Academy o/ Management Review

1993. Vol. 18, No. A. 621-656.

A REVIEW AND AN INTEGRATION OF RESEARCH

ON JOB BURNOUT
CYNTHIA L. CORDES
Binghamton University (SUNY)
i THOMAS W. DOUGHERTY
University of Missouri

Burnout is a unique type of stress syndrome, characterized by emo-

tional exhaustion, depersonalization, and diminished personal ac-
complishment. Although burnout has been shown to be potentially
very costly in the helping professions, such as nursing, education,
and social work, little work has been done thus iar to establish its
generalizability to industry. This article reviews the literature on
burnout and provides a conceptual framework designed to improve
the understanding of burnout. Propositions are presented that are
aimed at clarifying the dynamics of burnout, including determinants
of and interrelationships among the three burnout components.

Empirical evidence has shown that burnout has important dysfunc-

tional ramifications, implying substantial costs for both organizations
and individuals because of, for example, increases in turnover, absen-
teeism, reduced productivity, and human considerations {Jackson &
Maslach, 1982; Leiter & Maslach, 1988; Shirom, 1989). That same research
points out the need to understand the many factors and conditions that
contribute to burnout in a wide variety of service, industrial, and corpo-
rate settings. The effective implementation of individual, managerial,
and organizational practices to deal with burnout critically depends on
managers' clear and accurate understanding of the construct.
The purposes of this article are to examine the construct of burnout,
to consider how this construct has evolved, and to provide an overview of
the literature on the antecedents and consequences of bumout. In addi-
tion, propositions are presented regarding the process of burnout, the
determinants of burnout, and the interrelationships among the burnout
components. The study of burnout has been unnecessarily limited to the
helping professions, but it is experienced by a variety of occupational
groups beyond nurses, teachers, and social workers. This article points to
a more widespread occurrence of bumout than has previously been dis-

The authors thank Jim Wall, George Dreher, Dan Turban, Fran Yammarino, and Bruce
Avoiio for thoughtful suggestions on earlier drafts of this article.

821
622 Academy of Management Review October

cussed, and it presents burnout as a key construct in understanding stress

processes in many jobs and in organizations of all types.
This update focuses primarily on empirical studies that conceptual-
ized job burnout according to Maslach's three-component syndrome and
measured it with the Maslach Burnout Inventory (MBI). This focus facili-
tates comparisons across studies. Thus, the studies reviewed here (a) are
concerned with job burnout (thereby excluding, for instance, research on
student or athlete burnout); (b) define burnout as a response syndrome of
emotional exhaustion, depersonalization, and reduced personal accom-
plishment; (c) assess burnout using the most frequently used measure, the
MBI; and (d) empirically test the degree of association between burnout
and its antecedents or consequences. Studies were located by both man-
ual and computer searches, and they were published in major social
science journals primarily between 1983 and 1992. (Readers can review
some of the earlier literature in Perlman and Hartman, 1982.) Nonempir-
ical articles or those appearing prior to 1983 were included only as they
reflect the theoretical development of the area. (Readers might also be
interested in related forthcoming work by Schaufeli, Maslach, and Marek
that was not available when this article was being written.)
Although we support the conceptualization of burnout as a three-
pronged syndrome, we frequently refer to burnout and not to the specific
components. We do not believe the use of such an umbrella term is ap-
propriate for purposes of investigation, but we do believe it is acceptable
when discussing the phenomenon more generally. Thus, when we dis-
cuss results of empirical studies we generally are specific regarding the
relationships between the independent variables and the particular burn-
out components. At other times, however, for ease of communication, we
refer to burnout or the components o/ Jburnouf.

BACKGROUND
Research on burnout began as a result of work conducted on emotion,
arousal, and the way in which people cope with, or manage, the arousal
(Maslach & Jackson, 1984). Interviews were conducted with people
deemed most likely to experience the phenomena in question, namely
health care professionals. From the interviews emerged the realization
that the emotional stress inherent in these occupations often could be
excessively harmful or debilitating. This stream of investigation dove-
tailed with a similar set of phenomena in the legal services—phenomena
that lawyers who worked with the poor called burnout (Maslach & Jack-
son, 1984). In addition to providing a useful label, the parallel finding
suggested that the emotional strain was not unique to the health care
professions. The conclusion was that there was evidently something
unique about "people work" that could result in burnout.
In the next phase of development, researchers combined pieces of
information resulting from these early explorations into a working set of
1893 Coides and Dougherty 623

hypotheses and designed systematic studies to examine them. Much of

this work consisted of authors' personal experiences (e,g., Freuden-
berger, 1974, 1977a, 1977b) or narratives based on specific programs or
case studies (e.g., Maslach & Pines, 1977; Pines & Maslach, 1978, 1980),
and it was characterized by conceptual disagreement. Perlman and Hart-
man (1982) compiled a listing of the multiple conceptualizations used
during this period. Definitions of burnout included (a) to fail, wear out,
become exhausted; (b) a loss of creativity; (c) a loss of commitment for
work; (d) an estrangement from clients, co-workers, job, and agency; (e) a
response to the chronic stress of making it to the top; and finally (f) a
syndrome of inappropriate attitudes toward clients and toward self, often
associated with uncomfortable physical and emotional symptoms. Al-
though these notions were similar, they lacked a common and precise
measure of burnout.
It was not until the late 1970s and early 1980s that systematic empir-
ical studies on burnout were conducted and published (e.g., Iwanicki &
Schwab, 1981; Maslach & Jackson, 1981). During these years, the concept
of burnout was more clearly conceptualized and defined. Also during this
period, an accepted, standardized, and psychometrically sound instru-
ment, the Maslach Burnout Inventory, was developed to measure burnout
(Maslach & Jackson, 1981, 1986).
The Concept of Burnout
Today, the most commonly accepted definition of burnout is the three-
component conceptualization used by Maslach and colleagues (Maslach,
1982; Maslach & Jackson, 1981; Pines & Maslach, 1980). One component of
burnout, emofionaJ exhaustion, is characterized by a lack of energy and a
feeling that one's emotional resources are used up. This "compassion
fatigue" may coexist with feelings of frustration and tension as workers
realize they cannot continue to give of themselves or be as responsible for
clients as they have been in the past. A common symptom is dread at the
prospect of returning to work for another day. Another component, depei-
sonalization or dehumanization, is marked by the treatment of clients as
objects rather than people. Workers may display a detached and an emo-
tional callousness, and they may be cynical toward co-workers, clients,
and the organization. Visible symptoms include the use of derogatory or
abstract language (e.g., the "kidney" in room 212), strict compartmental-
ization of professional lives, intellectualization of the situation, with-
drawal through longer breaks or extended conversations with co-workers,
and extensive use of jargon (Maslach & Pines, 1977). One classic analogy
is that of a petty bureaucrat, going strictly "by the book" to deal with
individual clients rather than becoming personally involved enough to
tailor a solution or an approach to the client's needs {Daley, 1979).
The final component of burnout, diminished personaJ accomplish-
ment, is characterized by a tendency to evaluate oneself negatively. In-
dividuals experience a decline in feelings of job competence and success-
624 Academy oi Management Review October

ful achievement in their work or interactions with people. Frequently

there is the perception of a lack of progress or even lost ground. The
problem employee who is routinely receiving disciplinary citations from
the supervisor, or the feelings one might imagine if one were bailing out
a leaky boat and realized that the boat may sink, are symptomatic of this
component.
Burnout as a Process
The process of burnout, or the sequencing of the three components of
burnout, has been conceptualized differently among researchers of burn-
out. Maslach (1978, 1982) originally suggested that emotional exhaustion
appears first as excessive chronic work demands drain individuals' emo-
tional resources. Perhaps as a defensive coping strategy, they limit their
involvement with others and distance themselves psychologically. This
depersonalization provides an emotional buffer between the individual
and the imposing job demands. Finally, individuals recognize the dis-
crepancy between their current attitude and their original optimistic ex-
pectations about their potential contributions to society and to the agency
or organization. As a result, individuals experience a sense of inade-
quacy in terms of their ability to relate to people and to perform their jobs.
This sequence is based on findings from initial interviews, surveys, and
observations. Although she later retreated from such a process model
(Maslach & Jackson, 1981, 1986), Maslach has reasserted the appropriate-
ness of this original sequencing (Leiter & Maslach, 1988). Studies by Leiter
(1988), Leiter and Meechan (1986), and Leiter and Maslach (1988) provide
support for such a model, whereas a study by Lee and Ashforth (1993)
found only partial support. Others have argued, however, that there is no
fixed sequence; one component is not an inevitable consequence of an-
other (Schwab & Iwanicki, 1982a).
Another conceptualization of the burnout process is a sequence that
was advanced by Golembiewski and Munzenrider (1981, 1984) and Go-
lembiewski (1989). They hypothesized that significant depersonalization
is necessary to diminish feelings of personal accomplishment, and sig-
nificant reductions in personal accomplishment are necessary to result in
high levels of emotional exhaustion. Using the median to divide each
subscale into high and low groups, they generated eight phases based on
a "rule of sequential potencies"; that is, the effect of each component is
hypothesized to depend upon its place in the sequence. Each phase is
characterized by some combination of either high or low emotional ex-
haustion, depersonalization, and personal accomplishment, and it corre-
sponds to a higher total burnout level or score than the preceding phase.
This methodology has been used in studies of police officers (Burke &
Deszca, 1986; Burke, Shearer, & Deszca, 1984b), life care workers at a
retirement community (Rountree, 1984), teachers (Burke & Greenglass,
1989), middle managers across the public and private sector (Cahoon &
Rowney, 1989), and supervisors, managers, and executives from a variety
1993 Cordes and Dougherty 625

of types of organizations (Cahoon & Rowney, 1984). (See the Journal of

Health and Human Resources Administration. 1986, volume 9, and 1991,
volume 13, for additional discussion and support of the phase model of
burnout.) These studies are problematic, however, because the authors
used a cross-sectional research design to draw explicit conclusions
regarding causality or the process of burnout.
Burnout as a Type of Stress
Despite the growing consensus surrounding the concept of burnout,
the distinction between burnout and stress has not been clearly delin-
eated. It appears that, as Ganster and Schaubroeck (1991) have argued,
burnout is, in fact, a type of stress—specifically, a chronic affective re-
sponse pattern to stressful work conditions that features high levels of
interpersonal contact.
Similar to the case of burnout, several different conceptualizations of
stress have been proffered. Although most researchers define stress as an
outgrowth of person-environment interactions or "fit" (e.g., French &
Caplan, 1972; McGrath, 1976; Schuler, 1980) or as a result of dysfunctional
role relationships (Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964), there
has traditionally been little definitional or operational agreement among
job stress conceptualizations (Schuler, 1980). On this basis, a number of
authors advocated the treatment of stress as a general concept that can
provide a "framework" for research on a number of problems (e.g.,
McGrath, 1976).
Building on the work of McGrath (1976), however, Schuler (1980) pro-
vided some much needed conceptual clarity, which has been widely ac-
cepted, and which has cleared up much of the previous confusion about
the essential qualities of job stress. McGrath and Schuler defined stress
as a dynamic condition in which an individual is confronted with an
opportunity, constraint, or demand on being/having/doing what one
desires and for which resolution is perceived to have uncertainty but
which will lead to important outcomes. This conceptualization can sub-
sume many different stress problems, including burnout. Burnout is a
distinctive aspect of stress in that it has been defined and studied pri-
marily as a pattern of responses to stressors at work (Shirom, 1989). The
burnout response syndrome begins to a great extent as a result of
demands (Schuler and McGrath's terminology), including interpersonal
stressors. Thus, burnout represents a particular type of job stress, in
which a pattern of emotional exhaustion, depersonalization, and dimin-
ished personal accomplishment (strains) result from a variety of work
demands (stressors), especially those of an interpersonal nature. More-
over, the concepts of uncertainty and of the importance of outcomes
should be as relevant to the experience of burnout as to other kinds of
stress responses generated by various work demands and constraints.
The three-component model that burnout represents is unique as a
stress phenomenon. At its core is emotional exhaustion, which is a tra-
626 Academy of Management Review October

ditional stress variable. The second component, depersonalization, is a

new construct, not formerly appearing in the stress literature (Jackson,
Schwab, & Schuler, 1986). Finally, although feelings of personal accom-
plishment (related to such concepts as self-efficacy) are familiar to the
stress literature, the third component of burnout, a diminished level of
this variable, adds the assertion that self-evaluations are central to the
stress experience.

