this kind. Although
not gainsaying these possibilities, it is noteworthy that available
longitudinal research tends to support our model. For example,
Wright and Bonett (1997) found that emotional exhaustion pre-
dicted job performance ratings taken 2 years later. Moreover, the
effect of emotional exhaustion remained significant even after
Time 1 performance was taken into account. Given these findings,
the Wright and Bonett study suggested that emotional exhaustion
acts as a cause of at least some work outcomes. Nevertheless, it is
clear that more research is necessary to address this issue.
Another possibility is that of common method variance (Spec-
tor, 1987). When data are taken from a single source, one often
expects to find somewhat inflated relationships among the vari-
ables. To address this concern, we took the advice of Podsakoff
and Organ (1986) and included two sources of data—the employee
Study 2: Mediation Tests
ratings OCBO OCBS
Age 0.00 0.00 0.00 ⫺0.00 0.00 ⫺0.08 ⫺0.00 0.01 ⫺0.09
Gender 0.09 0.10 0.06 0.20 0.11 0.12* ⫺0.17 0.16 ⫺0.07
Ethnicity 0.01 0.12 0.00 ⫺0.03 0.12 ⫺0.02 0.25 0.19 0.09
Organizational commitment 0.06 0.04 0.12* 0.10 0.04 0.19** 0.14 0.06 0.17**
Age 0.00 0.00 0.00 ⫺0.01 0.00 ⫺0.08 ⫺0.01 0.01 ⫺0.10
Gender 0.07 0.10 0.05 0.19 0.11 0.12* ⫺0.18 0.16 ⫺0.07
Ethnicity 0.01 0.12 0.01 ⫺0.03 0.12 ⫺0.01 0.25 0.19 0.09
Organizational commitment 0.04 0.04 0.07 0.08 0.04 0.15* 0.12 0.06 0.14*
Emotional exhaustion ⫺0.07 0.03 ⫺0.14* ⫺0.05 0.04 ⫺0.10 ⫺0.06 0.05 ⫺0.07
Note. The first step is not shown, as it is identical to Step 1 in Table 5. OCBO ⫽ organizational citizenship
behaviors beneficial to organizations; OCBS ⫽ organizational citizenship behaviors beneficial to one’s super-
* p ⬍ .05, one-tailed. ** p ⬍ .01, one-tailed.
and his or her supervisor. Generally speaking, our predictions held
for the nonself-report measures (e.g., performance and OCB).
However, it is noteworthy that the effect sizes for the self-report
criterion (turnover intentions) was substantially stronger than any
obtained through nonself-report measurement (see Table 1). Thus,
method variance could have inflated the association between emo-
tional exhaustion and turnover intentions.
Allen, M. W., & Meyer, J. P. (1990). The measurement and antecedents of
affective, continuance, and normative commitment to the organization.
Journal of Occupational Psychology, 63, 1–18.
Baba, W., Jamal, M., & Tourigny, L. (1998). Work and mental health: A
decade of Canadian research. Canadian Psychology, 39, 94–107.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable
distinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychol-
ogy, 51, 1173–1182.
Bishop, J. W., & Scott, K. D. (2000). An examination of organizational and
team commitment in a self-directed team environment. Journal of Ap-
plied Psychology, 85, 439–450.
Bishop, J. W., Scott, K. D., & Burroughs, S. M. (2000). Support, commit-
ment, and employee outcomes in a team environment. Journal of Man-
agement, 26, 1113–1132.
Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.
Bobko, P., & Stone-Romero, E. F. (1998). Meta-analysis may be another
useful research tool, but it is not a panacea. In G. R. Ferris (Ed.),
Research in personnel and human resources management (Vol. 16, pp.
359–397). Stamford, CT: JAI Press.
Cordes, C. L., & Dougherty, T. W. (1993). A review and integration of
research on job burnout. Academy of Management Review, 18, 621–656.
Cropanzano, R., & Byrne, Z. S. (2000). Workplace justice and the dilemma
of organizational citizenship. In M. VanVugt, T. Tyler, & A. Biel (Eds.),
Collective problems in modern society: Dilemmas and solutions (pp.
