Linking Job Demands and Resources to Employee Engagement and Burnout: A Theoretical Extension and Meta-Analytic Test

Abstract

We refine and extend the job demands-resources model with theory regarding appraisal of stressors to account for inconsistencies in relationships between demands and engagement, and we test the revised theory using meta-analytic structural modeling. Results indicate support for the refined and updated theory. First, demands and burnout were positively associated, whereas resources and burnout were negatively associated. Second, whereas relationships among resources and engagement were consistently positive, relationships among demands and engagement were highly dependent on the nature of the demand. Demands that employees tend to appraise as hindrances were negatively associated with engagement, and demands that employees tend to appraise as challenges were positively associated with engagement. Implications for future research are discussed.

Full text

Linking Job Demands and Resources to Employee Engagement and Burnout:
ATheoreticalExtensionandMeta-AnalyticTest
Eean R. Crawford and Jeffery A. LePine
University of Florida Bruce Louis Rich
California State University San Marcos
We refine and extend the job demands–resources model with theory regarding appraisal of stressors to
account for inconsistencies in relationships between demands and engagement, and we test the revised
theory using meta-analytic structural modeling. Results indicate support for the refined and updated
theory. First, demands and burnout were positively associated, whereas resources and burnout were
negatively associated. Second, whereas relationships among resources and engagement were consistently
positive, relationships among demands and engagement were highly dependent on the nature of the
demand. Demands that employees tend to appraise as hindrances were negatively associated with
engagement, and demands that employees tend to appraise as challenges were positively associated with
engagement. Implications for future research are discussed.
Keywords: employee engagement, burnout, job demands and resources, challenge and hindrance stress,
meta-analysis
Among organizational managers and executives, there is wide-
spread interest in employee engagement, originally defined by
Kahn (1990, p. 694) as the harnessing of organization members’
selves to their work roles by which they employ and express
themselves physically, cognitively, and emotionally during role
performances, and later defined by Schaufeli, Salanova, Gonza´lez-
Roma´, and Bakker (2002, p. 74) as a positive, fulfilling, work-
related state of mind characterized by vigor, dedication, and ab-
sorption. Engagement has also been viewed as feeling responsible
for and committed to superior job performance (Britt, 1999, 2003),
and on the basis of their narrative review of the literature, Macey
and Schneider (2008) described engagement as a broad construct
consisting of state, trait, and behavioral forms that connote a blend
of affective energy and discretionary effort directed to one’s work
and organization. Managerial interest in engagement is under-
standable given claims from consultancies that having engaged
employees results in higher shareholder return, profitability, pro-
ductivity, quality, and customer satisfaction as well as lower
absenteeism and turnover (e.g., Wellins, Bernthal, & Phelps, n.d.).
Published scientific research on engagement has been surging over
the past 5 years, and results indicate that engagement is associated
with positive job attitudes (e.g., Harter, Schmidt, & Hayes, 2002;
Schaufeli, Taris, & van Rhenen, 2008), lower turnover (e.g., Bak-
ker, Demerouti, & Schaufeli, 2005; Harter et al., 2002; Saks, 2006;
Schaufeli & Bakker, 2004), and higher levels of performance at the
individual and unit levels (e.g., Harter et al., 2002; Rich, LePine,
& Crawford, in press; Salanova, Agut, & Peiro´, 2005; Schneider,
Macey, Barbera, & Martin, 2009; Xanthopoulou, Bakker, Heuven,
Demerouti, & Schaufeli, 2008).
Given the apparent desirability of having engaged employees,
researchers have focused a great deal of their attention on identi-
fying antecedent conditions of engagement that could, in turn,
inform managerial efforts to foster increased levels of engagement
among employees. Kahn (1990), for example, argued that the
psychological conditions of meaningfulness, safety, and availabil-
ity determine levels of employee engagement, and that these
psychological conditions are influenced in turn by multiple char-
acteristics of both the work environment and the individual em-
ployee. Britt’s (1999) triangle model of responsibility positions job
clarity, job control, and job relevance as keys to engagement.
Harter et al. (2002) suggested 12 diverse work characteristics and
management practices that function as key causes of engagement,
some examples of which include clarity of work expectations,
supportiveness of supervisors and coworkers, and opportunities for
growth and development. Maslach and Leiter (1997) proposed that
the main predictors of engagement are factors such as job de-
mands, job control, rewards and recognition, a community of
support, fairness, and compatibility of values between job require-
ments and personal principles. Finally, Macey and Schneider
(2008) proposed that work attributes such as variety, challenge,
and autonomy, in addition to personal characteristics and leader-
ship, should all influence employee engagement.
Common across these perspectives is the focus on perceived
working conditions as predictors of engagement. The job
demands–resources model (Demerouti, Bakker, Nachreiner, &
Schuafeli, 2001) parsimoniously organizes this research by repre-
senting the majority of job attributes and other related working
conditions with two overarching categories—demands and re-
sources. In fact, over half of all the published empirical research on
engagement and its antecedents has been explicitly grounded in
this perspective. Essentially, the job demands–resources model
suggests that job resources promote engagement through a moti-
Eean R. Crawford and Jeffery A. LePine, Department of Management,
University of Florida; Bruce Louis Rich, Department of Management,
California State University San Marcos.
Correspondence concerning this article should be addressed to Eean R.
Crawford, Department of Management, University of Florida, P.O. Box
117165, Gainesville, FL 32611-7165. E-mail: eean.crawford@cba.ufl.edu
Journal of Applied Psychology © 2010 American Psychological Association
2010, Vol. 95, No. 5, 834– 848 0021-9010/10/$12.00 DOI: 10.1037/a0019364
834

vational process and that job demands contribute to burnout—a
syndrome of weariness with work characterized by exhaustion,
cynicism, and inefficacy (Maslach & Leiter, 1997, 2008)—through
an energy depletion process. The job demands–resources perspec-
tive is useful because, to the extent that the relationships of
engagement with specific demands or resources do not vary sig-
nificantly within these overarching categories, the model provides
a good vehicle for summarizing these relationships in a concise
way. Unfortunately, however, the model may be overly parsimo-
nious. Because research grounded in this perspective has produced
conflicting, inconsistent, and unexpected findings on the relation-
ship between demands and engagement (e.g., Bakker, van
Emmerik, & Euwema, 2006), scholars have generally concluded
that demands are not relevant for predicting engagement
(Schaufeli & Bakker, 2004). In contrast, we argue that the job
demands–resources model fails to account for the important dis-
tinction among types of demands with respect to the way they tend
to be appraised by employees (e.g., Cavanaugh, Boswell, Roeh-
ling, & Boudreau, 2000; Lazarus & Folkman, 1984) and that when
this distinction is made, meaningful relationships between de-
mands and engagement emerge.
Thus, given the increasing interest in employee engagement and
the parsimony of the job demands–resources model in organizing
the antecedent influences of working conditions on engagement
and burnout, the general purpose of our research is to quantita-
tively summarize and test the validity of the job demands–
resources model as a theoretical basis for the influence of working
conditions on engagement and burnout. We focus on individual-
level perceptions of working conditions and their relationships
with individual perceptions of engagement and burnout; however,
for simplicity of presentation and consistency with terminology
used in prior research, we refer to relationships between job
demands, job resources, engagement, and burnout. In providing
the first quantitative summary of the research findings in this area,
we resolve inconsistencies in job demands–engagement relation-
ships by integrating knowledge from the transactional theory of
stress (Boswell, Olson-Buchanan, & LePine, 2004; Cavanaugh et
al., 2000; Lazarus & Folkman, 1984; LePine, Podsakoff, &
LePine, 2005) to propose that job demands tending to be appraised
by employees as hindrances are negatively related to engagement,
whereas job demands tending to be appraised by employees as
challenges are positively related to engagement. We tested hypoth-
eses from this refined and updated theory by obtaining meta-
analytic estimates from primary research of the relationships be-
tween job demands, resources, engagement, and burnout, and by
submitting that meta-analytic correlation matrix to a path analysis
to quantitatively summarize and statistically verify the proposi-
tions of this new model.
Our research makes several important theoretical contributions.
First, although prior meta-analyses have examined some of the
relationships in our model (e.g., Halbesleben, 2006; Lee & Ash-
forth, 1996; LePine et al., 2005), the most central relationships
with engagement have not been meta-analyzed, and no prior meta-
analytic research has examined the relationships simultaneously in
a way that allows for the falsification of a theoretical model in
which engagement plays a central role. Thus, our research tests a
theoretical model of engagement rather than isolated, disconnected
relationships that could change if considered in concert with other
variables. Second, by extending the distinction of challenges ver-
sus hindrances in the examination of the relationship between
demands and engagement, we advance theoretical understanding
of engagement by challenging status quo thinking in an area where
quite a bit of research has taken place. Importantly, we show that
true relationships exist between different types of demands and
engagement where scholars have generally concluded there are
none (e.g., Hallberg, Johansson, & Schaufeli, 2007; Llorens, Bak-
ker, Schaufeli, & Salanova, 2006; Mauno, Kinnunen, & Ruoko-
lainen, 2007; Schaufeli & Bakker, 2004; Schaufeli et al., 2008).
Third, although prior research has shown that challenge and hin-
drance stressors are related in opposing directions to outcomes
such as performance, motivation, and job attitudes (LePine et al.,
2005; Podsakoff, LePine, & LePine, 2007), none of this research
has examined relationships with engagement. Although it might
seem reasonable to presume that engagement should operate sim-
ilarly, researchers have demonstrated that engagement is a unique
concept that has functional relationships with more distal criteria
that differ from various other job attitudes and motivational con-
cepts. For example, Rich et al. (in press) showed that engagement
was distinct from job satisfaction, job involvement, intrinsic mo-
tivation, task performance, and citizenship behavior. Further, they
showed that the indirect effects of proposed antecedents on per-
formance and citizenship were mediated entirely through engage-
ment. None of the other attitudes (all included as simultaneous
mediators in their model) exhibited significant indirect effects.
Thus, our research provides evidence necessary to test, rather than
assume by virtue of its presumed similarity to other attitudinal or
motivational concepts, that engagement exhibits hypothesized re-
lationships with challenge and hindrance stressors. Finally, by
providing evidence as to the consistency of engagement relation-
ships within the broad categories of resources, challenge demands,
and hindrance demands, we provide an examination of the support
for the utility of the job demands–resources perspective and our
extension of it as a means of parsimoniously depicting relation-
ships among working conditions and engagement.
A Theory of Differentiated Job Demands and
Resources
The Job Demands and Resources Perspective
The job demands–resources perspective assumes that whereas
every occupation may have its own specific risk factors associated
with motivation and job stress, these factors can be classified in
two general categories—job demands and job resources—that
constitute an overarching model that may be applied to various
occupational settings, regardless of the particular demands and
resources involved (Bakker & Demerouti, 2007; Bakker, Demer-
outi, de Boer, & Schaufeli, 2003; Bakker, Demerouti, & Euwema,
2005; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Job
demands refer to those physical, social, or organizational aspects
of the job that require sustained physical or mental effort and are
therefore associated with certain psychological costs (e.g., exhaus-
tion) and include aspects such as workload, time pressure, and
difficult physical environments. Job resources refer to those as-
pects of the job that are functional in achieving work goals,
stimulate personal growth and development, and reduce job de-
mands and their associated physiological and psychological costs
and include aspects such as job control, opportunities for devel-
835
JOB DEMANDS, RESOURCES, ENGAGEMENT, AND BURNOUT

