Psychology Week 10 Assignment

EMOTIONAL EXHAUSTION AND TURNOVER INTENTION IN HUMAN SERVICE OCCUPATIONS: THE PROTECTIVE ROLE OF COWORKER SUPPORT

Lori J. Ducharme

Institute for Behavioral Research, University of Georgia, Athens, Georgia, USA

Hannah K. Knudsen

College of Behavioral Medicine, University of Kentucky, Lexington, Kentucky, USA

Paul M. Roman

Department of Sociology, University of Georgia, Athens, Georgia, USA

Human services occupations are prone to high rates of burnout and turn- over. These outcomes have adverse implications for service providers and the clients in their care. Several studies have assessed the structural causes and consequences of burnout and turnover, but little attention has been paid to the potentially protective role of coworker support. We estimate a structural equation model including job characteristics, co- worker support, and workplace justice to predict turnover intention, both directly and indirectly through emotional exhaustion. More than 1,800 substance abuse treatment counselors provided survey data for these analyses. Net of demographic and workload measures, low autonomy,

This research was supported by grants from the National Institute on Drug Abuse

(R01DA13110 and R01DA14482). The opinions expressed are those of the authors and do

not represent the official views of the funding agency.

Address correspondence to Lori J. Ducharme, Institute for Behavioral Research, 101

Barrow Hall, University of Georgia, Athens, GA, 30602-2401, USA. E-mail: [email protected]

Sociological Spectrum, 28: 81–104, 2008

Copyright # Taylor & Francis Group, LLC

ISSN: 0273-2173 print/1521-0707 online

DOI: 10.1080/02732170701675268

and a lack of distributive justice significantly predicted emotional exhaustion, while coworker support was inversely associated with exhaustion. In turn, exhaustion was a significant predictor of intent to quit. Coworker support exhibited direct inverse effects on intent to quit, while counselors reporting low autonomy and lack of workplace justice were more likely to be contemplating quitting. These findings extend pre- vious research by identifying specific and beneficial effects of coworker support on counselors’ job-related affect and well-being.

As the U.S. economy shifts increasingly to service work from production occupations, more attention is drawn to the unique stres- sors and health outcomes of workers engaged in emotional labor in the day-to-day context of interacting with clients and customers (Hochschild 1983; Wharton 1999). Much attention has been paid to the service sector, and the repeated surface acting roles in which workers engage in order to complete discrete, time-limited transac- tions (e.g., in-store sales, restaurant service, call centers). In routine service work, employee-customer exchanges are such that neither party has a high emotional investment in the encounter, nor is the interaction highly emotionally charged. Such interactions, however, occur with great frequency in these occupations. By contrast, other forms of service work—especially counseling and other human ser- vice occupations—are characterized by less frequent but more intense encounters whose success depends in large part upon the worker’s ability to make a significant and sustained emotional investment in the client ‘‘transaction.’’

Several studies have failed to find a link between frequency of service work interactions and emotional exhaustion or burnout (Bulan et al. 1997; Morris and Feldman 1996); however, emotional exhaustion has often been associated with occupations characterized by a greater intensity of interpersonal interactions, such as counseling, nursing, teaching, and other ‘‘caring’’ or ‘‘helping’’ professions (Maslach 1978). A key distinction between routine service work and the helping professions is that the former is characterized by the deliv- ery of a discrete service or completion of a fixed transaction, whereas the latter involves the evoking of some change in the client, whether physical, emotional, or cognitive. Counseling occupations represent one segment of the broader service industry in which the successful delivery of care is fundamentally dependent on the counselor’s ability to establish a strong, lasting, and meaningful bond with the client. Counselors generally refer to this relationship as the ‘‘therapeutic alliance,’’ literally referring to the ‘‘emotional bond and . . . shared pre- sumption regarding the tasks and goals of the treatment endeavor’’

82 L. J. Ducharme et al.

between counselor and client (Connors et al. 1997, p. 588). In this cir- cumstance, emotional labor requires deep acting—an ability to empathize with a client, identify her needs, advocate for her care, and stimulate change in her attitudes and behaviors. It seems reason- able, then, to posit that the experience and impact of stress emanating from this class of service encounters may be substantially different than the stressors encountered in jobs requiring a lesser investment on the part of other service workers. As such, the potential mediating effects of organizational contexts and work environments for these occupations warrants exploration.