BURNOUT: MEASUREMENT AND CONSTRUCT VALIDITY

In the initial development of the Maslach Burnout Inventory (Maslach

6 Jackson, 1981), 47 items were administered to a sample of 605 individ-
uals from a wide variety of service occupations. Ten factors emerged and,
based on a variety of elimination criteria, four factors, composed of 25
items, were then administered to a new sample of 420 individuals from
similar occupations. The same four factors emerged from this confirma-
tory sample, three of which had eigenvalues greater than 1. These three
factors, composed of 22 items, make up the subscales of the MBI. Respon-
dents indicate the frequency and the intensity with which each item is
experienced, with the scale ranging from 1 ("a few times a year") to 6
("every day") for frequency, and from 1 ("very mild, barely noticeable") to
7 ("very strong, major") for intensity. A space is provided to indicate if the
item is never experienced by the respondent. Although high scores on
emotional exhaustion and depersonalization reflect high burnout, the
personal accomplishment subscale is reverse scored; thus low scores
reflect high burnout. The instrument does not measure the presence or
absence of burnout per se. Rather, experienced levels of burnout fall on a
continuum. For instance, high scores on the emotional exhaustion and
depersonalization subscales and low scores on the personal accomplish-
ment subscale (this subscale is reverse scored) reflect high levels of burn-
out and vice versa. No predictions are made concerning critical threshold
levels.
Subsequent research using the MBI has revealed that the intensity
and frequency dimensions of burnout are highly correlated (Anderson &
Iwanicki, 1984; Brookings, Bolton, Brown, & McEvoy, 1985; Gaines &
Jennier, 1983), so the revised version of the MBI (Maslach & Jackson, 1986)
includes only the frequency dimension. Despite this correlation, a num-
ber of researchers have continued to use the intensity dimension alone
(Friesen, Prokop, & Sarros, 1988; Friesen & Sarros, 1989; Jackson, Turner,
& Brief, 1987; Leiter & Meechan, 1986). The user's manual of the MBI
research edition also includes normative data for a variety of demo-
graphic variables and helping occupations (Maslach & Jackson, 1986).
Researchers have analyzed the convergent and discriminant validity
of the scale in a number of ways (Maslach & Jackson, 1981, 1986). For
example, to test for convergent validity, correlations were analyzed be-
tween MBI scores and behavioral ratings made independently by a per-
1993 Cordes and DougJierfy 627

son who knew the respondent well. As predicted, co-workers' and

spouses' ratings were significantly positively correlated with the respon-
dents' own ratings of emotional exhaustion and depersonalization and
were significantly negatively correlated with the respondents' ratings of
personal accomplishment. Also, correlations were analyzed between MBI
scores and certain job characteristics hypothesized to contribute to burn-
out (e.g., a greater number of clients is positively correlated with emo-
tional exhaustion and depersonalization and negatively correlated with
personal accomplishment). Finally, correlations between MBI scores and
measures of various outcomes hypothesized to be related to burnout were
analyzed. Again, the empirical evidence is consistent with the hypothe-
sized relationships.
Support for the discriminant validity of the scale is seen in a signif-
icant, although small, positive correlation between MBI scores and job
satisfaction (Maslach & Jackson, 1986; Wolpin, Burke, & Greenglass,
1991). In addition, correlations between MBI scores and the Crowne-
Marlowe Social Desirability Scale were not significant. Thus, responses
to the burnout scale do not appear to be influenced by a social desirability
response set. Iwanicki and Schwab (1981) and Gold (1984) cross-validated
the MBI on a group of teachers, finding the same basic constructs. It
should be emphasized that a systematic assessment of the convergent
and discriminant validity using the Multitrait-Multimethod Matrix ap-
proach has not appeared in the literature.
Even though there is growing evidence of the psychometric sound-
ness of the MBI, several authors have reported high correlations between
emotional exhaustion and depersonalization (e.g., Koeske & Koeske, 1989;
Lee & Ashforth, 1990; Wolpin et al., 1991). Some researchers have found
support for a two-component model (Brookings et al., 1985), whereas oth-
ers have argued for a three-coinponent model (Green & Walkey, 1988; Lee
& Ashforth, 1990). Results of confirmatory factor analyses, however, seem
to generally support the existence of three distinct components of burnout
(Fimian & Blanton, 1987; Gold, 1984; Golembiewski & Munzenrider, 1981;
Green & Walkey, 1988; Maslach & Jackson, 1981, 1986).
It has also been suggested that the items assessing the three com-
ponents of burnout can be summed to form an overall measure of burnout
(Golembiewski & Munzenrider, 1981; Meier, 1984). Golembiewski and
Munzenrider (1981) found that a total burnout score covaried significantly
and in expected directions with a variety of variables hypothesized to be
important aspects of the worksite. Meier (1984) argued that a greater num-
ber of items would contribute to greater reliability for the burnout instru-
ment. He also argued that a total score has good internal consistency (.88
in his study).
Maslach, however, has argued against viewing burnout as a unitary
concept. Empirical evidence supports the idea that emotional exhaustion,
depersonalization, and personal accomplishment are conceptually dis-
tinct components (Iwanicki & Schwab, 1981; Maslach & Jackson, 1981). The
BBS Academy of Management Review October

importance of maintaining three distinct components is also illustrated by

the differential patterns of correlations between each component and
other study variables, such as age, workload, autonomy, job challenge,
satisfaction with status and recognition, role conflict, and role ambiguity
(Friesen et al., 1988; Jackson et al., 1986; Maslach & Jackson, 1984; Schwab
& Iwanicki, 1982a). That is, some variables are differentially correlated,
or even uncorrelated, with each of the three components, although find-
ings have not been highly consistent. Further, if other variables are dif-
ferentially associated with burnout components, then it is plausible that
intervention strategies would also be differentially effective, depending
on the particular burnout component that is being addressed. Thus, an
overall measure of burnout results in a loss of information.

ANTECEDENTS OF BURNOUT
The many demands, constraints, and other factors that compose the
antecedents of burnout can be grouped into three broad categories. The
first category consists of job and roJe characteristics. It appears from the
literature that the most critical variables are the characteristics of the
employee-client relationship. Role characteristics consist of role over-
load, ambiguity, and conflict. The second category, organizationai char-
acteristics, includes variables such as job context and contingency of
rewards and punishments. The final category addresses personal char-
acteristics.
Job and Role Characteristics
The role of the client in service-provision interactions, and the expec-
tations of the service providers themselves, have been shown to contrib-
ute to an explanation of experienced burnout. Client interactions that are
more direct, frequent, or of longer duration, for example, or client prob-
lems that are chronic (versus acute) are associated with higher levels of
burnout. Role conflict, role ambiguity, and role overload have been
shown to be associated with burnout to varying degrees. Individuals who
report higher levels of these role variables also report higher levels of
burnout.
Role of interpersonal relations. Most of the systematic research on the
concept of burnout has focused on individuals in the helping professions,
specifically health, social services, and teaching, where burnout is typi-
cally believed to be most frequently and intensely experienced because of
the high level of arousal from direct, frequent, and rather intense inter-
actions with clients. As a result, the term client has taken on a more
narrow, limiting connotation. In this article, client refers to any individ-
ual, internal or external, with whom one interacts on a professional basis.
Maslach (1978) theorized that the potential for emotional strain is
greatest for workers in the helping professions because they are con-
stantly dealing with other people and their problems. Their work involves
1993 Cordes and Dougherty 629

extensive and direct face-to-face contact with other people in emotionally

charged situations. In many of these professions, feedback from either the
client or the organization sources is either nonexistent, or is almost ex-
clusively negative.
The potential stressfulness of these interactions is likely to be af-
fected, in part, by the stance of the clients during these interactions.
Depending on the situation, clients may be aggressive, passive-
dependent, or defensive (Maslach & Jackson, 1984). Further, these clients
are relying on the service providers to help them. The belief of the service
provider that he or she alone is responsible for ensuring the future well-
being of this client can be an awesome and exhausting burden (Maslach,
1982). The problem or strain is often compounded by an excessively heavy
client load.
As a result, research has focused on the role of the client and the
employee's caseload in contributing to burnout. Jackson and colleagues
(1986) suggested that caseload can be divided into quantitative and qual-
itative dimensions. The quantitative dimensions include frequency of
contact, duration of contact, number of interactions, and percent of time
spent with clients. The quantitative aspects of client caseload most com-
monly investigated are number of interactions or clients and amount of
time spent with clients. Most of these investigations have reported on the
relationship of classroom size to burnout and the effect of amount of break
time, or time outs, on burnout (Maslach & Pines, 1977; Pines & Maslach,
1978). As the number of clients increases, the demands on the employee's
personal resources increase. If these demands are continuous, rather
than intermittent, the employee may be vulnerable to burnout.
Qualitative dimensions of client caseload involve interpersonal dis-
tance (e,g., phone contact versus face-to-face contact) and client charac-
teristics (e.g., chronic versus acute, child versus teenager). Both of these
qualitative dimensions represent a kind of psychological or interpersonal
intensity of contacts with clients. The qualitative dimension most com-
monly studied involves the effects of the type of client problem at issue.
Grade level taught (Anderson & Iwanicki, 1984; Gold, 1985; Schwab &
Iwanicki, 1982b) and severity of a student's mental handicap (Zabel &
Zabel, 1982), for example, affect experienced levels of the burnout com-
ponents for teachers. Theoretically, the nature of the client's situation can
be described as more or less stressful, it can be chronic or acute, and it
can have a higher or lower probability of success or cure. In many cases,
there may be a lack of obvious change or improvement in the situations of
a large proportion of the clients as a result of the services rendered.
Several clients may disappear after service provision, only to return later
with essentially the same needs. These factors hypothetically affect both
the intensity and the duration, which, in turn, affect the overall stress
associated with the interaction (Maslach, 1978).
Studies by VanYperen, Buunk, and Schaufeli (1992) and Leiter and
Maslach (1988) examined additional aspects of interactions. VanYperen
630 Academy of Management Review October