opment, participation in decision making, task variety, feedback,
and work social support.
The job demands–resources model assumes two processes that
explain the relationships with engagement and burnout (Bakker &
Demerouti, 2007; Schaufeli & Bakker, 2004). First, job demands
are assumed to activate an energy depletion process whereby an
employee’s sustained increases in effort to meet perceived job
demands are met with an increase in compensatory psychological
and physiological costs that drain the employee’s energy. The
depletion of energy and increased stress from responding to de-
mands gradually leads employees to feel used up and worn out.
Thus, job demands are assumed to have a direct positive relation-
ship with burnout. Second, job resources are assumed to activate a
motivational process whereby perceived resources that are instru-
mental in achieving work goals can also foster employees’ growth,
learning, and development; satisfy needs for autonomy and com-
petence; and increase willingness to dedicate one’s efforts and
abilities to the work task. These perceptions and beliefs increase
the degree to which individuals are willing to invest their selves
into their role performances. Thus, job resources are assumed to
have a direct positive relationship with engagement. Empirical
evidence from multiple studies in various occupations and coun-
tries confirms that job demands are positively associated with
burnout, whereas job resources are positively related to engage-
ment (Bakker & Demerouti, 2007).
Evidence from several studies (e.g., Bakker, Demerouti, &
Euwema, 2005; Bakker, Demerouti, & Schaufeli, 2003; Schaufeli
& Bakker, 2004) also suggests that job resources have a direct
negative relationship with burnout. The conservation of resources
theory suggests this is so because stress occurs when resources are
lost or threatened, and this stress will lead employees to burn out
over time (Hobfoll & Freedy, 1993). Individuals with larger pools
of resources are more easily able to meet demands and to protect
themselves from the strains of resource depletion (Lee & Ashforth,
1996), whereas individuals with limited resources to meet de-
mands more quickly accrue strains that over time result in burnout.
Thus, job resources are assumed to have a direct negative rela-
tionship with burnout.
Finally, although the job demands–resources model clearly pre-
dicts that job demands lead to burnout, the evidence concerning the
relationship between job demands and engagement is ambiguous,
and as a result, scholars have generally concluded that demands are
not relevant for predicting engagement. Further, when researchers
have found demands and engagement to be related, the findings
were most often unexpected rather than hypothesized. For exam-
ple, in a study of job demands, resources, burnout, and engagement
using four independent employee samples, Schaufeli and Bakker
(2004) concluded from their structural equation models that job
demands do not predict engagement. In a later study, Schaufeli et
al. (2008) unexpectedly found that job demands such as time
pressure (e.g., having to work very fast) were actually positively
related to engagement. As other examples of unexpected findings,
Sonnentag (2003) found in a study of 147 German public service
employees that although the job demand of situational constraints
(e.g., missing or defective equipment; missing or outdated infor-
mation) was significantly negatively associated with engagement,
the job demand of time pressure was not significantly related to
engagement. Xanthopoulou, Bakker, Demerouti, and Schaufeli
(2007) found in a study of 714 Dutch employees that although two
job demands (emotional dissonance, organizational changes) were
significantly negatively related to engagement, one other demand
(high workload) was significantly positively related to engage-
ment. Bakker et al. (2006) found that although physical workplace
demands were negatively related to engagement, work time pres-
sure demands were positively related to engagement, and the
positive relationships were of a greater magnitude than the nega-
tive ones (average r!.35 vs. r!".19, respectively).
As these examples reveal, the general conclusion from previous
research that demands and engagement are not related does not quite
tell the whole story. Unexpectedly, in some cases demands appear
to be unrelated to engagement, in other cases demands appear to
detract from engagement, and yet in other cases demands appear to
promote engagement. In the aggregate, these differences may
result in overall relationships that appear weak and insignificant,
especially if these differences are assumed to result from random
variation in the relationships of demands and engagement. A
shortcoming of the job demands–resources model is that it cur-
rently lacks theory to account for this variation, and given the
importance of predicting engagement in the model, it is crucial to
consider whether these differences in relationships with demands
are systematic, and if so, what explains them? It is here that we
clarify and update the job demands–resources model with reason-
ing from the transactional theory of stress to show that demands
vary by type. In essence, we argue that the failure to differentiate
the type of demand in the current job demands–resources model
masks relationships with engagement that in fact exist. Although
all demands require sustained effort, deplete energy, and result in
strain and burnout, demands also vary systematically in terms of
the psychological responses they tend to trigger. Thus, as we
explain below, whereas some demands should reduce engagement,
other demands may actually promote engagement.
Differentiating Job Demands
According to the transactional theory of stress, people appraise
stressful situations such as job demands in terms of their signifi-
cance for well-being as either potentially challenging or threaten-
ing (Lazarus & Folkman, 1984). Supporting this reasoning,
Cavanaugh et al. (2000) found two factors, labeled challenge
stressors and hindrance stressors, underlying scores on items from
several popular measures of stress. Challenges tend to be appraised
as stressful demands that have the potential to promote mastery,
personal growth, or future gains. Examples of challenges include
demands such as a high workload, time pressure, and high levels
of job responsibility. Employees tend to perceive these demands as
opportunities to learn, achieve, and demonstrate the type of com-
petence that tends to get rewarded. Hindrances tend to be appraised
as stressful demands that have the potential to thwart personal
growth, learning, and goal attainment. Examples of hindrances
include demands such as role conflict, role ambiguity, organiza-
tional politics, red tape, and hassles. Employees tend to perceive
these demands as constraints, barriers, or roadblocks that unnec-
essarily hinder their progress toward goal attainment and rewards
that accrue as a result of being evaluated as an effective performer.
We acknowledge that appraisals of demands vary somewhat as a
function of the characteristics of the individual doing the apprais-
ing (Lazarus & Folkman, 1984). However, as Brief and George
(1995) argued, work contexts evoke a fairly consistent economic
836 CRAWFORD, LEPINE, AND RICH

meaning to individuals, and consequently, individuals tend to
appraise work-related stressors in fairly consistent ways. Empirical
evidence from samples of executives (Cavanaugh et al., 2000),
lower level employees (Boswell et al., 2004), and part-time master
of business administration (MBA) students (LePine et al., 2005)
supports the notion that despite individual differences and experi-
ences that result in unique perceptions regarding the level of job
demands, certain types of demands are more likely to be appraised
as challenges, and other types of demands are more likely to be
appraised as hindrances.
The outcome of the initial appraisal of demands as challenges or
hindrances influences subsequent emotions and cognitions, which
in turn, influence how a person copes with the demand (Lazarus &
Folkman, 1984; LePine et al., 2005). Challenge demands, because
they tend to be appraised as having the potential to promote
personal growth or gains, tend to trigger positive emotions (e.g.,
eagerness, excitement, exhilaration) and an active or problem-
solving style of coping (e.g., strategizing, increases in effort).
Individuals should be more willing to invest themselves in re-
sponse to challenging demands because they are likely to feel more
confident and secure that expending their effort will allow them to
successfully meet these demands, and they are likely to see the
opportunity for growth or gain achieved by meeting them as
meaningful and desirable (Kahn, 1990; Lazarus & Folkman,
1984). As an example, people with a high level of job responsi-
bility are likely to feel that they can successfully meet this demand
by working very hard and that by doing so they will experience a
sense of personal accomplishment and perhaps receive formal
recognition. This is consistent with Macey and Schneider’s (2008)
view that challenging situations promote engagement when em-
ployees trust their investment of time and energy will be rewarded
in some meaningful way. Further, Britt, Adler, and Bartone’s
(2001) research has shown that perceiving opportunity and mean-
ing in the face of stressful demands predicts individuals’ deriving
benefits from coping with those demands. Other research has
shown that the experience of positive emotions and the experience
of meaning emanating from being challenged are both associated
with greater levels of motivation and engagement (Erez & Isen,
2002; May, Gilson, & Harter, 2004). This is also consistent with
reasoning and empirical evidence from the job characteristics
model, which has shown that job characteristics that promote the
experience of meaningfulness and responsibility are highly asso-
ciated with internal work motivation (Fried & Ferris, 1987; Hack-
man & Oldham, 1980). As people feel that coping efforts will be
effective and expect to experience meaning in meeting these
challenges, they become more willing to invest the energy to adopt
more active, problem-focused styles of coping, and such invest-
ments should be reflected in greater engagement. Thus, challenge
demands should have a positive relationship with engagement.
Hindrance demands, because they tend to be appraised as having
the potential to harm or block personal growth or gains, tend to
trigger negative emotions (e.g., fear, anxiety, anger) and a passive
or emotional style of coping (e.g., rationalization, withdrawing
from the situation). Individuals should be less willing to invest
themselves to respond to hindering demands because the negative
emotions they experience are likely to make them feel unable to
adequately deal with these demands. In fact, because people are
likely to believe that using up resources to cope with these de-
mands will block them from attaining meaningful outcomes, they
are apt to have little motivation to actively cope with these de-
mands and, thus, resort to a more passive, disengaging style of
coping to deal with the negative emotions associated with these
hindrances (Kahn, 1990). For example, people experiencing con-
flicting role demands at work are likely to believe that no reason-
able amount of effort will be sufficient to simultaneously satisfy
each demand, and they will likely devote energy that could oth-
erwise be dedicated to meeting demands associated with valued
outcomes to coping with the anxiety and frustration resulting from
leaving conflicting roles unfulfilled. This is consistent with Harter
et al.’s (2002) view that people are less likely to be cognitively and
emotionally engaged to the extent that they encounter obstacles
such as not knowing what is expected of them or lacking what they
need to do their work. Indeed, research has shown that resources
consumed dealing with negative emotions and the psychological
threat associated with hindering situations are associated with
decreased levels of motivation and engagement (May et al., 2004;
Porath & Erez, 2009). This is also consistent with reasoning and
findings from self-determination theory that show that threats to
the satisfaction of needs for competence, relatedness, and auton-
omy serve to weaken internal motivation (Deci et al., 2001; Meyer
& Gagne´, 2008). Thus, as people feel that they will be frustrated
in their efforts to overcome these hindrances or barriers and
perceive they will be blocked from attaining meaningful outcomes,
they become less willing to invest energy to deal with hindrances
directly and will resort to more passive, emotion-focused styles of
coping such as those reflected in withdrawal and decreased en-
gagement. Thus, hindrance demands should have a negative rela-
tionship with engagement.
To summarize our refinement of the job demands–resources
perspective, all job demands, whether challenges or hindrances,
should be positively related to burnout because the increased effort
associated with the appraisal of demands and coping with them
results in strain (e.g., anxiety, fatigue), which in turn, is dissatis-
fying and over time can lead to employees feeling exhausted and
worn out. However, this does not mean that individuals who may
at some point feel exhausted will necessarily be unwilling to invest
themselves (be engaged) when they are confronted with demands,
especially in regards to demands they appraise as being particu-
larly meaningful and important to address. As we have reasoned,
on the one hand, challenge demands should be positively related to
engagement because they trigger positive emotions and active,
problem-focused coping styles that increase willingness to invest
energy in efforts to meet these demands. Individuals may work
very passionately to respond to challenge demands because they
believe doing so is meaningful and important, even though they
simultaneously recognize that doing so may also leave them feel-
ing exhausted. On the other hand, hindrance demands should be
negatively related to engagement because they result in negative
emotions and passive, emotion-focused coping styles that reflect
withdrawal and reduced employee engagement. Individuals may
be very detached in addressing hindrance demands because they
believe they will be frustrated by having to waste additional energy
and personal resources, beyond the exhaustion they may experi-
ence from having to deal with these obstacles. Finally, and con-
sistent with previous theory, job resources should be negatively
related to burnout because individuals with greater resources are
more easily able to meet demands and to protect themselves from
the strains of further resource depletion, whereas those with fewer
837
JOB DEMANDS, RESOURCES, ENGAGEMENT, AND BURNOUT