In human services occupations such as substance abuse and mental health counseling, clients often bring a collection of disparate service needs and related issues to the treatment encounter. Addressing these co-occurring and ancillary needs, and coordinating the receipt of ser- vices with the treatment of the client’s underlying diagnosis, can tax counselors’ energy and emotional resources. Moreover, addiction cli- ents treated in even the ‘‘best’’ facilities are prone to high rates of relapse, yielding frustration among clinicians who invest significant emotional resources in building their therapeutic alliance. Over time, the cumulative effects of managing a client caseload can lead to fati- gue, exhaustion, and burnout (Budrionis et al. 1987; Niehoff 1984). As a result, turnover rates among addiction treatment counselors are high (McLellan et al. 2003), and are a source of constant concern for program managers (Gallon et al. 2003).

There is a substantial accumulated literature in the sociology of work and occupations that has examined the structural features of work environments and the content of work, and linked these to job satisfaction, organizational commitment, and retention. An increasing emphasis has been placed on the social climate of the work setting and its relationship to worker affect and well-being. Much of this social climate is a product of management practices and policies that lead employees to perceive that they are valued and integral members of the organization. Less emphasized in the context of work environment analyses is the potential beneficial role of coworker support in mediating the stresses inherent in the performance of emotional labor.

There is still much ground to be covered in integrating the structural and social elements of the work environment into a comprehensive model of workers’ health and behavior outcomes. As in many human services occupations, the addiction treatment counseling field is char- acterized by consistently high burnout and turnover rates despite var- iations in workload and management practices across occupational settings. Thus, an exploration of the interplay between the work

Exhaustion and Turnover Intention in Human Service 83

environment (including coworker relationships) and counselors’ retention and well-being not only has theoretical implications for the sociological study of work and health, but also practical implica- tions for the management of behavioral healthcare programs and, by extension, other human services organizations. In this article, we examine the effects of coworker support on the emotional exhaustion and turnover intentions of more than 1,800 behavioral health coun- selors employed in nationally-representative samples of substance abuse treatment programs.

BURNOUT AND EMOTIONAL EXHAUSTION IN COUNSELORS

‘‘Burnout’’ is a generic term that is commonly used to describe the exhaustion and distancing behaviors experienced by human services workers and other helping professionals in response to the excessive emotional demands that characterize their jobs (Jackson et al. 1986). In studies of human service workers, the various dimensions of burn- out have been associated with poorer job performance as well as adverse health outcomes. In terms of job performance, burnout has been associated with absenteeism, ineffectiveness, interpersonal con- flict, lower productivity, job dissatisfaction, reduced organizational commitment, and turnover (Maslach and Jackson 1986). Of parti- cular importance for clients is the finding that burnout can be ‘‘con- tagious’’ among staff, and unit-level burnout among counselors and nurses has been associated with lower client satisfaction with services received (Garman et al. 2002; Vahey et al. 2004).

The emotional exhaustion component of burnout has repeatedly been associated with stress-related health outcomes. Emotional exhaustion predicts increased rates of illness, fatigue, and substance abuse (Cherniss 1980; Pines and Maslach 1978), as well as depression, anxiety, and irritability (Jayaratne et al. 1986). In their review of pub- lished literature, Glass and McKnight (1996) found a nearly uni- formly positive association between burnout and depressive symptomatology. Thus, the experience of conditions leading to emotional exhaustion can have significant implications for counse- lors, in terms of both their employment and well-being. In the context of this article, we differentiate between burnout as a general concept and emotional exhaustion as the specific, proximal outcome of inter- est. Given the documented associations between emotional exhaus- tion and adverse health outcomes, we focus on exhaustion as a key dependent variable in these analyses.