and colleagues (1992) found that nurses' perceived imbalance in their

relationships with patients was associated with higher levels of burnout.
That is, nurses who believed they invested more in their patients than
they received in return, in the form of positive feedback, health improve-
ments, appreciation, and gratitude, also reported higher levels of emo-
tional exhaustion, depersonalization, and diminished personal accom-
plishment.
In a study of workers in a mental health organization, Leiter and
Maslach (1988) investigated the effects of different sources and types of
interpersonal contact. They looked at contact with co-workers and super-
visors, and distinguished between pleasant and unpleasant co-worker
and supervisor contacts. Among the group of nurses, unpleasant super-
visor contact was positively related to emotional exhaustion, whereas
pleasant supervisor contact was negatively related to depersonalization
and pleasant co-worker contact was positively related to personal accom-
plishment.
Although these variables have been viewed mainly but not exclu-
sively in the context of client or service recipient contacts (e.g., nurse-
patient or teacher-student interactions), they may provide insight into the
generalizability of burnout. Maslach and Jackson (1984) and Shirom (1989)
noted that there are many occupations not included under the rubric of the
helping professions, but where interpersonal contacts cause strain, in
which employees may therefore be vulnerable to burnout. Pines and
Aronson (1981), for example, found that dentists experience burnout. Jack-
son and Schuler (1983) and Jackson (1984) have speculated that managers
and supervisors also may experience burnout because they are required
to help their employees resolve job-related and personal difficulties.
Role conflict and role ambiguity. Role conflict occurs as a result of
incongruity or incompatibility of expectations communicated to a role
incumbent by his or her role senders (Kahn, 1978; Kahn et al., 1964). Role
ambiguity is associated with one's need for certainty and predictability,
especially regarding one's goals and means of accomplishing them. It
may occur if an individual lacks adequate information to accomplish
required activities, as, for example, when information is restricted or not
clearly defined or articulated (Jackson & Schuler, 1985). Lack of clarity
regarding proper procedures for performing job tasks or criteria for per-
formance evaluations (Miles & Perreault, 1976) results in role ambiguity.
Role conflict and ambiguity have received attention in the burnout
literature. Schwab and Iwanicki (1982a) found that these two variables
accounted for a significant amount of variance in the emotional exhaus-
tion and depersonalization dimensions for a sample of 469 teachers,
whereas role ambiguity accounted for a significant, though much
smaller, amount of variance in personal accomplishment. Brookings and
colleagues (1985) reported statistically significant relationships between
perceived role conflict and role ambiguity and all three burnout compo-
nents for 135 female human service professionals. Jackson and her col-
1993 Cordes and Dougherty 631

leagues (1986), however, found role conflict to be significantly associated

with emotional exhaustion, but not with personal accomplishment or de-
personalization in their study of 248 teachers; role ambiguity was found to
be significantly related to personal accomplishment. Leiter and Maslach
(1988) found role conflict was significantly related to emotional exhaus-
tion for a sample of nurses, whereas Jackson and her colleagues (1987)
found it was related to emotional exhaustion as well as depersonalization
for 391 public service lawyers. Fimian and Blanton (1987) found both role
variables were related to total burnout for a sample of teacher trainees
and first-year teachers.
Although only a few studies have investigated the effects of role
ambiguity and role conflict on burnout, the findings across them are very
consistent. Because role conflict and role ambiguity are not limited to
human service professions, the relationship between these role variables
and burnout would be expected to be equivalent in corporate and indus-
trial settings as well.
Role overload. As originally conceptualized, burnout was believed to
result partially from qualitative and quantitative overload (Maslach &
Jackson, 1984). Individuals experiencing qualitative overload feel they
lack the basic skills or talents necessary to complete the task effectively.
Quantitative overload refers to the individual's perception that the work
cannot be done in the allotted time (Kahn, 1978; Pines & Maslach, 1978). In
many organizations this may come about because of resource scarcity
and the continual threat of cutbacks (Jackson, 1984). As a result, staff
workers may often be overloaded with cases, clients, or students
(Maslach, 1976).
Empirical investigation to date has focused on the effects of quanti-
tative overload on individuals' burnout scores, with very consistent find-
ings. Higher staff-child ratios in day-care centers (Maslach & Pines, 1977)
and school classrooms (Russell, Altmaier, & Van Velzen, 1987), for exam-
ple, are associated with higher experienced levels of the burnout compo-
nents. Given the theoretical basis for observing an effect of qualitative
overload on burnout levels, however, future studies should focus on this
relationship, especially as it applies in corporate and industrial settings.
Organizational Characteristics
Even though many researchers have investigated how variables as-
sociated directly with the job or task contribute to burnout, the question of
how variables associated with the organization itself and its policies may
be related to burnout has received comparatively little attention. In this
section, we consider how this category of variables, specifically contin-
gency of organizational outcomes and job context, might affect burnout.
Contingency and noncontingency of organizational outcomes. The
role of contingency of organizational outcomes (i.e., the extent to which
rewards and punishment are linked to performance) in contributing to
burnout has both intuitive appeal and theoretical justification. The two
6^ Academy of Management Review October

problematic conditions are lack of contingent rewards and the presence of

noncontingent punishment. This phenomenon, however, has received lit-
tle attention in the literature. Jackson and colleagues (1986) studied con-
tingency of organizational outcomes in two ways. First, they examined
contingency of outcomes as a job condition that might contribute to burn-
out. Second, they examined the role of this variable in the experience of
unmet organizational expectations. In neither case, however, did they
find significant relationships. Although they suggested that the lack of
significant relationships may have been due to methodological problems,
the theoretical arguments appear to justify further investigation of this
variable as a potential contributor to burnout.
Job context. Empirical evidence indicates that specific context affects
the incidence of stress and burnout in the workplace (Gaines & Jermier,
1983; Parasuraman & Alutto, 1981; Pretty, McCarthy, & Catano, 1992). The
context is characterized by a variety of factors, such as subsystem, work
shift, and psychological environment, which have been shown to contrib-
ute to burnout. As discussed previously, a critical factor contributing to
burnout may be the nature of the employee-client relationship. If job con-
texts differ significantly by the types of interaction that characterize them
(e.g., frequency, intensity), context would, in tum, be differentially re-
lated to burnout. For example, managers who perform boundary-
spanning functions have a high frequency (and perhaps intensity) of in-
terpersonal contact, whereas production managers' environments are
characterized by much less frequent or intense interpersonal contact.
Similarly, service representatives or sales representatives, who act
mainly as boundary spanners, have more interpersonal contact than will
information system specialists.
Pretty and colleagues (1992) studied the effects of psychological en-
vironment, job level, and gender on burnout among managerial and non-
managerial telecommunications employees. They found that women ex-
perienced more emotional exhaustion and depersonalization if they were
nonmanagers, whereas men experienced more emotional exhaustion and
depersonalization if they were managers.
Gaines and Jermier (1983), in a study of police officers, found that
levels of emotional exhaustion differed across departments. They sug-
gested that this difference was due to the higher status and enriched
nature of the work in some departments versus others. It is more likely,
however, that specialty area is acting as a surrogate for other variables,
most notably interpersonal or client interactions. If the nature of interac-
tions had been assessed, and then their effects partialed out, it seems
doubtful that departmental context would have contributed significantly
to the variance in burnout.
Personal Characteristics
Finally, certain demographic characteristics have been shown to con-
tribute to an explanation of why some individuals experience burnout
and why others remain virtually unaffected by it. In addition, social sup-
1993 Vy Coides and Dougherty 633

port, defined in a variety of professional and personal ways, is generally

related to lower levels of the burnout components. Employees whose or-
ganization and achievement expectations are more discrepant from the
current realities of the workplace report higher levels of burnout. Finally,
the relationship between burnout and one's career progress has not yet
been investigated.
Demographic variables. Men and women often report differences in
levels of the three burnout components, but there is mixed evidence con-
cerning the pattern and complexity of relationships (Lemkau, Rafferty,
Purdy, & Rudisill, 1987; Maslach & Jackson, 1981, 1985; Pretty et al., 1992;
Schwab & Iwanicki, 1982b). Younger individuals consistently report
higher levels of the burnout components (Anderson & Iwanicki, 1984;
Maslach & Jackson, 1981; Russell et al., 1987; Schwab & Iwanicki, 1982b;
Stevens & O'Neill, 1983; Zabel & Zabel, 1982), but one study found that
more experienced employees reported lower levels of emotional exhaus-
tion and depersonalization (Anderson & Iwanicki, 1984). There is some
evidence that married individuals report lower levels of the burnout com-
ponents (Maslach & Jackson, 1985), and individuals with children consis-
tently report lower levels of the burnout components (Maslach & Jackson,
1985). These results are summarized in Table 1. The moderating effects of
these variables have either not been studied or their results were not
significant and thus not reported. An exception is Gaines and Jermier's
(1983) study of emotional exhaustion in a police organization. They found
a significant interactive effect between departmental context and gender,
but not between context and either years in position or marital status.
Social support. The effects of social support on stress and burnout
have received extensive attention in the literature (Caplan, 1974; Cohen &
Wills, 1985; Constable & Russell, 1986; Maslach & Jackson, 1984). Social
support appears to have a positive effect on individuals' well-being
through two different processes (Cherniss, 1980b). First, support has been
identified as a buffer or moderator between job-related stress and the
pathogenic influences of stressful events. This may occur in one of two
ways. Support can help individuals redefine the potential harm in the
situation, or it can enhance their belief that they can cope with the situ-
ation by increasing their perception that others will provide the necessary
resources (Cohen & Wills, 1985). In addition, the buffering model suggests
that social support buffers the relationship of stressors with use of adap-
tive coping (Kirmeyer & Dougherty, 1988).
Second, social support can have a main or direct effect on experi-
enced stress. Social support is positively related to psychological and
physical health, irrespective of the presence or absence of life or work
stressors (Dignam, Barrera, & West, 1986). The accumulating evidence
has provided support for both the buffering model and the direct model of
social support. Both models seem to be correct in certain situations, but
each represents a different process through which social support affects
well-being (Cohen & Wills, 1985).
Numerous studies relating support to burnout have considered the
 TABLE 1
Summary of Relationship Between Demographic Variables and the
Burnout Components''
Marital/Tamily
Study/Sample Gsnder Age Status Experience

Anderson & Young More experienced

Iwanicki (1984) experience > workers' experience
375 teachers EE. > EE & DP.
Gaines & leimier F experience > No relation No relation between
(1983) 169 EE. between MS experience and
police officers and any any burnout
and support burnout components.
personnel components.
Gold (1985) 462 F experience < Young Single No relation between
teachers DP and > PA. experience > experience > experience al
EE and DP. but EE and DP. same school and
also > PA. any burnout
components.
Lemkau et al. No relation
(1987) 67 family between
practice gender and
physicians any burnout
components.
Maslach & M experience > Young
lackson (1981) DP and > PA. experience >
1.025 EE and DP, and
health/service <PA.
professionals
Maslach & MBI Study: F MBI Study:
Jackson (1985) experience > Married
1.025 EE and < PA; experience <
health/service M experience EE.
professionals >DP. Agency Study:
845 Agency Study: M Manied
claims/service experience > experience <
& telephone DP; no relation EE and DP, and
reps. between > PA; childless
gender and EE experience >
or PA. EE and DP, and
<PA.
Pretty et al. (1992) M experience >
433 telecomm. EE, DP if
workers manaaers; F
experience >
EE, DP if
nonmanagers.
Russell et al. M experience > Young Married No relation between
(1987) 316 DP; no relation experience > experience > experience and
teachers between EE; no relation PA. any burnout
gender and EE between age components.
or PA. and DP or PA.
Schwab & M experience > Young No relation No relation between
Iwanicki DP; no relation experience > between MS experience and
(1982b) 469 between EE and DP. and any any burnout
teachers gender and EE bumout components.
or PA. components.
Zabel&Zabel Young No relation between
(1982) 100 ex;>erience experience and
special EE. any bumout
education conip)Onents.
teachers
° EE ~ emotional exhaustion, DP depeisonalization, PA = personal accomplishment, M = male.
F = female.
1993 Cordes and Dougherty 635