resources accrue strains that result in burnout more quickly. Job
resources should be positively associated with engagement be-
cause they are instrumental in achieving work goals, foster em-
ployees’ growth and development, satisfy needs for autonomy and
competence, and increase willingness to dedicate one’s efforts and
abilities to the work task.
Hypothesis 1: Job demands, whether challenges or hin-
drances, are positively associated with burnout.
Hypothesis 2: Job resources are negatively associated with
burnout.
Hypothesis 3: Job resources are positively associated with
engagement.
Hypothesis 4: The relationship between job demands and
engagement depends on the type of the demand such that job
challenge demands are positively associated with engage-
ment, and job hindrance demands are negatively associated
with engagement.
Method
To test our hypotheses, we used Hunter and Schmidt’s (1990)
method of meta-analysis. Accordingly, we estimated true popula-
tion correlations among variables by sample weighting correla-
tions from primary studies and by applying formulas that account
for sampling and measurement error in both the predictor and
criterion. We searched the PsycINFO and Web of Science data-
bases through April 2008 for articles with meta-analyzable results.
We used the keyword engagement independently and in conjunc-
tion with the keywords burnout,demands, and resources to con-
duct the search. To be considered, studies needed to include
employee engagement or burnout as an individual-level criterion
and at least one job demand or resource and report a correlation
coefficient (r) or information that could be used to compute this
coefficient.
1
To identify additional studies, we examined the ref-
erence sections of meta-analyses, narrative reviews, book chapters,
and conceptual articles on employee engagement and the job
demands–resources model. We also contacted 16 scholars who
have published in the employee engagement and/or job demands–
resources domain and asked for unpublished manuscripts or raw
data. Finally, we included previously unidentified manuscripts if
they were in our possession or were identified by friendly review-
ers. The final database included 55 manuscripts and articles re-
porting relationships from 64 samples.
The first and second authors jointly developed a standardized
procedure for coding the articles and identified an exhaustive set of
decision rules for all coding decisions. The first author then used
this procedure to independently code all studies. We recognized
that questions would arise in the course of coding the articles,
and we decided that when this happened, the other two authors
would be contacted, and the issue would be resolved though
consensus. To provide a check on the accuracy and reliability of
coding regarding (a) the way the variables in the primary studies
were categorized, (b) the effect sizes, and (c) the reliabilities of the
predictor and criterion, the second author independently coded 20
studies. Across these studies, we found that the first and second
authors agreed 97% on the categorization of the variables, 100%
on the effect sizes, and 95% on the reliabilities. We returned to the
primary studies to determine where the differences occurred.
Nearly all the discrepancies were due to mistakes in data entry, and
all other disagreements were minor and resolved through a con-
sensus discussion.
Measures of demands were categorized as challenges or hin-
drances on the basis of the two-dimensional framework that has
been validated by factor analysis, subject matter experts’ catego-
rizations, employee ratings of job demands as challenging and/or
hindering, critical incident techniques, and previous meta-analyses
(Boswell et al., 2004; Cavanaugh et al., 2000; LePine, LePine, &
Jackson, 2004; LePine et al., 2005; Podsakoff et al., 2007). When
there was doubt as to the appropriate category for a particular
measure, we examined the content of the measure and reached a
consensus using Lazarus and Folkman’s (1984) theory. Consistent
with this body of previous research, job challenge demands in-
cluded measures of the level of attention required by job/role
demands, job complexity, job responsibility, pressure to complete
tasks, time urgency, and quantitative and subjective workload;
whereas job hindrance demands included measures of situational
constraints, hassles, organizational politics, resource inadequacies,
role ambiguity, role conflict, and role overload. We note that our
search identified studies with measures of demands that combined
challenges and hindrances (e.g., Jackson, Rothmann, & van de
Vijver, 2006) or were difficult to classify as either challenges or
hindrances for other reasons (e.g., emotional demands; Llorens et
al., 2006). Although we did not report meta-analyses that included
effect sizes from these measures, they are available from the first
author upon request. Measures of resources included job control;
autonomy; coworker, supervisor, and organization support; feed-
back; access to information; work opportunities for development;
positive social climate; innovative organization climate; job vari-
ety; and positive workplace events.
In studies grounded explicitly in the job demands–resources
perspective, employee engagement has been measured predomi-
nantly using the Utrecht Work Engagement Scale (UWES;
Schaufeli et al., 2002). We found fewer studies relating job de-
mands or resources with engagement measured by scales on the
basis of other perspectives (e.g., Britt, 1999; Kahn, 1990; May et
al., 2004). One other common measure of employee engagement is
the Gallup Workplace Audit (Harter et al., 2002). This measure
comprises employee perceptions of work characteristics such as
supervisor feedback and coworker support considered to be “en-
gagement conditions,” each of which is a causal contributor to
engagement but not a measure of engagement itself (Harter &
Schmidt, 2008). Because associations between job resources and
this measure of engagement may be driven by overlap in predictor
and criterion, the Gallup measure of engagement was not included
in this analysis. Finally, Maslach and Leiter’s (1997) perspective
equates engagement and burnout as exact empirical opposites by
1
Salanova et al.’s (2005) study, which surveyed three employees from
each of 114 work units of restaurants and hotel chains and aggregated
employee data to the work-unit level, was included. This study contributed
only one estimate of the resource–engagement relationship to the meta-
analysis, and we chose to retain it in the analysis because the ratio of
employees to work units was very small (3:1), and the effect size was
consistent with those reported at the individual level of analysis.
838 CRAWFORD, LEPINE, AND RICH