84 L. J. Ducharme et al.

TURNOVER INTENTION

In addition to the health impacts of overwork, another adverse out- come linked to exhaustion is turnover. Addiction treatment counsel- ing is generally regarded as an inherently stressful and frustrating occupation, which may predispose treatment program employees to higher turnover rates (Layne et al. 2004). Indeed, voluntary turnover among addiction treatment counselors is a significant problem for treatment facilities and their clients (Hser 1995; Laundergan et al. 1986), and counselor turnover rates range from 25% to 50% annually in the public sector (Gallon et al. 2003; McLellan et al. 2003). These rates surpass the national average of 19.2% annual voluntary turn- over (quits) across all occupations, and the 17.7% annual quit rate among occupations in healthcare and social assistance (Bureau of Labor Statistics 2004).

Turnover among human services occupations has adverse implica- tions not only for the organization, but also for the clients in its care. Turnover among counselors creates instability for both the treatment facility and for clients who depend on a stable therapeutic relation- ship to support their recovery (Hser 1995). As a result, clients may suffer setbacks such as a loss of trust in the treatment facility or the therapeutic process when their clinical services are disrupted by their primary counselor’s departure. Turnover can intensify stress among remaining counselors whose caseloads increase, and can have adverse effects on quality of care when newer, less-experienced staff are hired into vacant positions (Powell and York 1992). When turn- over rates are high, large numbers of clients within a treatment facility can be personally affected by the resulting discontinuity in service delivery (Lum et al. 1998). It is also worth noting that, while turnover has many adverse effects, clients are also affected when counselors experience burnout but remain with the organization. Thus, both burnout and turnover in this occupational group adversely impact client recovery, service quality, and organizational effectiveness (Mulvey et al. 2003).

PREDICTORS OF BURNOUT AND TURNOVER

Given the implications of counselors’ emotional exhaustion and turn- over for treatment facilities and their clients, it is important to iden- tify work-related predictors of these outcomes that might effectively be altered to enhance counselors’ job satisfaction, commitment, and long-term retention. Among human services occupations, workload (e.g., client caseload) is widely cited as a major contributor to both

Exhaustion and Turnover Intention in Human Service 85

burnout and turnover intention (Alexander et al. 1998; Janssen et al. 1999; Schaufeli and Enzmann 1998). Relatedly, the content of work tasks also plays an important role. For counselors, the percentage of work hours spent on activities other than counseling—for example, administrative paperwork, supervision, case management—has had documented effects on reported emotional exhaustion (Acker 1999; Johnson and Rubin 1983; Rubin 1984). Burnout itself is a strong predictor of intent to quit in counseling occupations (Blankertz and Robinson 1997; Drake and Yadama 1996).

Beyond the stresses inherent in managing a client caseload, broader elements of the work environment have also been associated with exhaustion and turnover intention among counselors. In parti- cular, management practices that result in rewarding work experi- ences have shown direct effects on reducing turnover intention (Griffeth et al. 2000; Jamal 1990; Vinokur-Kalpan et al. 1994). Among the elements of ‘‘rewarding’’ work are autonomy and decision latitude (Karasek 1979). Turnover intention has been found to be higher among clinicians reporting low levels of autonomy at work (Arches 1991; Cherniss 1980; Knudsen et al. 2003; Lecroy and Rank 1987).

Another key element of the work environment is perceived fairness of program management. Notably, in studies of mental health profes- sionals, perceived inequity on the job led to increased resentment, which predicted turnover (Geurts et al. 1998). By contrast, work set- tings characterized by higher levels of procedural and distributive justice tend to have lower turnover levels (Lum et al. 1998; Hendrix et al. 1999).