buffering effect. Greater perceived social support from co-workers or su-

pervisors in day-care centers (Maslach & Pines, 1977), elementary and
secondary schools (Jackson et a l , 1986; Russell et al., 1987; Zabel & Zabel,
1982), hospitals (Constable & Russell, 1986), and U.S. legal agencies (Jack-
son et al., 1987) is associated with lower reported levels of the burnout
components.
In a series of studies, Leiter (1988, 1990, 1991) explored the direct effect
of several facets of social support. In a study of workers in a mental health
organization, he recognized the role of personal, informal, and profes-
sional support (Leiter, 1988). He found that informal contacts or support
was positively related to personal accomplishment. Professional support,
however, played a dual role of both alleviating and aggravating burnout.
Although professional support was related to higher feelings of personal
accomplishment, it was also positively related to emotional exhaustion.
In a related study, Leiter (1990) defined personal support as family re-
sources and organization support as skill utilization, that is, the opportu-
nity for implementation and development of skills. He found that personal
support was negatively related to emotional exhaustion and depersonal-
ization, whereas organization support was negatively related to deper-
sonalization and diminished personal accomplishment. In a third study,
Leiter (1991) examined three types of organization supports: skill utiliza-
tion, co-worker support, and supervisor support. As before, skill utiliza-
tion was positively related to personal accomplishment, but it was neg-
atively related to emotional exhaustion. Co-worker support was nega-
tively related to depersonalization and positively related to personal
accomplishment. Supervisor support was not significantly related to any
of the components.
This line of research highlights the distinct pattern of relationships
among variables and the three burnout components, and it also points out
the complicated nature of support and the role it plays in burnout. Not
only are the sources of support differentially effective on the burnout
components, they may also have negative effects. This research suggests
that professional and personal sources of support are largely indepen-
dent of one another.
It is important for future researchers to attempt to identify the specific
aspects of support that either contribute to or result in a reduction of
burnout. Furthermore, although conceptual research on burnout identi-
fies both professional and personal sources of social support, most of the
research to date has investigated the professional or organizational re-
sources. A better understanding of the effects of personal sources of social
support can contribute to an understanding of the relative importance of
the variables associated with burnout. This understanding becomes in-
creasingly important as the number of households with two employed
adults continues to increase. A better understanding of the intricacies of
the work-nonwork interface, and the implications that this interface has
for personal sources of support, is an important and timely concern.
636 Academy of Management Review October

Personal expectations. Employees' expectations about the profes-

sion, the organization, and their own personal efficacy also make a sig-
nificant contribution to burnout (Cherniss, 1980; Jackson & Schuler, 1983;
Maslach & Jackson, 1984) and represent a source of "demands" placed
upon themselves in their work. Jackson and colleagues (1986) referred to
these as achievement expectations and organizational expectations.
Achievement expectations are the individuals' beliefs about what they
will be able to accomplish with clients. Organizationai expectations refer
to the individuals' expectations about the nature of the professional sys-
tem in general and the job in particular. Many of these expectations are
inculcated by the individual's most recent training environment (Gold,
1985) or are shaped by recruiters interested in selling the organization to
applicants (Wanous, 1973).
Besides high expectations, unmet expectations can also be a source
of burnout. As individuals enter the profession or change their organiza-
tional affiliation, they compare their expectations with their experiences.
The resulting discrepancy will influence employees' reactions to their
jobs (Porter & Steers, 1973). The greater the discrepancy, the greater the
effects are likely to be for both the new employee and the organization
(Wanous, 1973, 1976).
Jackson and her colleagues (1986) studied the effects of unmet expec-
tations on reported burnout of 248 teachers. They hypothesized that emo-
tional exhaustion, feelings of depersonalization, and feelings of low per-
sonal accomplishment would be associated with higher levels of unmet
expectations, but they found no support for their hypothesis. They sug-
gested that this lack of association was because the unmet expectations
of their respondents were assessed without regard to tenure on the job.
Such a measure is contaminated by the effects of memory deterioration.
In addition, the measure they used required respondents to calculate and
report, in a single step, the discrepancy between their early expectations
and the current situation for a variety of job factors. A more precise and
direct measure of unmet expectations would result if respondents report
on their early expectations and the current situation separately, leaving
the discrepancy calculation for the researchers.
Another expectation variable that may contribute to burnout concerns
shifts in expectations. Research has shown that older, more experienced
employees tend to experience lower levels of burnout than do younger
employees. One explanation for this discrepancy may be that older em-
ployees have actually shifted their expectation set to fit reality based on
their experiences. For example, individuals may lower their expectations
of the levels of client gratitude and organizational resources. Stevens and
O'Neill (1983) examined the roles that expectations and expectation shifts
play in burnout and found that total expectation score was correlated with
perceptions of diminished personal accomplishment. Expectation shifts
were correlated with all three burnout scales; higher levels of burnout
were associated with negative expectation shifts (i.e., organizational re-
1993 - Cordes and Dougherty 637

alities are worse than originally expected). Less experienced employees

maintained client-focused expectations and thus reported higher levels of
burnout, whereas more experienced employees shifted their expectations
from client progress, for example, to their own competencies. These find-
ings, however, stemmed from a cross-sectional methodology.
Career progress. Individuals who have had greater upward career
movement may experience less burnout. If one keeps the symptoms of the
three components of burnout in mind, there may be three possible expla-
nations for this hypothesized relationship. First, repeated promotion is
generally accompanied by a reduction in client contact (Maslach, 1982).
This change reduces an individual's susceptibility to emotional exhaus-
tion resulting from the demands of client interaction. Second, greater
career advancement relative to peers may signify to individuals that they
are making a positive contribution. In fact, in some cases, this may be a
rare piece of positive feedback. This measure of accomplishment can help
counter any feelings of diminished personal accomplishment stemming
from other sources. Finally, individuals who have had reasonable career
progress are more likely to believe that the organization has appropriate
policies and procedures and that the policies and procedures regarding
promotions are fair and equitable. An environment that is perceived as
being predictable and fair is less likely to induce learned helplessness;
therefore, the depersonalization that often ensues from such learned help-
lessness is likely to be minimized. Despite its theoretical merit, this vari-
able has not received attention in the literature on burnout.

INDIVIDUAL AND ORGANIZATIONAL CONSEQUENCES OF BURNOUT

The importance of burnout as a practical concern is illustrated by its

association with negative organizational outcomes and various types of
personal dysfunction. Although many of the consequences discussed here
are not unique to burnout, they illustrate how potentially costly and dam-
aging burnout can be and highlight the importance of better management
to deal with the problem. In a recent review. Kahili (1988) grouped the
consequences of burnout into five categories: physical, emotional, inter-
personal, attitudinal, and behavioral. Although empirical evidence sup-
porting these relationships is scant (especially evidence that meets the
criteria set forth in the beginning of this article), certain general conclu-
sions can be drawn.

Physical and Emotional Consequences

The burnout components have been linked with a variety of mental

and physical health problems (Burke & Deszca, 1986; Maslach & Pines,
1977). Deterioration of mental health is characterized by decreases in
feelings of self-esteem, depression, irritability, helplessness, or anxiety
(Jackson & Maslach, 1982; Kahili, 1988). Physical health problems include.
636 Academy of Management Review October

for example, fatigue, insomnia, headaches, and gastrointestinal distur-

bances (Kahili, 1988),
In a study of supervisors and managers from a public welfare agency.
Lee and Ashforth (1990) found psychological and physiological strain and
helplessness to be associated with higher levels of emotional exhaustion
and depersonalization. Burke and Deszca (1986) assessed how often re-
spondents experienced 18 physical conditions, including poor appetite,
headaches, and chest pains, and found that this measure of psychoso-
matic symptoms was positively related to the bumout components. Sim-
ilarly, on the basis of self-reports and reports of spouses, Jackson and
Maslach (1982) reported that police workers who experienced significant
levels of the burnout components returned home from work tense and
anxious.
Interpersonal Consequences
The deleterious effects of job-related activities on individuals' rela-
tionships with family and friends and on their personal lives have gained
increasing recognition as researchers have become aware of the link
between work and nonwork domains (Jackson & Maslach, 1982). In par-
ticular, the links between burnout and the deterioration of social and
family relationships and the links between burnout and work-nonwork
conflict have received empirical support (Burke & Deszca, 1986; Jackson &
Maslach, 1982).
In a study of 142 couples (where one spouse was a police officer),
Jackson and Maslach (1982) found that individuals experiencing burnout
tended to withdraw from their friends and reduced their socializing. In
addition, the police officers reported being unable to shake off the pro-
fessional role when they left the job. The spouses reported that the offic-
ers, when dealing with the children, handled them as they would some-
one in a professional situation.
Burke and Deszca (1986) and Burke and colleagues (1984a) studied the
relationship between burnout and work-nonwork conflict also among po-
lice workers. They assessed the impact of the officers' job and job de-
mands on nine areas of personal, home, and family life, and they com-
bined these areas into a single measure of work-nonwork conflict. They
found that those individuals who reported higher levels of the burnout
components also reported a greater negative impact of the job demands
on their personal, home, and family lives.
Interpersonal consequences also include changes in the nature or
frequency of interactions with clients and co-workers (Jackson & Schuler,
1983). In studies conducted at day-care centers, child care workers who
experienced higher levels of the burnout components experienced greater
impatience and moodiness and less tolerance. They also reported with-
drawing more from clients, either by talking with other staff more or by
taking longer breaks and lunch periods (Maslach & Pines, 1977). In their
study of public contact workers, Maslach and Jackson (1985) also found
1993 Cordes and Dougherty 639

support for the link between the bumout components and the desire to
spend less time with the public as well as a link between the burnout
components and poorer co-worker relations.
Attitudinal Consequences
Attitudinal consequences involve the development of negative atti-
tudes toward clients, the job, the organization, or oneself (Kahili, 1988). In
studies of police workers (Burke et al., 1984a; Jackson & Maslach, 1982)
and public contact workers (Maslach & Jackson, 1985), for example, em-
ployees reported higher levels of dissatisfaction. In Jackson and Mas-
lach's (1982) study, even the spouses of the workers developed negative
attitudes toward the police officer's job. The burnout components also
have been linked to lower levels of organizational commitment for public
service lawyers (Jackson et al., 1987) and nurses (Leiter & Maslach, 1988).
Behavioral Consequences
Behavioral consequences of burnout entail work- or organization-
related behaviors as well as consumption behaviors. Organizational out-
comes include turnover (Jackson et al., 1986), absenteeism (Firth & Britton,
1989), and decreases in the quality and quantity of job performance
(Maslach & Jackson, 1985). Consumption behaviors include behaviors
such as smoking and drug and alcohol use.
In studies of police workers (Burke & Deszca, 1986; Burke, Shearer, &
Deszca, 1984a; Jackson & Maslach, 1982), individuals reporting higher
levels of the burnout components were more likely to report intentions to
leave their jobs. They also reported higher levels of drug, alcohol, and
tobacco use. Jackson and her colleagues (1986) found that teachers'
thoughts about leaving their jobs were significantly associated with their
reported levels of all three burnout components, but actual turnover lev-
els were significantly associated with emotional exhaustion.
In another study. Firth and Britton (1989) investigated the relationship
between burnout and turnover and between burnout and absenteeism
among nurses. They found that actual turnover of nurses was associated
only moderately with depersonalization, and there was no significant
relationship between turnover and either emotional exhaustion or feel-
ings of reduced personal accomplishment. Absenteeism was reported to
be higher for those nurses experiencing higher levels of emotional ex-
haustion (Firth & Britton, 1989), although this relationship was significant
only for longer periods of absence. Finally, among public contact employ-
ees in a federal service agency, Maslach and Jackson (1985) found that the
burnout components are linked not only to intent to leave a job but also to
poorer job preparation.
Thus, as the scant but growing body of evidence illustrates, the con-
sequences of burnout have some very real physical, emotional, interper-
sonal, attitudinal, and behavioral implications. Not only does the indi-
vidual suffer, but the employee's family and friends, the organization.
6W Academy oi Management fleview October

and the people with whom the employee interacts during the work day all
bear the costs of this organizational problem.