definition rather than by hypothesis, and for this reason studies
measuring engagement by reverse-scoring measures of burnout
(e.g., Maslach & Leiter, 2008) were excluded from the analysis, as
a test for covariation between this measure of engagement and
burnout is not possible. As for burnout, it has been measured in job
demands–resources research nearly exclusively using some form
of the Maslach Burnout Inventory (MBI; Maslach, Jackson, &
Leiter, 1996; Schaufeli, Leiter, Maslach, & Jackson, 1996). Fewer
studies used alternative measures of burnout (e.g., Little, Sim-
mons, & Nelson, 2007). We compared the results of our meta-
analyses using data from all samples regardless of which measure
of engagement or burnout was used to restricted meta-analyses
using data from research that only used the UWES and the MBI.
Results and conclusions are virtually identical between the two
sets of analyses, and we report the meta-analyses from the studies
using the full set of engagement and burnout measures here.
Meta-analyses of samples using only the UWES and MBI mea-
sures are available from the first author upon request. In studies
with multiple measures of a construct, we calculated average
correlations and reliabilities ensuring that each sample contributed
one independent relationship to the meta-analysis.
2
Results
Table 1 reports the results of the meta-analyses among the study
constructs, giving the number of correlations (k), total sample size
(N), estimated sample-weighted mean correlation (r), estimated
true correlation corrected for sampling error and unreliability (#),
and the 95% confidence interval around the estimated true corre-
lation (95% CI). Confidence intervals provide an estimate of the
variability around the estimated mean correlation; a 95% confi-
dence interval excluding zero indicates that if we repeatedly sam-
pled the population of correlations, 97.5% or more of the associ-
ated confidence intervals would exclude zero (the other 2.5% of
the average correlations would lie in the other tail of the distribu-
tion) (Judge, Ilies, Bono, & Gerhardt, 2002).
The first column of Table 1 includes meta-analyses with undif-
ferentiated job demands to estimate a path model of the conven-
tional job demands–resources perspective. This column shows that
undifferentiated demands have a positive relationship with burnout
(#!.27, p$.05) and a negative relationship with engagement
(#!".07, p$.05). We note here that our estimate for the job
demands–burnout relationship is very consistent with Lee and
Ashforth’s (1996) estimate (average corrected correlation !.28).
These results appear to support the conventional propositions of
the job demands–resources model that job demands are primarily
predictive of burnout, whereas job demands are much less relevant
to engagement.
The second and third columns of Table 1, however, include
meta-analyses with job demands differentiated as challenges and
hindrances, respectively, to estimate the refined model we have
proposed. These columns show that although both types of job
demands still have positive relationships with burnout (challenge
with burnout, #!.16, p$.05; hindrance with burnout, #!.30,
p$.05), job demands differentiated as challenges and hindrances
now also have opposing relationships with engagement. Challenge
demands have a positive relationship with engagement (#!.16,
p$.05), whereas hindrance demands have a negative relationship
with engagement (#!".19, p$.05). Of note, both of these
estimates are larger in magnitude than is the estimate of undiffer-
entiated demands with engagement, and their signs are in the
direction expected. That is, challenge demands are associated with
higher levels of engagement, whereas hindrance demands are
associated with lower levels of engagement. Thus, the separation
of job demands into challenges and hindrances uncovers true
relationships that exist between demands and engagement that
have thus far been masked by their nondistinction in the traditional
job demands–resources model.
The fourth column of Table 1 presents relationships of resources
with burnout and engagement. This column shows that job re-
sources have a positive relationship with engagement (#!.36,
p$.05) and that job resources have a negative relationship with
burnout (#!".27, p$.05). These relationships support the
predictions of the job demands–resources model that job resources
positively predict engagement but can also negatively predict
burnout. The final column shows that burnout and engagement
have a strong negative relationship (#!".48, p$.05), which is
consistent with their positioning in the job demands–resources
model as independent, yet negatively correlated constructs
(Schaufeli & Bakker, 2004).
Because we are interested in the simultaneous effects of the job
demands and resources antecedents on the engagement and burn-
out outcomes, and to formally test our hypotheses, we input
matrices of the relevant estimated true correlations using the
harmonic means of the cell samples sizes into LISREL 8.52
(Jo¨reskog & So¨rbom, 2002) to estimate meta-analytic path models
using procedures outlined by Viswesvaran and Ones (1995). As
shown in the top panel of Figure 1, we first estimated a conven-
tional job demands–resources model without differentiating chal-
lenge demands from hindrance demands. As shown in the bottom
panel of Figure 1, we then estimated the updated job demands–
resources model we have proposed by differentiating challenge
demands from hindrance demands. Because previous job
demands–resources research has shown that, on the antecedent
side, demands and resources are negatively correlated, and that, on
the outcome side, burnout and engagement are negatively corre-
lated (Schaufeli & Bakker, 2004), we allowed these correlations to
be estimated in both of our models. Further, because prior research
has shown that challenge demands and hindrance demands are
moderately related (#!.33, p$.05; LePine et al., 2005), and
because our estimate reported in Table 1 is consistent with this
relationship (#!.23, p$.05), we also allowed this correlation to
be estimated in our path model. As a result, each model has the
maximum number of paths specified, and although we were the-
oretically and empirically justified in doing so, this created satu-
rated models that fit the data perfectly (i.e., %
2
!0.00). Thus, we
do not report fit indices in our focus on the substantive relation-
ships to test our hypotheses.
The parameter estimates to test our hypotheses are found in the
bottom panel of Figure 1. In regards to relationships with burnout,
2
Although we could have calculated composite correlations using
Hunter and Schmidt’s (1990) formulas, the population estimates would
have been marginally stronger, and this would have given us an advantage
in confirming our hypotheses. Thus, because our study focused more on
testing hypotheses rather than providing population parameters, we chose
the more conservative approach.
839
JOB DEMANDS, RESOURCES, ENGAGEMENT, AND BURNOUT

Hypothesis 1 is supported, as both challenge demands (&!.10,
p$.05) and hindrance demands (&!.25, p$.05) are positively
and significantly related to burnout. Hypothesis 2 is also sup-
ported, as job resources (&!".24, p$.05) are negatively and
significantly related to burnout. Together, challenge demands,
hindrance demands, and job resources explain 15% of the variance
in burnout. In regards to relationships with engagement, Hypoth-
esis 3 is supported, as job resources (&!.34, p$.05) are
positively and significantly related to engagement. Hypothesis 4 is
also supported, as challenge demands (&!.21, p$.05) are
positively and significantly related to engagement, whereas hin-
drance demands are negatively and significantly related to engage-
ment (&!".19, p$.05). Together, challenge demands, hin-
drance demands, and job resources explain 19% of the variance in
engagement. As shown in the contrasting top panel of Figure 1, the
variance explained in engagement by the conventional model was
only 13%. Thus, distinguishing demands as challenges and hin-
drances increased the variance explained in engagement by nearly
half.
Previous meta-analyses have provided estimates of the relation-
ships of burnout with specific demands and resources (e.g., Hal-
besleben, 2006; Lee & Ashforth, 1996). However, there is no
meta-analysis reporting the relationships of engagement with sim-
ilar specific demands and resources. Because a purpose of this
article was to clarify inconsistencies in previous relationships of
demands with engagement and to provide evidence as to the
consistency of relationships within categories, we report in Table
2 meta-analytic estimates of the relationships of engagement with
specific types of resources, challenge demands, and hindrance
demands. Doing so allowed us to understand whether the clarified
relationships of engagement with challenge demands and hin-
drance demands occur as a function of one or two specific de-
mands, or whether this pattern is observed generally across each
demand within the challenge and hindrance categories. These
results also provide clearer information regarding the frequency
with which different types of demands and resources have been
related to engagement. Results are reported for nine main catego-
ries of resources (autonomy and job control; feedback and access
to information; opportunities for development; positive workplace
climate; recovery time; rewards and recognition; organizational,
supervisor, coworker, and social support; job variety; and work
role fit), three main categories of challenge demands (job respon-
sibility, time urgency, and workload), and six main categories of
hindrance demands (administrative hassles, emotional conflict,
organizational politics, resource inadequacies, role conflict, and
role overload).
In addition to reporting the number of correlations (k), total
sample size (N), estimated sample-weighted mean correlation (r),
estimated true correlation corrected for sampling error and unre-
liability (#), the 95% confidence interval around the corrected
correlation (95% CI), we also report the 80% credibility interval
around the corrected correlation (80% CV), the percentage of
variance in the correlations explained by statistical artifacts (per-
centage of variance explained), and the Qstatistic. Credibility
intervals provide an estimate of the variability of individual cor-
relations in the population; an 80% credibility interval excluding
zero indicates that more than 90% of the individual correlations in
the population will exclude zero (another 10% will lie above the
upward limit of the interval). Thus, as noted before, whereas
Table 1
Meta-Analytic Results for the Relationships Among Demands, Resources, Burnout, and Engagement
Variable
Demands (undifferentiated) Challenge demands Hindrance demands Resources Burnout
kN r #95% CI kN r #95% CI kN r #95% CI kN r #95% CI kN r #95% CI
Demands
Challenge
demands
Hindrance
demands 11 5,604 .19 .23 [.14, .32]
Resources 32 18,348 ".06 ".07 [".12, ".02] 23 11,629 ".01 ".01 [".06, .04] 19 11,127 ".10 ".13 [".19, ".06]
Burnout 27 16,607 .22 .27 [.22, .32] 18 9,794 .14 .16 [.10, .22] 15 9,439 .24 .30 [.23, .37] 24 14,392 ".22 ".27 [".31, ".24]
Engagement 43 26,724 ".06 ".08 [".13, ".03] 26 12,487 .14 .16 [.11, .21] 28 18,540 ".16 ".19 [".23, ".16] 54 27,200 .30 .36 [.33, .39] 54 25,998 ".39 ".48 [".51, ".45]
Note. k is the number of correlations used for the meta-analysis; Nis the combined sample size for the meta-analysis; ris the sample-weighted correlation; #is the estimated true correlation corrected
for sampling error and unreliability; 95% CI is the 95% confidence interval around the estimated true correlation. As indicated by the 95% confidence intervals, all estimated true correlations are
significant at p$.05 except for the challenge demands–resources relationship ( p'.05).
840 CRAWFORD, LEPINE, AND RICH

confidence intervals estimate variability in the estimated mean
correlation, credibility intervals—as described here—estimate
variability of the individual correlations in the population of stud-
ies (Judge et al., 2002). The percentage of variance explained by
artifacts and the Qstatistic are two indicators of the amount of
heterogeneity in the corrected correlations across studies. A low
percentage of variance explained and a statistically significant Q
suggest heterogeneous effect sizes, which indicate the presence of
potential moderators. However, because of the small number of
studies for most variables, it was not possible to conduct any
moderator analysis on the correlates with heterogeneous effects
sizes. Thus, their individually corrected correlations should be
interpreted with caution, especially where the credibility interval
includes zero (Whitener, 1990).
As can be seen in Table 2, a pattern of fairly uniform positive
correlations with engagement is observed across the nine different
types of job resources. Autonomy (#!.37, p$.05), feedback
(#!.35, p$.05), opportunities for development (#!.47, p$
.05), positive workplace climate (#!.28, p$.05), recovery (#!
.29, p$.05), rewards and recognition (#!.21, p'.05), support
(#!.33, p$.05), job variety (#!.53, p$.05), and work role
fit (#!.52, p$.05) all had significant positive relationships with
engagement. The majority of the confidence intervals overlap with
each other and with the confidence interval for job resources in
general reported in Table 1. The confidence interval for positive
workplace climate [.23, .33] is slightly lower than that of job
resources in general [.33, .39], whereas the confidence intervals for
job variety [.49, .57] and work role fit [.43, .61] are somewhat
higher. These two intervals do overlap, however, with those of
opportunities for development [.30, .63] and recovery [.07, .51] as
well as with each other. All of the confidence intervals and
credibility intervals exclude zero except for the credibility interval
for rewards and recognition. Thus, in general, it appears that the
positive relationship of job resources with engagement is fairly
robust across the specific types of resources.
A similar pattern of consistently positive correlations with en-
gagement is observed across the three different types of challenge
demands, albeit these relationships are somewhat smaller than the
positive relationships of engagement with the various job re-
sources. Job responsibility (#!.15, p$.05), time urgency (#!
.21, p$.05), and workload (#!.13, p$.05) all had positive
relationships with engagement. All of the confidence intervals
overlap and exclude zero, and all of the credibility intervals over-
lap and exclude zero. Further, the confidence and credibility in-
Demands
Reso urce s
Burnout
Engagement
.25*
.36*
-.07* -.37*
-.06*
-.25*
Challenge
Demands
Hindrance
Demands
Burnout
Engagement
.23*
-.36*
.21*
.25*
Reso urce s
-.13*
-.01
-.19*
.10*
-.24*
.34*
R
2
= .15
R
2
= .19
R
2
= .14
R
2
= .13
Figure 1. Summary of meta-analytic path analyses of the conventional (top panel—job demands undifferentiated;
harmonic M!20,212) and the differentiated (bottom panel—job demands differentiated as challenges and hin-
drances; harmonic M!11,889) job demands–resources models. Values are standardized estimates.
!
p$.05.
841
JOB DEMANDS, RESOURCES, ENGAGEMENT, AND BURNOUT