As in any environment, counselors bring with them to the work- place certain characteristics that are not amenable to change, but that nevertheless influence their propensity to experience exhaustion or quit their jobs. Among these, age and educational attainment signi- ficantly predict turnover, as younger, more highly-educated, and better-trained counselors are more marketable and mobile, and thus prone to seek more rewarding employment opportunities elsewhere (Blankertz and Robinson 1997; Jinnett and Alexander 1999). Simi- larly, professionalism (as measured by credentials such as certifi- cation and licensure) has been found to predict nursing turnover in hospitals (Bloom et al. 1992); however, once counselors become established within a treatment center, they may be less likely to leave, as is evidenced by the inverse association between tenure and turn- over (Bloom et al. 1992; Gray and Phillips 1994; Somers 1995).

The characteristics of individual counselors, their client caseloads, and their work environments have been examined in the context of

86 L. J. Ducharme et al.

studies on burnout, turnover intention, and actual turnover. While the relationships among these categories of variables have been fairly well documented, less attention has been paid to the role of interper- sonal relations between counselors and its potential protective effects against exhaustion, adverse health outcomes, job dissatisfaction, and turnover. This article seeks to extend existing research by considering the impact of coworker support on counselors’ affective and beha- vioral reactions to demanding and unrewarding work.

COWORKER SUPPORT AS A PROTECTIVE FACTOR

Few studies have directly examined the role of coworker support in the experience of exhaustion and intent to quit among counseling professionals. Within the limited published literature, studies that incorporate support measures tend to have methodological or measurement weaknesses that detract from the potential value of their findings. Notably, many of these studies measure supervisor or organizational support, but not coworker support, or they com- bine these three sources of workplace support into a single measure; others confound work-based support with more general coping stra- tegies or with support received from family and friends outside the workplace (Bennett et al. 1994; Layne et al. 2004; Lee and Ashforth 1993; Shoptaw et al. 2000). A more common issue is that these studies often rely on small sample sizes that do not represent a sufficiently diverse array of counselors or work settings to yield confidence in the substantive significance or generalizability of their findings (Layne et al. 2004; Shoptaw et al. 2000; Elman and Dowd 1997). In terms of studies that specifically focus on burnout or turnover among substance abuse counselors, these have not examined the potential influence of coworker support (e.g., Gallon et al. 2003; Knudsen et al. 2003, 2006).

Despite the limitations of these studies, their accumulated findings suggest that work-based social support may effectively reduce coun- selors’ experience of burnout and may directly or indirectly reduce their likelihood of leaving their jobs. In a field such as addiction counseling, in which high turnover rates are an ongoing challenge, there is great potential benefit to be gained from a better understand- ing of available resources that can be leveraged to enhance workforce stability within provider organizations. For example, Bennett and colleagues (1994) found an association between a general measure of social support and burnout among HIV=AIDS counselors, while Shoptaw and colleagues (2000) reported a negative association between work-based support and exhaustion. A number of studies

Exhaustion and Turnover Intention in Human Service 87

report significant inverse associations between coworker support and burnout generally (Pines and Maslach 1978; Poulin and Walter 1993; Lee and Ashforth 1993; Schaufeli and Enzmann 1998), as well as exhaustion specifically (Janssen et al. 1999). Notably, Bowden (1994) found that practical support (as distinct from affective or emotional support) from coworkers had a significant negative effect on burnout among mental health staff, while Brown and colleagues (2003) found that support from coworkers had more influence on exhaustion than did support from family, friends, or other nonwork sources. As in the broader support literature, debate persists as to whether support effectively buffers (i.e., moderates) the relationship between suboptimal working conditions and employees’ well-being, and mixed findings are reported in studies focusing on counseling occupations (Himle et al. 1989; Jayaratne and Chess 1984; Koeske and Koeske 1989; Shinn et al. 1984).