RESEARCH PROPOSITIONS: TOWARD AN UNDERSTANDING

OF BURNOUT

Stress theorists such as McGrath (1976) and Schuler (1980) have de-
scribed stress as resulting from demands, constraints, or opportunities
presented to employees. Employees experience the strongest responses
("strains") to these demands when they perceive uncertainty about the
ability to handle the demands and when the consequences of handling
the demands are important. Bumout is a response to demand stressors
(e.g., workload) placed upon an employee, and it is distinguishable from
other forms of stress because it represents a set of responses to a high
level of chronic work demands, entailing very important interpersonal
obligations and responsibilities. Because of the high level of arousal,
employees begin to feel emotionally exhausted after repeated exposure to
these important demands, using depersonalization of clients as a coping
strategy. Finally, they begin to feel a diminished sense of personal ac-
complishment, especially when their work environment provides little
feedback and few rewards for work accomplishments. These relation-
ships are summarized in Figure 1.
Many researchers consider emotional exhaustion to be the key to the
experience of burnout, and the first stage of the burnout process (Gaines
& Jermier, 1983; Maslach, 1982). Burnout (and especially emotional ex-
haustion) has been consistently shown to be directly related to high levels
of work demand (Shirom, 1989). We believe that the key determinants of
emotional exhaustion reflect both organizational and personal demands
placed upon employees. For example, quantitative work overload (the
perception of too much work to accomplish in the time available) is an
important determinant of emotional exhaustion. Individuals' attempts to
maintain performance standards despite insufficient time and staff may
lead to an excessive expenditure of time and emotional energy, thus
leading to emotional exhaustion. Despite a lack of comparability across
the literature on burnout, workload has been consistently linked to emo-
tional exhaustion in the vast majority of studies. Particular aspects of
workload that were linked to emotional exhaustion have been found in
the works of Friesen and Sarros (1989), Jackson and her colleagues (1986,
1987), Maslach and Pines (1977), and Pines and Maslach (1978).
Role conflict is also a job demand that can contribute to emotional
exhaustion in the burnout process. Employees may experience a variety
of different kinds of conflicting role expectations from different sources or
roJe senders (Kahn et al., 1964). Intersender conflict, for example, occurs
when demands from one's supervisor conflict with the demands from cli-
ents. Person-role conflict occurs when expectations from one's job conflict
with one's values or personal beliefs. Attempts to reconcile conflicting
1993 Cordes and Dougherty 641

m
ion:

"3 1
"D
s .1
CD
D
a O
uantitallive role.

terpersianal inte
ob/Role,^Organiz

M
0)
M
Si
'y
Directn ess

o
a B
Freque:

a
Intensil
Duratic

S 0
9 u
o Q
O EC B
lievemeni: expecta-

:pec-

n
and Stres;sors

0)
•3 Ul
ife intere:
[ndividua

anization

d (D
vement

3
izat nal

Ul D tn
<D H 01 01 B
U 0
.9 u u 0
SOU!

.e 3 -g
S
d)
u D)
0 [fl 0
Ul
C
0
0
>
„0
3
0
in • -
1-3
c B
D
c
0
Ul
D
B
Q <0 G 0 0 01 B 0 01 B 0
^
Dl .0
C
Dl a .s 0 in X! D u D Ul
1 support:—org
DZI]
loping

01 Ol O' !: D) D) 11
1 support per
ing

f
X " 0
B U < u .
B c
.2 a
i; 0
'a
0
DOS

•J u u 0 0
"0 aa "o "o a a
of.essi onal

a
3 a a a
ility

3 3 3
tn Ul tn tn

1
cial
cial

IS D 0
D .2 .2 £
y 0 y -o
Avail

0
Avai

0 0 a
UJ CO '3 CO
^ %
W
< Dl IA
B B _
0 C
BO 0) B *- 0
-*• D (D D
0) A Dl Tt
B u u (D "5
u
Factoi sug esting 0

ce-contii
unappj iciat d< ineffe

skill ut izat: n or job

a (D
equate
litative ole< ferload

a
0)
"D
Dt <D
"5
r-l a G 0
B 01
et orga izat:
DUIJ

uiaA

•
ity

.2
n w 0 0 3 B s
ol perf

D)
01 !3
D
S
-s D
D
il D 0> g
3
S M "o
05
B B
1
642 Academy of Management Review October

demands may be frustrating and emotionally taxing {Fisher & Gitelson,

1983). Numerous studies have found links between role conflict and emo-
tional exhaustion (e.g., Jackson et al., 1986; Jackson et al., 1987).
In addition to organization and role demands, individuals may place
demands on themselves that lead to emotional exhaustion. Maslach
(1982), for example, has written about expectations for achievement as a
key factor, suggesting that those who begin to burn out (i.e., experience
emotional exhaustion) tend to be overachievers who have unrealistic ex-
pectations and are typically younger employees. These characteristics
have been confirmed by links between age and emotional exhaustion in
the literature (Anderson & Iwanicki, 1984; Gold, 1985; Russell et a l , 1987;
Zabel & Zabel, 1982). Thus, young, idealistic employees who have high
expectations about their own accomplishments are subject to exhaustion
in response to these high demands (Gold, 1985; Maslach, 1982; Maslach &
Jackson, 1981; Maslach & Jackson, 1984; Saxton, Phillips, & Blakeney,
1991).
Individuals also may generate high demands in terms of expectations
for the organization in which they work. New employees often expect
much from themselves, and they also expect a lot from the organization;
frequently these organizational expectations are especially high when
the employee enters the organization (Wanous, 1973). High expectations
in terms of work challenge, rewards and recognition, career advance-
ment, and many other aspects of work can create intrinsic demand stress.
New employees may also believe that they can change conditions in the
organization, expectations that are often unrealistic (Maslach, 1982). Sim-
ilarly, employees with high levels of job involvement (Lodahl & Kejner,
1965) and who hold work as a central life interest (Dubin, 1956) are more
likely candidates for emotional exhaustion because they view their work
and its consequences as being extremely important.
Finally, as already noted, researchers suggest several qualitative
and quantitative dimensions pertinent to interpersonai interactions, each
representing different types of demands, which are critical variables af-
fecting the levels of emotional exhaustion (Jackson et al., 1986; Maslach,
1982). Frequent face-to-face interactions that are intense or emotionally
charged will likely be more demanding and can be expected to be asso-
ciated with higher levels of emotional exhaustion (e.g., this might occur
with line managers) (Maslach, 1982). Individuals who have little direct
contact with clients, who deal primarily with "paper people" (e.g., per-
sonnel staff), or who deal with more "neutral" client issues should expe-
rience less emotional exhaustion (Maslach, 1978). Between these ex-
tremes, individuals who handle a large volume of clients on a daily basis
but address "neutral" problems may experience a moderate amount of
emotional exhaustion. Some individuals' client contact may even be quite
ego enhancing, rewarding, and pleasant (Leiter & Maslach, 1988). The
nature of the demands of client contacts is likely to be a key factor in the
bumout of managers and professionals.
1993 Cordes and Dougherty 643

Figure 2 is a matrix designed to stimulate ideas about the general-

izability of emotional exhaustion from client relationships to a wide va-
riety of jobs. The four cells are based on two key aspects of client rela-
tionships: (a) the frequency of interpersonal interactions with clients and
(b) the intensity of interactions with clients. {Although four aspects of
interpersonal interaction have been discussed—directness, frequency,
intensity, and duration—only two are used here for ease of illustration.)
Jobs that we speculate are low or high in frequency and intensity are
provided as examples in the four cells. Individuals employed in jobs that
tend to have high levels of both aspects (e.g., nurse, social worker) would
be expected to experience the highest levels of emotional exhaustion from
interpersonal relations. Not surprisingly, jobs with both high intensity
and high frequency have dominated the burnout literature to date. Con-
versely, individuals employed in jobs that tend to have low levels of both
aspects {e.g., research physicist) would be expected to experience the
lowest levels of emotional exhaustion. Individuals with mixed levels of
both aspects, that is, jobs with high frequency and low intensity (e.g.,
receptionist) and those with low frequency and high intensity (e.g., para-
medic), would be expected to experience a moderate level of emotional
exhaustion. This conceptualization suggests that many jobs outside the
traditional helping professions are likely to lead to moderate and, in
some cases, even high levels of emotional exhaustion (e.g., customer
service representatives).
Of course, service provider-recipient relationships are not the only
ones characterized by demands associated with interpersonal strain. Su-
pervisor-subordinafe and co-worker con(ac/s include interpersonal inter-
actions that also may contribute to emotional exhaustion (Leiter &

HGURE 2
Predicted Emotional Exhaustion in Jobs Based Upon Nature of
Interpersonal Contact
Social worker
Receptionist
Customer service
High Sales representative
representative
Librarian
Schoolteacher
Benefits representative
Nurse
Frequency of (Moderate burnout)
(High burnout)
Interpersonal
Contact
Research physicist Paramedic
Forest ranger Public defender
Oil refinery operator Fire fighter
Low Laboratory technician Police detective
(Low burnout) (Moderate burnout)