tervals for the specific types of demands all overlap with those of
challenge demands in general reported in Table 2. This indicates
that the positive relationship of challenge demands with engage-
ment is robust across the specific types of challenge demands, or
in other words, the specific types of challenge demands do not
exhibit differential relationships with engagement. In terms of
whether the relationships of challenge demands with engagement
are significantly weaker than the relationships of job resources
with engagement, the confidence interval for overall challenge
demands [.11, .21] does not overlap with that of overall job
resources [.33, .39], indicating that job resources have a signifi-
cantly stronger relationship with engagement. However, when
analyzed across the narrower types, this conclusion is not without
exception, as the confidence interval for the challenge demand
time urgency [.10, .31] overlaps with those of the job resources
feedback [.29, .40], opportunities for development [.30, .63], pos-
itive workplace climate [.23, .33], recovery [.07, .51], rewards and
recognition [.06, .37], and support [.29, .36]. Further, the confi-
dence intervals for the challenge demands job responsibility [.14,
.17] and workload [.09, .18] both overlap with those of recovery
[.07, .51] and rewards and recognition [.06, .37]. In the aggregate,
although resources generally have a stronger relationship with
engagement than do challenge demands, this difference is not
necessarily universal in terms of specific resources and challenges.
A pattern of consistently negative correlations with engagement
is observed across the six different types of hindrance demands.
Administrative hassles (#!".17, p$.05), emotional conflict
(#!".19, p$.05), organizational politics (#!".25, p$.05),
resource inadequacies (#!".18, p$.05), role conflict (#!
".24, p$.05), and role overload (#!".20, p$.05) all had
negative relationships with engagement. All of the confidence
intervals exclude zero, and the majority of the confidence intervals
overlap. The confidence interval for organizational politics [".27,
".23], however, did not overlap with those of administrative
hassles [".23, ".12], emotional conflict [".22, ".15], and re-
source inadequacies [".22, ".15]. However, this result should be
interpreted cautiously, as the results for organizational politics
were derived from only four correlations. All of the credibility
intervals for the specific types of hindrance demands exclude zero.
Thus, it appears that the negative relationship of hindrance de-
mands with engagement is robust across the specific types of
hindrance demands. The specific types of hindrance demands
generally do not exhibit differential relationships with engage-
ment.
Discussion
The general purpose of our research was to examine and clarify
the job demands–resources model as theoretical basis for under-
standing how working conditions function as antecedents of en-
gagement and burnout. In doing so, we integrated knowledge from
the transactional theory of stress to resolve inconsistencies in
relationships between job demands and engagement by showing
that job demands typically appraised as challenges are consistently
positively related to engagement, whereas job demands typically
appraised as hindrances are consistently negatively related to en-
gagement. Research aimed at this purpose is important given (a)
the increasing interest in job engagement among scholars and
practitioners, (b) the parsimonious use of the job demands–
resources model as a framework to explain how working condi-
Table 2
Meta-Analytic Relationships of Specific Types of Resources, Challenge Demands, and Hindrance Demands With Engagement
Variable kN r #95% CI 80% CV
% variance
explained Q
Resources
Autonomy 32 18,344 .31 .37 [.33, .42] [.21, .53] 11.60 275.90
!
Feedback 19 12,125 .28 .35 [.29, .40] [.20, .49] 13.99 135.83
!
Opportunities for development 6 4,980 .38 .47 [.30, .63] [.20, .73] 3.10 193.74
!
Positive workplace climate 13 10,322 .23 .28 [.23, .33] [.16, .39] 17.45 74.52
!
Recovery 3 350 .26 .29 [.07, .51] [.07, .51] 23.81 12.60
!
Rewards and recognition 7 6,372 .16 .21 [.06, .37] [".04, .57] 3.91 179.12
!
Support 33 17,029 .27 .33 [.29, .36] [.23, .42] 28.57 115.51
!
Job variety 6 6,739 .42 .53 [.49, .57] [.47, .59] 31.38 19.12
!
Work role fit 6 4,559 .43 .52 [.43, .61] [.39, .66] 10.62 56.49
!
Challenge demands
Job responsibility 7 2,583 .13 .15 [.14, .17] [.15, .15] 100.00 0.63
Time urgency 9 6,561 .18 .21 [.10, .31] [.02, .40] 7.57 118.97
!
Workload 16 6,963 .11 .13 [.09, .18] [.03, .24] 35.31 45.31
!
Hindrance demands
Administrative hassles 7 7,187 ".15 ".17 [".23, ".12] [".25, ".10] 25.99 26.93
!
Emotional conflict 4 3,220 ".16 ".19 [".22, ".15] [".19, ".19] 100.00 3.00
Organizational politics 4 3,042 ".21 ".25 [".27, ".23] [".25, ".25] 100.00 0.93
Resource inadequacies 11 11,770 ".15 ".18 [".22, ".14] [".24, ".11] 33.51 32.82
!
Role conflict 12 3,689 ".20 ".24 [".29, ".18] [".34, ".14] 42.35 28.34
!
Role overload 5 6,152 ".14 ".20 [".32, ".08] [".37, ".03] 6.85 72.94
!
Note. k is the number of correlations used for the meta-analysis; Nis the combined sample size for the meta-analysis; ris the sample-weighted correlation;
#is the estimated true correlation corrected for sampling error and unreliability; 95% CI is the 95% confidence interval around the estimated true correlation;
80% CV is the 80% credibility interval around the estimated true correlation; % variance explained is the percentage of variance explained by statistical
artifacts; Qis the homogeneity statistic.
!
p$.05.
842 CRAWFORD, LEPINE, AND RICH

tions influence employee engagement and burnout in theories of
employee well-being and organizational life, and (c) the limita-
tions of the model to explain inconsistencies in results of empirical
research on this topic that have grown dramatically in the past
decade. We highlight that our research is the first to examine these
relationships simultaneously in a way that allows for the falsifica-
tion of a theoretical model in which engagement plays a central
role. We make an important contribution by showing that true
relationships exist between demands and engagement, even though
a substantial body of scholars has supposed there are none. Finally,
because other research shows that engagement is functionally
different than other attitudes and motivational concepts, we pro-
vide the evidence to test rather than assume that engagement
exhibits similar relationships with challenge and hindrance stres-
sors. Our results support our integrated model and have numerous
theoretical implications with practical value.
Theoretical Implications
At the most general level, our research has important theoretical
implications to the job demands–resources model as articulated by
scholars who have used the theory to understand the antecedents of
engagement and burnout. On the one hand, we found results that
are consistent with the model in that job demands have positive
relationships with burnout, and this is the case whether demands
are differentiated as challenges or hindrances. We found that job
resources have positive relationships with engagement and that
these relationships were fairly consistent across the various types
of resources. We also found that job resources have negative
relationships with burnout. On the other hand, we found results
that suggest updates for the job demands–resources model. Spe-
cifically, whereas the theory does not expect direct relationships
between job demands and engagement, we found that they are not
only meaningfully related but also that the direction of the rela-
tionship varies systematically as a function of the type of demand
in question. We note that by simply differentiating demands as
challenges and hindrances, we were able to increase the variance
explained in engagement by nearly half. Further, their relation-
ships with engagement were rather consistent across the specific
types of demands within the broader challenge and hindrance
categories, providing support for the utility of the job demands–
resources perspective and our extension of it as a means of parsi-
moniously depicting relationships among working conditions, en-
gagement, and burnout.
Thus, we have proposed and empirically supported a refined job
demands–resources model that requires distinguishing job de-
mands as challenges and hindrances to model true relationships
that exist between demands and engagement. Propositions of this
refined job demands–resources model, which we refer to as the
differentiated job demands–resources model, are as follows: (a)
job characteristics can be broadly categorized in terms of challenge
demands, hindrance demands, and job resources; (b) job demands,
whether challenges or hindrances, activate an energy depletion
process that builds up strain and results in increased burnout; (c)
job resources activate a motivational process that increases will-
ingness to dedicate one’s efforts and abilities to the work task
resulting in increased engagement; (d) job resources protect indi-
viduals from strains related to resource depletion accruing over
time to result in burnout; (e) challenge demands trigger positive
emotions and cognitions that result in active, problem-focused
coping styles reflected in increased engagement; (f) hindrance
demands trigger negative emotions and cognitions that result in
passive, emotion-focused coping styles reflected in decreased en-
gagement.
We acknowledge that our theoretical extension and supportive
findings for the linkages between job demands and engagement
may seem fairly straightforward given what we know from the
prior meta-analytic research on challenge and hindrance stressors.
However, this is something that becomes obvious only in hind-
sight. Indeed, all of the studies on engagement in our meta-analysis
have been published since the initial work on the challenge and
hindrance stressor framework (Cavanaugh et al., 2000), and 39 of
the 55 studies have been published since LePine et al.’s (2005)
meta-analysis, and yet not a single one includes a hypothesis that
recognizes the idea that relationships between demands and en-
gagement are a function of the nature of the demand. Further, it is
unlikely that any future individual primary study could resolve the
observed inconsistencies in relationships between demands and
engagement, as we have done in this meta-analysis, because of the
inability of primary studies to correct for sampling error and other
artifacts that may attenuate or mask true underlying relationships.
This illustrates the usefulness of our meta-analysis to clarify ex-
isting research findings and to serve as a guidepost for future
research on this topic, for without it authors of primary studies and
narrative reviews may continue to suggest there is no relationship
between job demands and engagement.
Beyond clarifying relationships among job demands, resources,
engagement, and burnout, our research suggests an additional
implication to the job demands–resources perspective that is some-
what broader in scope. Whereas the job demands–resources theory
suggests that resources primarily impact more distal behavioral
outcomes through engagement, and that demands primarily impact
more distal behavioral outcomes through strain, our research re-
sults suggest that demands may function through the engagement
pathway as well as the strain pathway. In other words, job de-
mands likely have an additional important indirect relationship
with more distal criteria that has not been specified and accounted
for in previous research from the job demands–resources perspec-
tive. Being able to account for a total effect with additional indirect
effects contributes to our theoretical knowledge because it in-
creases our understanding of why important concepts are related to
one another. Although job demands may negatively affect organi-
zational outcomes through strain, which was previously under-
stood in the job demands–resources perspective, those job de-
mands appraised as challenges may simultaneously have a positive
effect on organizational outcomes through increasing engagement,
which is a new understanding in the job demands–resources per-
spective; likewise, those job demands appraised as hindrances may
have a simultaneous negative effect on organizational outcomes
through decreased engagement, which is also a new understanding.
Because relationships among demands and engagement differ in
direction as a function of the type of demand being considered, it
would appear that it is necessary to consider the nature of the
demand when developing and testing hypotheses grounded in the
job demands–resources perspective.
Our research also has a theoretical implication related specifi-
cally to employee engagement as well. The patterns of the rela-
tionships in our model provide theoretical support for the distinc-
843
JOB DEMANDS, RESOURCES, ENGAGEMENT, AND BURNOUT