142–161). London: Routledge.
Cropanzano, R., Howes, J. C., Grandey, A. A., & Toth, P. (1997). The
relationship of organizational politics and support to work behaviors,
attitudes, and stress. Journal of Organizational Behavior, 18, 159–180.
Cropanzano, R., Prehar, C., & Chen, P. Y. (2002). Using social exchange
theory to distinguish procedural from interactional justice. Group and
Organizational Management, 27, 324–351.
Cropanzano, R., Rupp, D. E., Mohler, C. J., & Schminke, M. (2001). Three
roads to organizational justice. In J. Ferris (Ed.), Research in personnel
and human resources management (Vol. 20, pp. 1–113). Greenwich,
CT: JAI Press.
Deckop, J. R., Mangel, R., & Cirka, C. C. (1999). Getting more than you
pay for: Organizational citizenship behavior and pay-for-performance
plans. Academy of Management Journal, 42, 420–428.
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001).
The job demands–resources model of burnout. Journal of Applied Psy-
chology, 86, 499–512.
Freudenberger, H. J. (1983). Burnout: Contemporary issues, trends, and
concerns. In B. A. Farber (Ed.), Stress and burnout (pp. 23–28). New
York: Pergamon Press.
Gaines, J., & Jermier, J. M. (1983). Emotional exhaustion in a high stress
organization. Academy of Management Journal, 26, 567–586.
Grandey, A. A., & Cropanzano, R. (1999). The conservation of resources
model applied to work–family conflict and strain. Journal of Vocational
Behavior, 54, 350–370.
Hendrix, W. H., Robbins, T., Miller, J., & Summers, T. P. (1998). Effects
of procedural and distributive justice on factors predictive of turnover.
Journal of Social Behavior and Personality, 13, 1–21.
Holmbeck, G. N. (2002). Post-hoc probing of significant moderational and
mediational effects in studies of pediatric populations. Journal of Pedi-
atric Psychology, 27, 87–96.
Jackson, S. E., Turner, J. A., & Brief, A. P. (1987). Correlates of burnout
among public service lawyers. Journal of Occupational Behavior, 8,
James, L. R., & Brett, J. M. (1984). Mediators, moderators, and tests for
mediation. Journal of Applied Psychology, 69, 307–321.
Jones, R. G., & Best, R. G. (1995, August). Further examination of the
nature and impact of emotional work requirements. Paper presented at
the annual meeting of the Academy of Management, Vancouver, British
Kahill, S. (1988). Symptoms of professional burnout: A review of the
empirical evidence. Canadian Psychology, 26, 284–297.
Konovsky, M. A., & Cropanzano, R. (1991). The perceived fairness of
employee drug testing as a predictor of employee attitudes and job
performance. Journal of Applied Psychology, 76, 698–707.
Konovsky, M. A., & Pugh, S. D. (1994). Citizenship behavior and social
exchange. Academy of Management Journal, 37, 656–669.
Lee, R. T., & Ashforth, B. E. (1990). On the meaning of Maslach’s three
dimensions of burnout. Journal of Applied Psychology, 75, 743–747.
Lee, R. T., & Ashforth, B. E. (1993). A further examination of managerial
burnout: Toward an integrated model. Journal of Organizational Behav-
ior, 14, 3–20.
Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the
correlates of the three dimensions of burnout. Journal of Applied Psy-
chology, 81, 123–133.
Leiter, M. P., Harvie, P., & Frizzell, C. (1998). The correspondence of
patient satisfaction and nurse burnout. Social Science and Medicine, 47,
Leiter, M. P., & Maslach, C. (1988). The impact of interpersonal environ-
ment on burnout and organizational commitment. Journal of Organiza-
tional Behavior, 9, 297–308.
Malatesta, R. M. (1995). Understanding the dynamics of organizational
and supervisory commitment using a social exchange framework. Un-
published doctoral dissertation, Wayne State University, Detroit,
Maslach, C. A. (1982). Burnout: The cost of caring. Englewood Cliffs, NJ:
Maslach, C. A., & Jackson, S. E. (1981). The measurement of experienced
burnout. Journal of Occupational Behavior, 2, 99–113.