Low High

Intensity of Interpersonal Contact

644 Academy of Management Review October

Maslach, 1988). Boundary-spanning positions also can include significant

interpersonal interactions (Parkington & Schneider, 1979). The boundary
spanner, who functions as an information processor or a filter between
the organization and the client, represents the organization and acts as
its agent in influencing the decision making of the client. Boundary span-
ners are caught in a difficult position when they perceive that client de-
mands cannot or will not be met by the organization.
In summary, emotional exhaustion is the first stage of burnout. It is
primarily a response to demand stressors placed upon employees, espe-
cially work overload, interpersonal interactions, role conflict, and high
levels of both personal and organizational expectations.
Proposition la: High levels of work demands are the pri-
mary determinants of emotional exhaustion. These de-
mands include work overload, role conflict, and direct,
intense, frequent, ox lengthy interpersonal contacts.
Corollary la: Individuals in supervisory positions and in
boundary-spanning positions will experience higher
levels of emotional exhaustion compared to those in
nonsupervisory or in non-boundary-spanning positions.
Proposition lb: High levels of personal demands are also
critical determinants of emotional exhaustion. Younger
employees and those with high expectations for job per-
formance, high expectations for the organization, high
job involvement, and work as a central life interest are
more likely to experience emotional exhaustion.
Depersonalization (cynical, dehumanizing, and negative attitudes to-
ward one's clients) is a stress response that is unique to burnout, and it
has not been examined in the job stress literature (Jackson et al., 1986). It
is most appropriately conceptualized as a (defensive) coping response to
emotional exhaustion—an expedient alternative for dealing with emo-
tional exhaustion when other coping resources are not available (Ash-
forth & Lee, 1990; Lee & Ashforth, 1990; Leiter, 1990; Maslach, 1982).
Employees' professional socialization may be at least partly respon-
sible for the use of depersonalization in response to high work demands
and the resulting emotional exhaustion (Leiter, 1990; Lemkau et al., 1987;
Maslach, 1982). As a result of both school and on-the-job training, with-
drawing from the job demands by depersonalizing may be viewed not
only as an acceptable response, but a professional one as well. Leiter
(1990), for example, pointed out that human service professionals gener-
ally follow a code of ethics that explicitly calls for depersonalization of
clients. Similarly, depersonalization might be expected to result from
helping professionals following the "rules of the game" such as expecta-
tions for maintaining a highly professional demeanor (Maslach, 1982). If
professional socialization is an important aspect of depersonalization.
1993 ^ ^ Cordes and Dougherfy 645

then a question may be raised about the relevance of depersonalization in

the burnout of employees ofher than helping professionals (Garden, 1987).
That is, the depersonalization component of burnout may not occur to the
same extent for employees in occupations that do not have strong norms
for maintaining a professional, impersonal, or even stoic demeanor in the
face of interpersonal stress.
Additionally, gender may be a factor in the occurrence of deperson-
alization, and men are more likely to cope with emotional exhaustion by
depersonalizing. Maslach and Jackson {1985) noted that, because of sex-
role socialization, women emphasize caring, nurturing, and showing con-
cern for others, and they would be less likely than men to exhibit callous,
impersonal behaviors. Several studies have found this to be true {Gold,
1985; Maslach & Jackson, 1981, 1985; Schwab & Iwanicki, 1982b).
A related but distinct factor in depersonalization is an employee's
experience of working in a depersonalizing work environment, one that is
characterized by rigid and controlling administrative practices, a lack of
participation (Jackson et al., 1986; Savicki & Cooley, 1983), or receipt of
noncontingent punishment (Jackson et al., 1986). Lack of participation
gives employees a feeling of lack of "control" over critical aspects or
demands of their work. Control is important for the perception of stress
{McGrath, 1976). Jackson and her colleagues (1986), for example, hypoth-
esized and found that lack of participation in decision making was linked
to depersonalization. Further, when an individual perceives environmen-
tal conditions as being random or uncontrollable, as after receipt of non-
contingent punishment, a feeling of helplessness or uncertainty ensues
{Chemiss, 1980b). In order to cope with the situation, individuals will
mechanize, or depersonalize, their relationships with co-workers, clients,
or the organization. Thus, impersonal, dehumanizing organizational cul-
tures can lead to employees' use of impersonal, dehumanizing styles with
their clients.

Proposition 2a: A high level of emotionaJ exhaustion

from one's worJc is the primary determinant of deperson-
alization. This is especially true for men and for individ-
uals who have been socialized to act in a highly profes-
sional and impersonal fashion. That is, when a person
experiences emotional exhaustion, depersonalizing
clients is an expedient coping mechanism.

Proposition 2h: When experiencing emotional exhaus-

tion, employees are more likely to depersonalize clients
when they operate in an impersonal, bureaucratic, or
rigid and controlling work environment, -which includes
factors such as (a) a lack of participation in decision
making and (b) the receipt of noncontingent punish-
ment.
646 Academy o/ Management Review October

Diminished personal accomplishment results in part from high levels

of depersonalization. When individuals develop negative, cynical atti-
tudes and withdraw physically or psychologically from the situation, they
find that they are no longer willing, or perceive that they are no longer
able, to perform their jobs effectively. Feelings of diminished personal
accomplishment also result from factors that suggest that one is unap-
preciated, that one's efforts are ineffective (Jackson et al., 1987), or that
one's competence or performance is low (Burke et al., 1984b). The percep-
tion of self-efficacy, defined as "people's judgments of their capabilities to
organize and execute courses of action required to attain designated
types of performances" (Bandura, 1986; 391), is at the core of the personal
accomplishment component (Lee & Ashforth, 1990). Self-referent misgiv-
ings are central to the stress experience (Bandura, 1986). Therefore, vari-
ables that cause one to doubt one's competence, abilities, and so forth,
should lead to feelings of diminished personal accomplishment.
The causes and consequences of diminished personal accomplish-
ment have not been as well understood as those associated with emo-
tional exhaustion (Maslach & Jackson, 1984). However, because individ-
ual assessments of task attributes (Gist & Mitchell, 1992) and attributions
of performance successes and failures (Bandura, 1986) influence self-
efficacy, we would expect variables that tend to make one feel lacking in
control, helpless, inadequate, or incompetent will contribute to feelings
of diminished personal accomplishment. For example, qualitative work
overload—the perception that one lacks the skills or abilities to ade-
quately perform the job (Kahn, 1978)—should be an important determi-
nant of feelings of personal accomplishment. When successful service is
defined as curing an illness or helping a person in distress (Maslach,
1982), and clients return only if the problems continue or recur, seeing the
same clients return again and again can leave the employee feeling that
he or she does not possess the capabilities to perform effectively. This
attribution of personal failure has been found to lead to lowered self-
efficacy (Gist & Mitchell, 1992; Maslach, 1978, 1982).
Roie ambiguity also affects self-efficacy and one's feelings of per-
sonal accomplishment. It is difficult to develop strong feelings of efficacy
when one feels unsure of what is expected of one's performance (Cher-
niss, 1980b) or when there is ambiguity concerning performance feedback
(Bandura, 1986). An individual's sense of personal accomplishment may
be undermined by ambiguity-induced suboptimal performance (whether
real or imagined). This ambiguity makes it difficult for employees to per-
form (or perceive they are performing) at an optimal level. Several studies
among teachers (Jackson et al., 1986; Schwab & Iwanicki, 1982a) and hu-
man service professionals (Brookings et al., 1985) support the relationship
between role ambiguity and feelings of diminished personal accomplish-
ment.
Another factor that can lower self-efficacy is the lack of performance-
contingent rewards (Jackson et al., 1986; Maslach, 1982) because feedback
information influences self-efficacy; employees infer meaning from infor-
1993 Cordes and Dougherty 647

mation cues in the work environment (Gist & Mitchell, 1992). Because
receipt of rewards is often the only way in which employees can gauge
their performance and what others think of their work (Schwab, Jackson,
& Schuler, 1986), and employees expect organizations to recognize and
reward good performance, a lack of contingent rewards can cause em-
ployees to feel that their performance does not warrant rewards or that
their efforts simply are not noticed or appreciated.
Unmet organization and achievement expectations are also likely to
contribute to a sense of diminished personal accomplishment. Whereas
level oi expectations serves as an internal demand as discussed previ-
ously, unmet expectations, regardless of the initial level of expectations,
suggests a failure to perform. When that failure is attributed to personal
inadequacies, feelings of personal accomplishment will be diminished.
Employees enter the workplace with certain expectations regarding both
what they will be able to accomplish professionally and what the orga-
nization will offer to reward and facilitate their work. Over time, these
expectations are contrasted with daily realities. The magnitude of these
unmet expectations, that is, the size of the discrepancy between initial
expectations and the current situation, influences the perception of per-
sonal accomplishment through the individual's attributions regarding the
cause for these discrepancies. If the discrepancy is small, then the effect
on perceived personal accomplishment should be small. In contrast,
when these expectations vary substantially from the realities of the work-
place, the employee is likely to perceive his or her performance as inad-
equate or ineffective {Stevens & O'Neill, 1983).
The focus in both the conceptual and empirical literature has been on
the extent to which individuals' achievement expectations are unmet.
Burnout has been most often studied in the helping professions, where the
difference between experience and expectations about client interactions
has been a central focus. Organizational expectations may also contrib-
ute to feelings of diminished personal accomplishment, especially in in-
dustrial and corporate settings.
In summary, diminished personal accomplishment results primarily
from depersonalization and factors that suggest one is ineffective, incom-
petent, or unappreciated. Especially relevant factors include qualitative
work overload, role ambiguity, lack of performance-contingent rewards,
and unmet organization and achievement expectations.
Proposition 3a: High levels of depersonalization will
cause individuals to alter their attitudes and interac-
tions with ciients, co-workers, or the organization in
such a way as to interfere with or inhibit the perception
of effective performance, resulting in feelings of dimin-
ished personal accomplishment.
Proposition 3b: High levels of factors that suggest one's
efforts are inadequate, ineffective, or unappreciated are
the primary determinants of feelings of diminished per-
646 Academy ol Management Review October

sonai accompJishment. These factors include qualita-

tive work overload, role ambiguity, lack of perfor-
mance-contingent rewards, and unmet organization
and achievement expectations.

Finally, availability of coping resources moderates the burnout pro-

cess at three different points. First, it moderates the relationship between
the demand stressors and emotional exhaustion. Second, it moderates the
relationship between emotional exhaustion and depersonalization. Fi-
nally, it moderates the relationship between depersonalization and feel-
ings of personal accomplishment. Two common types of coping resources
are organizational social support and personal social support.
Social support, in general, leads the individual to perceive that oth-
ers can and will provide the resources necessary for them to deal suc-
cessfully with the interaction. In the organization, for example, credible
senior employees or peers may convince one with reassuring words that
one is capable of successful performance (Eden & Kinnar, 1991). Also,
social support may cause the individual to redefine the potential for harm
posed by the situation, or it may bolster the individual's perceived ability
to cope with the demands imposed by the situation. Several researchers
have reported the effects of informal support (Leiter, 1988), professional
support (Leiter, 1988), co-worker support (Jackson et al., 1986; Leiter, 1991),
and supervisor support (Jackson et al., 1986). A related conceptualization
of social support is skill utilization (Leiter, 1990) or job challenge (Friesen
et al., 1988). Leiter (1990, 1991) argued that skill utilization is a form of
organization support because it provides opportunities for the implemen-
tation and development of skills and presents empirical evidence that
skill utilization is positively related to personal accomplishment.
Personal resources also represent a source of support that enhances
one's ability to cope with the demands of and reactions to the work envi-
ronment. Individuals who are married, for example, report lower levels of
burnout than their single counterparts (Gold, 1985; Maslach & Jackson,
1985; Russell et al., 1987). A possible explanation for this is that a spouse
is perceived as a source of support, providing a buffer between the stress-
ful work environment and adverse reactions to it (Leiter, 1990). Alterna-
tively, a spouse may enforce a more balanced life perspective by requir-
ing the employee to spend time away from the work environment.
Likewise, individuals who have children report lower levels of all three
components of burnout than do their childless peers (Maslach & Jackson,
1985). It is also possible that people with families are simply older and
thus more mature and stable. Whether older or not, people with families
may be more experienced in dealing with personal problems and emo-
tional conflicts. It may be that the family, in addition to being a source of
role conflict, is also a source of emotional support and comfort (Cohen &
Wills, 1985). The family helps the individual cope with work by providing
a much-needed balance of perspective. The family also may fill the indi-
1993 Cordes and Dougheity 649

vidual's needs for affection and approval, which are not being fulfilled
professionally.
Proposition 4a: The availability of coping resources is an
especially relevant buffer in the burnout process be-
cause it moderates (a) the relationship between the
demand stressors and emotional exhaustion, (b) the re-
lationship between emotional exhaustion and deper-
sonalization, and (c) the relationship between deperson-
alization and feelings of personal accomplishment.

Proposition 4b: More specifically, the availability of or-

ganizational social support and personal social support,
in particular spouse and family, moderate the experi-
ences of emotionai exhaustion and depersonaiization
and enhance feelings of personal accomplishment.