tiveness of the engagement and burnout constructs. If it were true
that engagement and burnout were on opposite poles of the same
continuum as some have suggested (e.g., Maslach & Leiter, 1997,
2008), then, because we found that challenges and hindrances were
both positively related to burnout, we should have found both
types of demands to be negatively related to engagement. This was
not the case for challenge demands, which were positively related
to both burnout and engagement. The relationships of the two
types of demands with burnout were not mirrored by opposite
relationships of these demands with engagement. This suggests
that although engagement and burnout are negatively related (#!
".48, 95% CI [".51, ".45], 80% CV [".61, ".34]), the two
constructs are not empirical opposites, and that additional variance
in relationships with antecedents can be explained by maintaining
their distinctiveness.
Limitations and Future Research
There are some limitations of our work that we should acknowl-
edge. First, the primary research we used in our meta-analysis was
not conducive to assessing the underlying mechanisms that link
job resources and demands to engagement and burnout. Future
research could address this issue by considering the intervening
theoretical processes we outlined earlier. Most important, perhaps,
researchers could examine how demands are appraised as a chal-
lenge or a hindrance and how these appraisals impact the cogni-
tions, emotions, and coping strategies that ultimately translate to
self-perceptions of engagement. Although scholars have long been
interested in the appraisal process (e.g., Lazarus & Folkman,
1984), the concept of engagement is new to this process and should
be studied further. Beyond insights in regards to the appraisal
process itself, such research could identify functional characteris-
tics of engagement that provide further evidence of its distinctive-
ness relative to other affective-motivational concepts.
Second, there are limitations related to the primary research in
this literature. For example, the modal research design is cross-
sectional, and thus our meta-analytic estimates should not be
interpreted as directly supporting the causal ordering of concepts
as suggested by job demands–resources theory. Although there are
certainly compelling theoretical reasons that support the presumed
causal ordering, strong inferences regarding causality require ex-
perimental research in which the theoretical antecedents—the re-
sources and demands—can be manipulated. Next, the majority of
the primary research used self-report measures for both indepen-
dent and dependent variables, creating the potential for same-
source bias to inflate the size of the reported meta-analytic rela-
tionships. This concern is minimized by findings from Crampton
and Wagner’s (1994) meta-analysis of over 42,000 correlations
from a 25-year period indicating that inflation from same-source
bias occurs more as the exception than the rule. Further, their
analyses revealed no significant differences in the size of same-
source versus multi-source correlations in subsample clusters of
relationships between motivation, stress, and work characteristics
similar to those under consideration in the present research. Thus,
although we cannot rule out the presence of same-source bias, it is
not likely that our findings could be explained or our conclusions
altered solely because of same-source bias in the primary research.
Finally, because of the limited primary research examining certain
relationships, our meta-analysis included some estimates based on
a relatively small number of studies. This was especially true in
our supplemental analyses in which we considered narrowly de-
fined resources and demands. Although meta-analyses produce
estimates of true relationships that are superior to those relation-
ships resulting from data collected in individual studies, it is likely
that additional primary research will result in estimates that are
more precise. However, given that the additional studies should
reduce the size of the confidence intervals around the estimates,
the additional studies would not likely result in changes that
contradict the support we found for our hypotheses.
A characteristic of the literature on job demands and resources
is that it generally focuses on perceived working conditions with-
out considering the role of objective job characteristics or job-level
aggregate measures. Although Hackman and Lawler (1971) argued
that, in terms of working conditions, it is not their objective state
that affects employee attitudes and behaviors but rather how em-
ployees experience them that affects their reactions to the job,
other researchers have suggested there may be value in considering
objective job characteristics to test theories of how working con-
ditions influence affective and motivational job reactions (Spector
& Jex, 1991). Although our findings on the relationships between
perceived working conditions and engagement and burnout are
important and interesting in and of themselves, future research
designs grounded in the job demands–resources perspective could
incorporate the use of job analysis, independent raters, or other
external rating sources—such as the occupational information net-
work (O*NET)—to provide multiple non-job-incumbent measures
of working conditions to examine the influence of objective job
characteristics on engagement and burnout that may be potentially
mediated through their relationships with perceived working con-
ditions.
A limitation of the job demands–resources model is that it does
not include all relevant predictors of employee engagement or
burnout. Its greatest use is to broadly categorize working condi-
tions as either resources or demands in predicting engagement, and
to the extent that the relationships of resources, challenge de-
mands, and hindrance demands are consistent within these over-
arching categories, the model and our extension of it summarizes
these relationships in a concise way. However, other perspectives
have proposed relevant predictors that are not included in the job
demands–resources framework. For example, Kahn (1990) origi-
nally proposed that the three main predictors of engagement were
the psychological conditions of meaningfulness, safety, and avail-
ability; however, these psychological conditions fall outside the
scope of the job demands–resources model. Likewise, Macey and
Schneider (2008) proposed that traits such as conscientiousness,
proactive personality, and positive affectivity are important indi-
vidual difference predictors of engagement; and Britt (2003)
showed that aspects of personal identity relevant to the work
domain also predict engagement. Yet, the job demands–resources
model cannot incorporate individual differences as a job demand
or a job resource, as these characteristics emanate from the indi-
vidual and not the characteristics of the job. Finally, researchers
have argued that transformational leadership is a key factor in
promoting engagement (Harter et al., 2002; Macey & Schneider,
2008); however, the job demands–resources model does not incor-
porate transformational leadership directly as either a demand or
resource. It captures leadership at best indirectly through narrower
variables such as supervisor support, feedback, and coaching,
844 CRAWFORD, LEPINE, AND RICH

which do not constitute the leadership construct broadly. Thus,
future research is necessary to summarize and clarify the relation-
ships of engagement with all its studied antecedents, including
those beyond working conditions.
A possible contingency exists regarding our observed pattern of
relationships between challenge demands, engagement, and burn-
out. It may be possible that the appraisal of demand as a challenge
can change as a result of experience over time, and as a conse-
quence, the relationships between the demand and both burnout
and engagement may change as well. For example, individuals
might come to believe over time that the demands they face are
more of a hindrance than a challenge (i.e., if the demands cannot
be addressed, they become threatening to the self and a hindrance
to accomplishing a sense of fulfillment), and as a consequence, the
relationship between the level of the demands and engagement
becomes negative. We also acknowledge that over time the level of
burnout may become so overwhelming that it becomes impossible
for the individual to marshal any personal resources to deal with
his or her job. Unfortunately, research conducted to date has used
cross-sectional designs, and we are limited in being able to assess
these ideas. Accordingly, we urge researchers to use experience-
sampling methods to elucidate the boundary conditions of how
challenges are related to engagement and burnout.
Finally, as alluded to earlier, there is a need for studies that
extend the differential relationships of challenge and hindrance
demands to important behavioral outcomes—such as job perfor-
mance, withdrawal, and turnover—with engagement and burnout
as dual mediators. The limited research that has linked job de-
mands to outcomes (e.g., Bakker, Demerouti, & Verbeke, 2004;
Schaufeli & Bakker, 2004) has not differentiated between chal-
lenges or hindrances, and it has generally focused on burnout as
the lone mediator to transmit those effects, assuming that job
demands have no relationship with engagement. Our research
shows that once job demands are differentiated as challenges and
hindrances, challenge demands have a positive relationship with
engagement, whereas hindrance demands have a negative relation-
ship with engagement. Future primary research is necessary that
incorporates this distinction of demands and specifies relationships
with organizationally valued criteria through both burnout and
engagement as dual mediators. With an increase in hindrance
demands, organizationally valued criteria are not only likely to be
negatively impacted though an increase in burnout but also by a
decrease in engagement. With an increase in challenge demands,
organizationally valued criteria are likely to be negatively im-
pacted through an increase in burnout, but this negative effect
should be offset by an increase in engagement. Past research
suggests that the positive indirect path can offset the negative
indirect path, and accordingly, challenge demands may have net
positive relationships with the more distal criteria (e.g., LePine et
al., 2005; Podsakoff et al., 2007). Although this line of reasoning
needs to be directly addressed in future research, our findings also
suggest this may be the case, as challenge demands exhibited a
stronger effect on engagement (&!.22, p$.05) than they did on
burnout (&!.10, p$.05). In sum, our research implies that to
understand relationships among job demands on the one hand, and
behavioral criteria such as job performance, withdrawal, and turn-
over on the other hand, it is necessary to account for a more
complex and nuanced system of relationships than has previously
been considered.
Practical Implications
Notwithstanding the limitations we noted above, there are some
potentially important practical implications of our research. First,
it is clear that individual employee engagement is positively asso-
ciated with perceived resources and that individual burnout is
positively associated with perceived demands. Although we cannot
rule out the possibility that employees who report higher levels of
engagement receive additional resources such as supervisor sup-
port and rewards in the form of more job variety, it does seem
reasonable at this point to suggest that a manager could potentially
foster an individual’s employee engagement by providing addi-
tional resources. Similarly, although we cannot rule out that em-
ployees who feel burned out simply perceive that they have greater
job demands, it is also reasonable to suggest that managers could
potentially reduce an employee’s burnout by limiting the demands
the employee must cope with, especially those demands perceived
as hindrances or roadblocks. Likewise, although we cannot rule
out that burned out employees simply feel they have fewer re-
sources at their disposal, it is reasonable to suggest that an addi-
tional reason managers may wish to provide more resources for
their employees is that doing so has the potential to reduce em-
ployee burnout because of the strain of having to meet demands
with few resources. This suggests that in situations in which
reducing demands is not a possibility, such as in a high workload
environment, the strains of dealing with such demands may be
reduced by providing additional resources.
Second, whereas job demands–resources theory suggests that
job demands are not associated with engagement, our results
suggest otherwise, and this opens the possibility that individual
employee engagement could be influenced by managers through
practices aimed at changing the level of demands confronted by
employees. However, it is important to emphasize that decisions to
adjust levels of demands to influence individuals’ engagement are
more complex and depend on the type of the demands being
considered. On the one hand, our results suggest that individual
employee engagement could be facilitated by a manager’s attempts
to reduce perceived hindrance demands such as administrative
hassles, politics, and role conflicts. Although the positive relation-
ship between these types of hindrance job demands and strain is
well known, our results suggest that an additional reason why
managers should attempt to minimize them is because of their
negative relationship with employee engagement. On the other
hand, our results suggest that individual employee engagement
might be promoted in contexts in which the level of perceived
challenge demands is increased. However, given that challenge job
demands are also positively associated with strains, it would be
premature to suggest that challenge job demands be increased
simply for the sake of promoting employee engagement. Before
such a recommendation could be made, it may be necessary to
identify how the associated increase in strain could be reduced. As
we have suggested earlier, it may be possible to identify commen-
surate levels of increased resources that could be put in place to
reduce the strain associated with the increases in challenge de-
mands.
Conclusion
We have provided the first quantitative summary of the state of
the research examining the influence of working conditions on
845
JOB DEMANDS, RESOURCES, ENGAGEMENT, AND BURNOUT