Maslach, C. A., & Leiter, M. P. (1997). The truth about burnout: How
organizations cause personal stress and what to do about it. San
Moorman, R. H., Blakely, G. L., & Niehoff, B. P. (1998). Does perceived
organizational support mediate the relationship between procedural jus-
tice and organizational citizenship behavior? Academy of Management
Journal, 41, 351–357.
Murphy, K. R. (1983). Fooling yourself with cross-validation: Single
sample designs. Personnel Psychology, 36, 111–118.
Nowack, K. M., & Hanson, A. L. (1983). The relationship between stress,
job performance, and burnout in college student resident assistants.
Journal of College Student Personnel, 24, 545–550.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory. New York:
Organ, D. W. (1988). Organizational citizenship behavior: The good
soldier syndrome. Lexington, MA: Lexington Books.
Organ, D. W. (1990). The motivational basis of organizational citizenship
behavior. In B. M. Staw & L. L. Cummings (Eds.), Research in orga-
nizational behavior (Vol. 12, pp. 43–72). Greenwich, CT: JAI Press.
Pines, T. L., & Aronson, E. (1988). Career burnout: Causes and cures.
New York: Free Press.
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational
research: Problems and prospects. Journal of Management, 12, 531–544.
Quattrochi-Turbin, S., Jones, J. W., & Breedlove, V. (1983). The burnout
syndrome in geriatric counselors and service workers. Activities, Adap-
tation, and Aging, 3, 65–76.
Randall, M. L., Cropanzano, R., Bormann, C. A., & Birjulin, A. (1999).
Organizational politics and organizational support as predictors of work
attitudes, job performance, and organizational citizenship behaviors.
Journal of Organizational Behavior, 20, 159–174.
Rupp, D. E., & Cropanzano, R. (2002). The mediating effects of social
exchange relationships in predicting workplace outcomes from multifoci
organizational justice. Organizational Behavior and Human Decision
Processes, 89, 925–946.
Settoon, R. P., Bennett, N., & Liden, R. C. (1996). Social exchange in
organizations: Perceived organizational support, leader–member ex-
change, and employee reciprocity. Journal of Applied Psychology, 81,
Shirom, A. (1989). Burnout in work organizations. In C. L. Cooper & I.
Robertson (Eds.), International review of industrial–organizational psy-
chology (pp. 25–48). New York: Wiley.
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in
structural equation models. In S. Leinhart (Ed), Sociological methodol-
ogy 1982 (pp. 290–312). San Francisco, CA: Jossey-Bass.
Sobel, M. E. (1988). Direct and indirect effect in linear structural equation
models. In J. S. Long (Ed.), Common problems/proper solutions: Avoid-
ing error in quantitative research (pp. 46–64). Beverly Hills, CA: Sage.
Spector, P. E. (1987). Method variance as an artifact in self-reported affect
and perceptions at work: Myth or significant problem? Journal of
Applied Psychology, 72, 438–553.
Wayne, S. J., Shore, L. M., & Liden, R. C. (1997). Perceived organiza-
tional support and leader–member exchange: A social exchange per-
spective. Academy of Management Journal, 40, 82–111.
Westman, M., & Eden, D. (1997). Effects of a respite from work on
burnout: Vacation relief and fade-out. Journal of Applied Psychol-
ogy, 82, 516–527.
Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organiza-
tional commitment as predictors of organizational citizenship and in-role
behaviors. Journal of Management, 17, 601–618.
Wright, T. A., & Bonett, D. G. (1997). The contribution of burnout to work
performance. Journal of Organizational Behavior, 18, 491–499.
Wright, T. A., & Cropanzano, R. (1998). Emotional exhaustion as a
predictor of job performance and voluntary turnover. Journal of Applied
Psychology, 83, 486–493.
Zohar, D. (1997). Predicting burnout with a hassle-based measure of role
demands. Journal of Organizational Behavior, 18, 101–115.
Received July 29, 1999
Revision received March 15, 2002
Accepted March 17, 2002 䡲