RESEARCH STRATEGIES

Despite the extensive literature on burnout to date, strategies other

than the existing ones may offer a significant benefit to this field. Thus
far, studies have used only a few types of methodologies. Of the studies
referenced here, only the earliest investigations by Maslach and Pines
used a case study design (Maslach & Pines, 1977; Pines & Maslach, 1980),
seven others used some type of structural equation modeling (Dignam et
al., 1986; Lee & Ashforth, 1990, 1993; Leiter, 1988, 1990, 1991; Saxton et al.,
1991), and only five used a longitudinal design (Cherniss, 1992; Fimian &
Blanton, 1987; Firth & Britton, 1989; Jackson et al., 1986; Leiter, 1990). The
remaining studies used cross-sectional correlational designs.
The more prevalent cross-sectional correlational studies of burnout
would benefit from increased attention to research rigor, especially using
statistical control to rule out third-variable explanations of correlation or
regression findings. For example, the nature of interpersonal interactions
is a key factor in the experience of burnout, but (unmet) job expectations
are also important. Multivariate analyses could clarify these relation-
ships, compared to studies examining individual zero-order correlations.
Similarly, some of the mixed findings about gender effects and "specialty
area" may be clarified if variables such as occupation and nature of
interpersonal interactions are controlled for in statistical analyses. Mul-
tivariate designs are also necessary to adequately assess the role of the
components in the burnout process. For example, emotional exhaustion
and depersonalization, in particular, are very highly correlated (Koeske &
Koeske, 1989; Lee & Ashforth, 1990; Wolpin et al., 1991), so it is difficult to
determine the unique contribution of noncontingent punishment without
controlling for the effect of emotional exhaustion.
Another improvement would be the measurement of precursors and
650 Academy of Managemenf fieview October

consequences of burnout at two or more points in time (e.g., Jackson et al.,

1986). Such action can at least diminish method variance problems (Pod-
sakoff & Organ, 1986), if not provide convincing support for causal rela-
tionships. Similarly, causal modeling (e.g.. Lee & Ashforth, 1993; Leiter,
1990, 1991) could provide a more informative examination of these pro-
cesses, compared to the cross-sectional correlational designs that have
dominated the burnout literature thus far. For example, Leiter (1991) used
causal modeling to establish the causal link between types of coping and
support and each of the burnout components.
Conversely, qualitative research also could be valuable in the study
of burnout, especially in capturing richer descriptions of contextual fac-
tors and personal meanings surrounding burnout processes. In-depth in-
terviews may be optimal for discovery of personal meaning, whereas
observation might be more effective for understanding everyday actions,
organizing structures, and contexts (Handy, 1988).
In addition, experimental or quasi-experimental research, which is
virtually nonexistent in the burnout literature to date, would enhance the
internal validity of conclusions about burnout processes. A potentially
fruitful strategy would be the experimental investigation of intervention
strategies, such as training programs, for ameliorating bumout. It ap-
pears that organizations have, in fact, provided bumout training pro-
grams to help employees to identify the symptoms of burnout, the factors
most likely to contribute to it, and specific ways to cope with it. Evalua-
tions of the effects of these training programs have not been published.
Naturally occurring experiments would also provide insights into the
causes and consequences associated with bumout. For example, if an
organization is downsizing and the remaining employees now have
larger workloads, burnout levels could be monitored to identify changes.
More attention should be given to the process of bumout, including
the sequencing of the three phases of burnout. The phase model (Golem-
biewski & Munzenrider, 1984) and Leiter's (1989) evolving process model
are steps in this direction, although more testing and refinement are
needed (see, e.g., Lee & Ashforth, 1993). We have suggested that emo-
tional exhaustion leads to depersonalization, which, in turn, contributes
to diminished feelings of personal accomplishment. Low correlations be-
tween these first two components and feelings of personal accomplish-
ment, however, suggest the latter may be either independent of emotional
exhaustion and depersonalization or related in some very complex way.
Further, the temporal aspects of the antecedents and consequences of
burnout, as well as the three components themselves, have received no
attention in the literature. Investigators should examine when the indi-
vidual components of burnout emerge, how quickly burnout progresses,
and what factors affect the onset of burnout and the speed of progression.
Pairs of measurements have been taken at a variety of intervals, such as
6 months (Leiter, 1990), 12 months (Jackson et al., 1986), 18 months (Fimian
& Blanton, 1987), 2 years (Firth & Britton, 1989), and 12 years (Chemiss,
1993 Cordes and Dougherty 651

1992). Although these measurements represent a start, more than two

measurement points will be necessary to establish patterns and pro-
cesses over time.
A related measurement issue pertains to the individual's career
stage. Individuals' susceptibility to burnout or coping success will be a
reflection of their career stages; therefore, career stage is expected to
affect relationships between the antecedent variables and the burnout
components. This issue has important implications for the timing of the
measurement of antecedent and burnout variables. Also, the standard set
of background variables has not been fully or sufficiently investigated,
especially as it moderates the relationships between the other indepen-
dent and dependent variables.
Finally, although the convergent and discriminant validity of the MBI
has been investigated {Iwanicki & Schwab, 1981; Maslach & Jackson,
1981, 1986; Wolpin et ai., 1991), the process has been limited to factor
analyses and the investigation of patterns of correlations among vari-
ables. There has been no thorough, convincing analysis of the discrimi-^
nant and convergent validity using the Multitrait-Multimethod framework
{Campbell & Fiske, 1959). Because the MBI has emerged as the primary
means of measuring this phenomenon, it is only appropriate that its va-
lidity be rigorously and comprehensively investigated.
CONCLUSIONS
Based on the literature to date, we believe the burnout phenomenon
deserves more attention by researchers studying stress processes in or-
ganizations. Burnout appears to be a unique type of stress syndrome,
which includes perceptions of emotional exhaustion, a dehumanization of
clients in one's work, and perceptions of diminished personal accom-
plishment, and it has been clearly distinguished, both conceptually and
empirically, from other forms of stress. Burnout can be measured in a
reliable and valid fashion. A developing literature on burnout has begun
to clarify the position of burnout in a network of variables included in the
study of organizational behavior. This article has attempted to provide
some necessary structural and conceptual clarity to aid in future research
efforts.
REFERENCES
Anderson, M. B. G., & Iwanicki, E. F. 1984. Teacher motivation and its relationship to bum-
out. Educational Administration Quarteriy, 20: 109-132.
Ashforth, B. E.. & Lee. R. T. 1990. Defensive behavior in organizations: A preliminary model.
Human fle/a(ions, 43: 621-648.
Bandura, A. 1986. Sociai foundations of thought and action: A sociaJ cognitive theory. En-
gelwood Cliffs. NJ: Prentice-Hall.
Brookings, J. B., Bolton, B., Brown, C. E., & McEvoy, A. 1985. Self-reported job bumout
among female human service professionals. Journal oi Occupational Behavior, 6: lAi-
150.
 652 Academy of Management Review October

Burke, R. J. 1989. Toward a phase model of bumout: Some conceptual and methodological
concerns. Group & Organization Studies, 14: 23-32.
Burke, R. J., & Deszca, E. 1986. Correlates of psychological bumout phases among police
officers. Human Relations. 39: 487-502.
Burke, R. J., & Greenglass, E. R. 1989. Psychological bumout among men and women in
teaching: An examination of the Cherniss model. Human Heiations, 3: 261-273.
Burke, R. J., Shearer, J., & Deszca. G. 1984a. Burnout among men and women in police work:
An examination of the Chemiss model. Journal of Health and Human Resources Admin-
istration, 7: 162-188.
Burke, R. J., Shearer, I., 8f Deszca, G. 1984b. Correlates of bumout phases among police
oificers. Group & O/ganization Studies. 9: 451-466.
Cahoon, A. R., & Rowney, J. I. A. 1984. Managerial burnout: A comparison by sex and level
of responsibility. Journal of Health and Human Resources Administration. 7: 249-263.
Cahoon, A. R., & Rowney, J. I. A. 1989. A comparison of bumout across public/private sector
managers. Journal of Health and Human Resources Administration. 12: 79-94.
Campbell D. T., & Fiske, D. W. 1959. Convergent and discriminant validation by the Mul-
titrait-Multimethod Matrix. Psychological Bulletin. 56: 81-105.
Captan, G. 1974. Support systems and community mental health: Lectures on concept de-
veiopinenf. New York: Behavioral Publications.
Cherniss. C. 1980a. Professional burnout in human service organizations. New York: Prae-
ger.
Chemiss, C. 1980b. Staff bumout: Job stress in the human services. Beverly Hills. CA: Sage.
Chemiss, C. 1992. Long-term consequences of burnout: An exploratory study. Journal of
Organizational Behavior. 13: 1-11.
Cohen, S., & Wills, T. A. 1985. Stress, social support, and the buffering hypothesis. Psycho-
logical Bulletin. 98: 310-357.
Constable, J. F., 8E Russell, D. W. 1986. The effect of social support and the work environ-
ment upon bumout among nurses. Journal of Human Stress. Spring: 20-26.
Daley, M. R. 1979. Bumout: Smoldering problem in protective services. Sociai Wort, Sep-
tember: 375-379.
Dignam, J. T., Barrera, M., Jr., & West. S. G. 1986. Occupational stress, social support, and
bumout among correctional officers. American Journal of Community Psychologv 14-
177-193.
Dubin, R. 1956. Industrial workers' worlds: A study of the central life interests of industrial
workers. Social Problems. 3: 131-142.
Eden, D., & Kinnar, J. 1991. Modeling galatea: Boosting self-efficacy to increase volunteer-
ing. Journal of Applied Psychology. 76: 759-780.
Fimian, M. I., & Blanton, L. P. 1987. Stress, bumout, and role problems among teacher
trainees and first year teachers. Journal of Occupational Behaviour. 8: 157-165.
Firth, H., & Britton, P. 1989. Bumout. absence and turnover amongst British nursing staff.
Journal of Occupational Psychology. 62: 55-59.
Fisher, C. D., & Gitelson, R. 1963. A meta-analysis of the correlates of role conflict and
ambiguity. Journal o/ Appiied Psychology. 68: 320-333.
French, J. R. P., Jr., & Caplan, R. D. 1972. Organizational stress and individual strain. In
A. J. Marrow (Ed.). The failure of success: 30-86. New York: AMACOM.
Freudenberger, H. J. 1974. Staif bum-out. Journal oi Social Issues. 30, 159-165.
1993 Coides and Dougherty 653