engagement and burnout organized according to the propositions
of the job demands–resources model. We have also refined and
updated the model with the transactional theory of stress to suggest
that relationships between job demands and engagement vary as a
function of the nature of the demand with respect to how it tends to
be appraised by employees. Our meta-analyses show that job de-
mands are indeed associated with increased burnout, that job
resources are associated with increased engagement, and that job
resources are also associated with decreased burnout. Our meta-
analyses also confirm our reasoning that in regards to engagement,
not all demands are created equal. When it comes to increasing
levels of employee engagement, it appears that demands appraised
as challenges tend to help, and demands appraised as hindrances
demands tend to hurt.
References
References marked with an asterisk indicate studies included in the
meta-analysis.
*Bakker, A. B., & Bal, P. M. (2007). Work engagement and performance:
A diary study among starting teachers. Unpublished manuscript.
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., de Boer, E., & Schaufeli, W. B. (2003). Job
demands and job resources as predictors of absence duration and fre-
quency. Journal of Vocational Behavior, 62, 341–356.
Bakker, A. B., Demerouti, E., & Euwema, M. C. (2005). Job resources
buffer the impact of job demands on burnout. Journal of Occupational
Health Psychology, 10, 170–180.
Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2003). Dual processes
at work in a call centre: An application of the job demands–resources
model. European Journal of Work & Organizational Psychology, 12,
393–417.
*Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2005). The crossover
of burnout and work engagement among working couples. Human
Relations, 58, 661–689.
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., 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.
*Bakker, A. B., van Emmerik, H., & Euwema, M. C. (2006). Crossover of
burnout and engagement in work teams. Work & Occupations, 33,
464–489.
*Bakker, A. B., van Emmerik, I. J. H., Geurts, S. A. E., & Demerouti, E.
(2008). Recovery turns job demands into challenges: A diary study on
work engagement and performance. Unpublished manuscript.
*Beckers, D. G. J., van der Linden, D., Smulders, P. G. W., Kompier,
M. A. J., van Veldhoven, M. J. P., & van Yperen, N. W. (2004).
Working overtime hours: Relations with fatigue, work motivation, and
the quality of work. Journal of Occupational & Environmental Medi-
cine, 46, 1282–1289.
*Bledow, R., & Schmitt, A. (2008, April). Work engagement as a dynamic
process: Events, emotions, and resources. Paper presented at the 23rd
Annual Conference for the Society for Industrial and Organizational
Psychology, San Francisco, CA.
Boswell, W. R., Olson-Buchanan, J. B., & LePine, M. A. (2004). Relations
between stress and work outcomes: The role of felt challenge, job
control, and psychological strain. Journal of Vocational Behavior, 64,
165–181.
Brief, A. P., & George, J. M. (1995). Psychological stress and the work-
place: A brief comment on Lazarus’ outlook. Philadelphia, PA: Taylor &
Francis.
*Britt, T. W. (1999). Engaging the self in the field: Testing the triangle
model of responsibility. Personality and Social Psychology Bulletin, 25,
696–706.
*Britt, T. W. (2003). Aspects of identity predict engagement in work under
adverse conditions. Self & Identity, 2, 31–45.
*Britt, T. W. (2009). [Engagement and work perceptions survey]. Unpub-
lished raw data.
Britt, T. W., Adler, A. B., & Bartone, P. T. (2001). Deriving benefits from
stressful events: The role of engagement in meaningful work and har-
diness. Journal of Occupational Health Psychology, 6, 53–63.
*Britt, T. W., & Bliese, P. D. (2003). Testing the stress–buffering effects
of self-engagement among soldiers on a military operation. Journal of
Personality, 71, 245–266.
*Britt, T. W., Castro, C. A., & Adler, A. B. (2005). Self-engagement,
stressors, and health: A longitudinal study. Personality and Social Psy-
chology Bulletin, 31, 1475–1486.
*Britt, T. W., Thomas, J. L., & Dawson, C. R. (2006). Self-engagement
magnifies the relationship between qualitative overload and performance
in a training setting. Journal of Applied Social Psychology, 36, 2100–
2114.
Cavanaugh, M. A., Boswell, W. R., Roehling, M. V., & Boudreau, J. W.
(2000). An empirical examination of self-reported work stress among
U.S. managers. Journal of Applied Psychology, 85, 65–74.
Crampton, S. M., & Wagner, J. A. (1994). Percept–percept inflation in
microorganizational research: An investigation of prevalence and effect.
Journal of Applied Psychology, 79, 67–76.
Deci, E. L., Ryan, R. M., Gagne´, M., Leone, D. R., Usunov, J., &
Kornazheva, B. P. (2001). Need satisfaction, motivation, and well-being
in the work organizations of a former eastern bloc country: A cross-
cultural study of self-determination. Personality and Social Psychology
Bulletin, 27, 930–942.
*DeGraff, J. E. (2007). Improving employee attitudes through HR pro-
grams and recognition practices. Unpublished bachelor’s honors thesis,
Brigham Young University, Provo, UT.
*Demerouti, E., Bakker, A. B., de Jonge, J., Janssen, P. P. M., & Schaufeli,
W. B. (2001). Burnout and engagement at work as a function of demands
and control. Scandinavian Journal of Work, Environment & Health, 27,
279–286.
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.
Erez, A., & Isen, A. M. (2002). The influence of positive affect on the
components of expectancy motivation. Journal of Applied Psychology,
87, 1055–1067.
*Fluegge Woolf, E. R. (2008). Who put the fun in functional? Fun at work
and its effects on job performance. Unpublished doctoral dissertation,
University of Florida, Gainesville.
Fried, Y., & Ferris, G. R. (1987). The validity of the job characteristics
model: A review and meta-analysis. Personnel Psychology, 40, 287–
322.
*Gorter, R. C., Te Brake, H. J., Hoogstraten, J., & Eijkman, M. A. (2008).
Positive engagement and job resources in dental practice. Community
Dentistry and Oral Epidemiology, 36, 47–54.
Hackman, J. R., & Lawler, E. E. (1971). Employee reactions to job
characteristics. Journal of Applied Psychology, 55, 259–286.
Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading, MA:
Addison-Wesley.
*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. J., Bakker, A. B., & Schaufeli, W. B. (2006). Burnout and
846 CRAWFORD, LEPINE, AND RICH

work engagement among teachers. Journal of School Psychology, 43,
495–513.
*Hakanen, J. J., Perhoniemi, R., & Toppinen-Tanner, S. (2008). Positive
gain spirals at work: From job resources to work engagement, personal
initiative and work-unit innovativeness. Journal of Vocational Behavior,
73, 78–91.
Halbesleben, J. R. B. (2006). Sources of social support and burnout: A
meta-analytic test of the conservation of resources model. Journal of
Applied Psychology, 91, 1134–1145.
*Hallberg, U. E., Johansson, G., & Schaufeli, W. B. (2007). Type A
behavior and work situation: Associations with burnout and work en-
gagement. Scandinavian Journal of Psychology, 48, 135–142.
*Hallberg, U. E., & Schaufeli, W. B. (2006). “Same same” but different?
Can work engagement be discriminated from job involvement and
organizational commitment? European Psychologist, 11, 119–127.
Harter, J. K., & Schmidt, F. L. (2008). Conceptual versus empirical
distinctions among constructs: Implications for discriminant validity.
Industrial and Organizational Psychology, 1, 36–39.
Harter, J. K., Schmidt, F. L., & Hayes, T. L. (2002). Business-unit-level
relationship between employee satisfaction, employee engagement, and
business outcomes: A meta-analysis. Journal of Applied Psychology, 87,
268–279.
Hobfoll, S. E., & Freedy, J. (1993). Conservation of resources: A general
stress theory applied to burnout. In W. B. Schaufeli, C. A. Maslach, &
T. Marek (Eds.), Professional burnout: Recent developments in theory
and research. (pp. 115–133). Philadelphia, PA: Taylor & Francis.
Hunter, J. E., & Schmidt, F. L. (1990). Methods of meta-analysis: Cor-
recting error and bias in research findings. Thousand Oaks, CA: Sage.
*Jackson, L. T. B., Rothmann, S., & van de Vijver, F. J. R. (2006). A
model of work-related well-being for educators in South Africa. Stress
& Health: Journal of the International Society for the Investigation of
Stress, 22, 263–274.
Jöreskog, K. G., & Sörbom, D. (2002). LISREL 8: Structural equation
modeling with the SIMPLIS command language (5th ed.). Lincolnwood,
IL: Scientific Software International.
Judge, T. A., Ilies, R., Bono, J. E., & Gerhardt, M. W. (2002). Personality
and leadership: A qualitative and quantitative review. Journal of Applied
Psychology, 87, 765–780.
Kahn, W. A. (1990). Psychological conditions of personal engagement and
disengagement at work. Academy of Management Journal, 33, 692–724.
*Ku¨hnel, J., Sonnentag, S., & Westman, M. (2009). Does work engage-
ment increase after a short respite? The role of job involvement as a
double-edged sword. Journal of Occupational and Organizational Psy-
chology, 82, 575–594.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New
York, NY: Springer.
Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the
correlates of the three dimensions of job burnout. Journal of Applied
Psychology, 81, 123–133.
LePine, J. A., LePine, M. A., & Jackson, C. L. (2004). Challenge and
hindrance stress: Relationships with exhaustion, motivation to learn, and
learning performance. Journal of Applied Psychology, 89, 883–891.
LePine, J. A., Podsakoff, N. P., & LePine, M. A. (2005). A meta-analytic
test of the challenge stressor–hindrance stressor framework: An expla-
nation for inconsistent relationships among stressors and performance.
Academy of Management Journal, 48, 764–775.
*LePine, J. A., Zhang, Y., Rich, B. L., & Crawford, E. R. (2008). Under-
standing the relationship between transformational leadership and job
performance: The mediating role of overall attitude and employee
engagement. Unpublished manuscript, University of Florida, Gaines-
ville.
*Little, L. M., Simmons, B. L., & Nelson, D. L. (2007). Health among
leaders: Positive and negative affect, engagement and burnout, forgive-
ness and revenge. Journal of Management Studies, 44, 243–260.
*Llorens, S., Bakker, A. B., Schaufeli, W., & Salanova, M. (2006). Testing
the robustness of the job demands–resources model. International Jour-
nal of Stress Management, 13, 378–391.
*Llorens, S., Schaufeli, W., Bakker, A., & Salanova, M. (2007). Does a
positive gain spiral of resources, efficacy beliefs and engagement exist?
Computers in Human Behavior, 23, 825–841.
*Lorente, L., Salanova, M., Martı´nez, I. M., & Schaufeli, W. B. (2008).
Extension of the job demands–resources model in the prediction of
burnout and engagement among teachers over time. Psicothema, 20,
354–360.
Macey, W. H., & Schneider, B. (2008). The meaning of employee engage-
ment. Industrial and Organizational Psychology, 1, 3–30.
Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach Burnout
Inventory manual (3rd ed.). Palo Alto, CA: Consulting Psychologists
Press.
Maslach, C., & Leiter, M. P. (1997). The truth about burnout: How
organizations cause personal stress and what to do about it (1st ed.).
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.
*Mauno, S., Kinnunen, U., Ma¨kikangas, A., & Na¨tti, J. (2005). Psycho-
logical consequences of fixed-term employment and perceived job in-
security among health care staff. European Journal of Work & Organi-
zational Psychology, 14, 209–237.
*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 engage-
ment of the human spirit at work. Journal of Occupational and Orga-
nizational Psychology, 77, 11–37.
Meyer, J. P., & Gagne´, M. (2008). Employee engagement from a self-
determination theory perspective. Industrial and Organizational Psy-
chology, 1, 60–62.
*Montgomery, A. J., Peeters, M. C. W., Schaufeli, W. B., & Den Ouden,
M. (2003). Work–home interference among newspaper managers: Its
relationship with burnout and engagement. Anxiety, Stress & Coping, 16,
195–211.
*Mostert, K., & Rothmann, S. (2006). Work-related well-being in the
South African police service. Journal of Criminal Justice, 34, 479–491.
*Naude´, J. L. P., & Rothmann, S. (2006). Work-related well-being of
emergency workers in Gauteng. South African Journal of Psychology,
36, 63–81.
*Nembhard, I. M., & Edmondson, A. C. (2006). Making it safe: The effects
of leader inclusiveness and professional status on psychological safety
and improvement efforts in health care teams. Journal of Organizational
Behavior, 27, 941–966.
Podsakoff, N. P., LePine, J. A., & LePine, M. A. (2007). Differential
challenge stressor–hindrance stressor relationships with job attitudes,
turnover intentions, turnover, and withdrawal behavior: A meta-analysis.
Journal of Applied Psychology, 92, 438–454.
Porath, C. L., & Erez, A. (2009). Overlooked but not untouched: How
rudeness reduces onlookers’ performance on routine and creative tasks.
Organizational Behavior & Human Decision Processes, 109, 29–44.
*Rich, B. L. (2006). Job engagement: Construct validation and relation-
ships with job satisfaction, job involvement, and intrinsic motivation.
Unpublished doctoral dissertation, University of Florida, Gainesville.
Rich, B. L., LePine, J. A., & Crawford, E. R. (in press). Job engagement:
Antecedents and effects on job performance. Academy of Management
Journal.
*Richardsen, A. M., Burke, R. J., & Martinussen, M. (2006). Work and
health outcomes among police officers: The mediating role of police
cynicism and engagement. International Journal of Stress Management,
13, 555–574.
847
JOB DEMANDS, RESOURCES, ENGAGEMENT, AND BURNOUT

*Rothbard, N. P. (2001). Enriching or depleting? The dynamics of engage-
ment 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.
*Saks, A. M. (2006). Antecedents and consequences of employee engage-
ment. Journal of Managerial Psychology, 21, 600–619.
*Salanova, M., Agut, S., & Peiro´, J. M. (2005). Linking organizational
resources and work engagement to employee performance and customer
loyalty: The mediation of service climate. Journal of Applied Psychol-
ogy, 90, 1217–1227.
*Salanova, M., & Schaufeli, W. B. (2008). A cross-national study of work
engagement as a mediator between job resources and proactive behav-
iour. International Journal of Human Resource Management, 19, 116–
131.
*Salanova, M., Schaufeli, W. B., Martı´nez, I. M., & Breso´, E. (2007). How
obstacles and facilitators predict academic performance: The mediating
role of students’ burnout and engagement. Unpublished manuscript.
*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., & van Rhenen, W. (2007). How changes
in job demands and resources predict burnout, work engagement, and
sickness absence. Unpublished manuscript.
Schaufeli, W. B., Leiter, M. P., Maslach, C., & Jackson, S. E. (1996).
Maslach Burnout Inventory—General Survey (MBI–GS). In C. A.
Maslach, S. E. Jackson, & M. P. Leiter (Eds.), MBI manual (3rd ed., pp.
19–26). Palo Alto, CA: Consulting Psychologists Press.
Schaufeli, W. B., Salanova, M., Gonza´lez-Roma´, 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., & van Rhenen, W. (2008). Workaholism,
burnout, and work engagement: Three of a kind or three different kinds
of employee well-being? Applied Psychology, 57, 173–203.
Schneider, B., Macey, W. H., Barbera, K. M., & Martin, N. (2009). Driving
customer satisfaction and financial success through employee engage-
ment. People & Strategy, 32(2), 22–27.
*Simon, L. S., Methot, J. R., & Colquitt, J. A. (2007). [Employee attitude
and supervisor performance rating survey]. Unpublished raw data.
*Sonnentag, S. (2003). Recovery, work engagement, and proactive behav-
ior: A new look at the interface between nonwork and work. Journal of
Applied Psychology, 88, 518–528.
Spector, P. E., & Jex, S. M. (1991). Relations of job characteristics from
multiple data sources with employee affect, absence, turnover intentions,
and health. Journal of Applied Psychology, 76, 46–53.
*Thomas, C. (2007, August). An expanded model of employee engage-
ment: An organizational test of determinants and outcomes. Paper pre-
sented at the annual meeting of the Academy of Management, Philadel-
phia, PA.
*Vansteenkiste, M., Neyrinck, B., Niemiec, C. P., Soenens, B., De Witte,
H., & Van Den Broeck, A. (2007). On the relations among work value
orientations, psychological need satisfaction and job outcomes: A self-
determination theory approach. Journal of Occupational & Organiza-
tional Psychology, 80, 251–277.
Viswesvaran, C., & Ones, D. S. (1995). Theory testing: Combining psy-
chometric meta-analysis and structural equations modeling. Personnel
Psychology, 48, 865–885.
Wellins, R. S., Bernthal, P., & Phelps, M. (n.d.). Employee engagement: The
key to realizing competitive advantage. Retrieved September 3, 2008, from
http://www.ddiworld.com/pdf/ddi_employeeengagement_mg.pdf
Whitener, E. M. (1990). Confusion of confidence intervals and credibility
intervals in meta-analysis. Journal of Applied Psychology, 75, 315–321.
*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.
(2009a). Reciprocal relationships between job resources, personal re-
sources, and work engagement. Journal of Vocational Behavior, 74,
235–244.
*Xanthopoulou, D., Bakker, A. B., Demerouti, E., & Schaufeli, W. B.
(2009b). Work engagement and financial returns: A diary study on the
role of job and personal resources. Journal of Occupational and Orga-
nizational Psychology, 82, 183–200.
*Xanthopoulou, D., Bakker, A. B., Heuven, E., Demerouti, E., &
Schaufeli, W. B. (2008). Working in the sky: A diary study on work
engagement among flight attendants. Journal of Occupational Health
Psychology, 13, 345–356.
Received June 11, 2009
Revision received February 10, 2010
Accepted February 15, 2010 !
848 CRAWFORD, LEPINE, AND RICH