Freudenberger, H. I. 1977a. Burn-out: Occupational hazard of the child care worker. Child
Care Quarterly. 6: 90-99.
Freudenberger, H. J. 1977b. Speaking from experience. Burn-out: The organizational men-
ace. Training and DeveJopment Journal. 31{July): 26-27.
Friesen, D.. & Sarros, J. C. 1989. Sources of burnout among educators. Journal of Organiza-
tional Behavior. 10: 179-188.
Friesen, D.. Prokop, C. M., & Sarros, J. C. 1988. Why teachers bum out. Educationai Re-
search Quarterly, 12: 9-19.
Gaines, J., & Jermier, J. M. 1983. Emotional exhaustion in a high-stress organization. Acad-
emy of Management Journal, 26: 567-586.
Ganster, D. C , & Schaubroeck, ]. 1991. Work, stress and employee health. youmaJ of Man-
agement, 17, 235-271.
Garden, A. 1987. Depersonalization: A valid dimension of bumout? Human ffeJafions. 40:
545-560.
Gist, M. E., & Mitchell, T. R. 1992. SeH-efficacy: A theoretical analysis of its determinants
and malleability. Academy of Management fleview, 17: 183-211.
Gold, Y. 1984. The factorial validity of the Maslach Burnout Inventory in a sample of Cali-
fornia elementary and junior high school classroom teachers. Educationai and Psycho-
logical Measurement. 44: 1009-1016.
Gold, Y. 1985. Does teacher bumout begin with student teaching. Educafion. 105: 254-257.
Golembiewski, R. T. 1984. An orientation to psychological bumout: Probably something old,
definitely something new. Journal ol Health and Human Resources Administration, 7:
153-161.
Golembiewski, R. T. 1989. A note on Leiter's study: Highlighting two models of bumout.
Group St Organization Studies, 14: 5-13.
Golembiewski, R. T., 8E Munzenrider, R. 1981. Efficacy of three versions of one burn-out
measure: MBI as total score, sub-scale scores, or phases? Journal oi Health and Human
Resources Administration, 228-246.
Golembiewski, R. T., & Munzenrider. R. 1984. Phases of psychological burn-out and orga-
nizational co-variants: A replication using norms from a large population. Journal oi
Health and Human Resources Administration, 7, 290-323.
Green. D., & Walkey, F. H. 1988. Confirmation of the three-factor structure of the Maslach
Burnout Inventory. Educational and Psychological Measurement, 48: 579-585.
Handy, J. A. 1988. Theoretical and methodological problems within occupational stress and
burnout research. Human Relations, 41: 351-369.
Iwanicki, E. F., & Schwab, R. L. 1981. A cross-validation study of the Maslach Bumout
Inventory. Educational and PsychoJogicai Measurement, 41: 1167-1174,
Jackson, S. E. 1984. Organizational practices for preventing bumout. In A. S. Sethi & R. S.
Schuler (Eds.). HandbooJr of organizationa/ stress coping strategies; 89- 111. Cambridge,
MA: Ballinger.
Jackson, S. E.. & Maslach, C. 1982, After-effects of job-related stress: Families as victims.
Journal oi Occupational Behaviour. 3: 63-77.
Jackson, S. E., & Schuler. R. S. 1983. Preventing employee burnout. Personnel 60(2): 58-68.
Jackson, S. E., & Schuler, R. S. 1985. A meta-analysis and conceptual critique of research on
role ambiguity and role conflict in work settings. OrganizafionaJ Behavior and Human
Decision Processes, 36: 16-78.
654 Academy of Management Review October

Jackson, S. E., Schwab, R. L., & Schuler, R. S. 1986. Toward an understanding of the bumout
phenomenon. Journal ol Applied Psychology. 71: 630-640.
Jackson, S. E., Tumer. J. A., & Brief, A. P. 1987. Correlates of burnout among public service
lawyers. Journal of Occupational Behavior. 8: 339-349.
Kahili, S. 1988. Symptoms ol professional bumout: A review of the empirical evidence.
Canadian Psychology. 29: 284-297.
Kahn, H- 1978. Job bumout: Prevention and remedies. PuWic Welfare. 36: 61-63.
Kahn, R. L., Wolfe, D. M., Quinn, R. P,, Snoek, J. D., &Rosenthal, R. A. 1964. Organiiational
stress: Studies in role conflict and ambigrufty. New York: Wiley.
Kirmeyer, S. L., & Dougherty, T. W. 1988. Work load, tension, and coping: Moderating effects
of supervisor support. Personnel Psychology. 41: 125-139.
Koeske, G. F., & Koeske, R. D. 1969. Construct validity of the Maslach Burnout Inventory: A
critical review and reconceptualization. Journal of Applied Behavioral Science, 25(2):
131-132.
Lee, R. T., & Ashforth, B. E. 1990. On the meaning of Maslach's three dimensions of bumout.
Journal of Applied Psychology. 75: 743-747.
Lee, R. T., & Ashforth, B. E. 1993. A further examination of managerial buxnout: Toward an
integrated model. Journal of Organizational Behavior. 14: 3-20,
Leiter, M. P. 1988. Bumout as a function of communication pattems. Group 8t Organization
Studies. 13: 111-126.
Leiter, M. P. 1989. Conceptual implications of two models of bumout. Group & Organization
Studies, 14: 15-22.
Leiter, M. P. 1990. The impact of family resources, control coping, and skill utilization on the
development of burnout: A longitudinal study. Human fleiaffons, 43: 1067-1063.
Leiter, M. P. 1991. Coping pattems as predictors of burnout: The function of control and
escapist coping pattems. Journal of Organizational Behaviour. 12: 123-144.
Leiter. M. P., & Maslach, C. 1986. The impact of interpersonal environment on bumout and
organizational commitment. Journal oi Organizational Behavior. 9: 297-306.
Leiter, M. P., & Meechan, K. A. 1986. Role stmcture and bumout in the field oi human
services. Journal of Applied Behavioral Science. 22: 47-52.
Lemkau. J. P.. Rafferty, J. P., Purdy, R. R., & Rudisill, J. R. 1967. Sex role stress and job
burnout among family practice physicians. Journal of Vocational Behavior. 31: 81-M.
Lodahl, T. M., & Kejner, M. 1965. The definition and measurement of job involvement. Jour-
nal of Applied Psychology. 49: 24-33.
Maslach, C. 1976. Burned-out. Human Behavior, 5(9): 16-22.
Maslach. C. 1976. The client role in staff bumout./ouraai o/Social Issues, 34: 111-124.
Maslach, C. 1982. Bumout: The cost of caring. Englewood Cliffs, NJ: Prentice-Hall.
Maslach, C , & Jackson, S. E. 1981. The measurement of experienced burnout, /oumal o/
Occupationai Behaviour, 2: 99-113.
Maslach, C , & Jackson, S. E, 1984. Burnout in organizational settings. In S. Oskamp (Ed.),
Applied social psychology annual: Applications in organizational settings, vol. 5: 133-
153. Beverly Hills. CA: Sage.
Maslach, C , & Jackson, S. E. 1965. The role of sex and family variables in bumout. Sex
Roles, 12: 637-851.
Maslach. C . 8t Jackson, S. E. 1986. The Maslach flumout Inventory. Palo Alto, CA: Consult-
ing Psychologists Press.
1993 Cordes and Dougherty 655

Maslach. C . & Pines, A. 1977. The bum-out syndrome in the day care setting. Child Care
Quarterly, B: 100-113.
McGrath. J. E, 1976. Stress and behavior in organizations. In M. D. Dunnett (Ed.). Handbook
oi industrial and organizational psychology: 1351-1396. Chicago: Rand-McNally.
Meier, S. T. 1983. Toward a theory of burnout. Human Relations, 36: 899^910.
Meier, S. T. 1984. The constmct validity of bumout. Journal oi Occupational Psychology. 57:
211-219.
Miles, R. H., & Perreault, W. D., Jr. 1976. Organizational role conflict: Its antecedents and
consequences. Organizationai Behavior and Human Perfonnance. 17: 19-44.
Parasuraman, S., & Alutto, J. 1981. An examination of the organizational antecedents of
stressors at work. Academy of Management/oumaJ, 24: 48-67.
Parkington. ]. J., & Schneider, B. 1979. Some correlates of experienced job stress: A boundary
role study. Academy ot Management Journal. 22: 270-281.
Perlman, B., & Hartman, E. A. 1982. Burnout; Summary and future research. Human ReJa-
(ions, 35: 283-305.
Pines, A.. & Maslach, C. 1978. Characteristics of staff bumout in mental health settings.
Hospital and Community Psychiatry. 29: 233-237.
Pines, A., & Maslach. C. 1980. Combatting staff bimi-out in a day care center: A case study.
Chiid Care Quarterly, 9: 5-16.
Pines, A. M., & Aronson, E. 1981. Burnout: From tedium to personal growth. New York:
Macmillan.
Podsakoff, P. M., & Organ, D. W. 1986. Self-reports in organizational research: Problems and
prospects. Journal of Management, 12: 531-544.
Porter, L. W.. & Steers. R. M. 1973. Organizational, work, and personal factors in employee
tumover and absenteeism. PsychoJogicai Bulletin, 80: 151-176.
Pretty, G. M. H., McCarthy, M. E., & Catano, V. M. 1992. Psychological environments and
bumout: Gender considerations within the corporation. Journal oi Organizational Be-
havior, 13:701-711.
Roimtree, B. H. 1984. Psychological bumout in task groups: Examining the proposition that
some task groups of workers have an affinity for bumout, while others do not. Journal oi
Health and Human Resources Administration, 7: 235-248.
Russell, D. W., Altmaier, E.. & Van Velzen. D. 1987. Job-related stress, social support, and
burnout among classroom teachers. Journal of Applied Psychology, 72: 269-274.
Savicki, R.. & Cooley, E. 1983. The relationship of work environment and client contact to
bumout in mental health professionals. Paper presented at the annual meeting of the
American Psychological Association. San Francisco.
Saxton, M. J., Phillips. J. S.. & Blakeney. R. N. 1991. Antecedents and consequences of emo-
tional exhaustion in the airline reservations service sector. Human Relafions, 44: 583-
595.
Schaufeli. W. B., Maslach, C , & Marek. T. (Eds.). In press. Professiona/ bumouf: Recent
developments in theory and research. Washington, DC: Taylor & Francis.
Schuler, R. S. 1980. Definition and conceptualization of stress in organizations. Organization
Behavior and Human Performance. 25: 184-215.
Schwab, R. L.. & Iwanicki, E. F. 1982a. Perceived role conflict, role ambiguity, and teacher
bumout. Educational Administration Quarterly, 18: 60-74.
Schwab. R. L., & Iwanicki, E. F. 1982b. Who are our burned out teachers? Educational Re-
search Quarterly. 7: 5-16.
656 Academy of Management Review October

Schwab, R. L., Jackson, S. E., & Schuler, R. S. 1966. Educator bumout: Sources and conse-
quences. Educational Research Quarterly. 10: 14-30.
Shirom, A. 1969. Bumout in work organizations. In C. L. Cooper & I. Robertson (Eds.), Inter-
national review of industrial and organizational psychology: 25-46. New York: Wiley.
Stevens, G. B., & O'Neill, P. 1963. Expectation and bumout in the developmental disabilities
field. American Journal of Community Psychology. 11: 615-627.
VanYperen, N. W.. Buunk, B. P., & Schaufeli. W. B. 1992. Communal orientation and the
bumout syndrome among nurses. Journal of Applied Social Psychology. 22: 173-189.
Wanous. J. P. 1973. Effects of a realistic job preview on job acceptance, job attitudes, and job
survival. Joumal of Applied Psychology. 58: 327-332.
Wanous, J. P. 1976. Organizational entry: From naive expectations to realistic belieis. Jour-
nal of Applied Psychology. 61: 22-29.
Wolpin, J., Burke, R. J., & Greenglass. E. R. 1991. Is job satisfaction an antecedent or a
consequence of psychological burnout? Human Relations, 44: 193-209.
Zabel, R. H.. & Zabel, M. K. 1982. Factors in bumout among teachers of exceptional children.
Exceptional Children. 49: 261-263.

Cynthia L. Cordes received her Ph.D. in management from the University of Mis-
souri. She is a visiting assistant professor of management at Binghamton University
{SUNY). Her current research interests include job stress and burnout, mentoring,
and information asymmetry.
Thomas W. Dougherty received his Ph.D. in industrial-organizational psychology
irom the University of Houston. He is a professor of management at the University oi
Missouri. His current research interests include job stress, determinants of career
success, and recruiting/interviewing processes.