Research also suggests an inverse association between work-based social support and turnover in human services occupations. Lichtenstein and colleagues (2004) found that satisfaction with coworkers was inversely related to turnover intention among clinicians treating cli- ents with serious mental illness. In a study of child life specialists, Munn and colleagues (1996) found that supervisor support had direct effects on job satisfaction and turnover intention, but not on exhaustion. Alexander and colleagues (1998) found that coworker support had a significant negative effect on both intent to quit and actual turnover in a study of nurses in long-term psychiatric settings. In a review of the literature on turnover intention and actual turn- over in human services professionals, Acker (1999) concluded that social support (variously measured) had consistent negative effects on intent to quit and turnover. Similarly, other studies have found a direct relationship between work-based support and retention (Jayaratne and Chess 1984; Siefert et al. 1991; Schaefer and Moos 1996; Blankertz and Robinson 1997; Alexander et al. 1998; Jinnett and Alexander 1999).

While the findings of these studies generally support the contention that positive coworker relations can have beneficial effects for coun- selors’ careers and well-being, more research is needed to establish the influence of coworker support on these processes. Our review of the published literature on burnout and=or turnover among counselors, nurses, and case workers found that none of these studies simul- taneously measured coworker support, burnout=exhaustion, and intent to quit. As a result, the direct and indirect effects of support on counselors’ affective and behavioral reactions to stressful working environments have not yet been fully specified. To address this gap,

88 L. J. Ducharme et al.

this article examines the effects of workload, working conditions, and coworker support on the emotional exhaustion and turnover intention of counselors in substance abuse treatment settings. The proposed model replicates some prior exploratory research using a large, broadly representative sample of addiction counselors, and extends prior research by focusing explicitly on instrumental support received from coworkers.

METHODS

Sampling of Treatment Programs and Counselors

Data are drawn from questionnaires collected as part of a larger study of addiction treatment programs and their services. Data were gathered from counselors within two nationally representative sam- ples of treatment facilities. Treatment facilities were selected via a two-stage random sampling process, stratifying first on geographic location (county) and then sampling treatment facilities within strata. A total of N ¼ 401 privately funded substance abuse treatment pro- grams and N ¼ 362 publicly-funded programs participated in data collection activities occurring between Fall 2002 and Spring 2004. These facilities include for-profit and non-profit, hospital-based and freestanding programs providing inpatient, residential, and=or outpatient addiction treatment services. The number of participating centers represents 87% of the sampled private centers and 80% of the sampled public centers.

Administrators of each participating facility were asked to provide a list of all substance abuse counselors employed by the program. Each counselor was then mailed a questionnaire focusing on program management, workplace climate, services provided, and awareness and use of a variety of evidence-based treatment practices. The response rate for counselors was 62% in the private programs, and 61% in the public programs. Analyses comparing the demographic characteristics and credentials of responding counselors to the known characteristics of all counselors surveyed indicated no significant response biases in either sample (data not shown). Unweighted data from both counselor samples (total N ¼ 2,304) were pooled for these analyses, with the study sample (i.e., public sector or private sector) included as a control variable to account for any systematic differ- ences in working conditions between the two samples. All variables used in these analyses are person-level data obtained from counselor questionnaire responses.

Exhaustion and Turnover Intention in Human Service 89

Measures

Work Environment We use several objective measures of counselors’ workload as well as three more general measures of the work environment. First, counse- lors reported the total number of hours worked per week. In the frequency distributions presented in Table 1, actual hours are reported; in the subsequent models, the natural log transformation of this variable is used to adjust for skew. Counselors’ reports of exhaustion and intent to quit are expected to be positively associated with weekly work hours. Second, counselors reported the number of hours per week spent in direct patient care (i.e., counseling), as well as the number of hours in administrative activities such as meetings or paperwork, and hours in training activities. We computed the percentage of time per week spent in direct patient care. Previous research (cited above) suggests that counselors spending proportion- ately less time in direct patient care activities should be more prone to exhaustion and intent to quit; however, it is equally plausible that exhaustion increases directly with the amount of client contact. Third, we expect that caseload size (i.e., the number of patients assigned to the counselor) will be positively associated with exhaustion and turnover intention. We use dichotomous variables defining three roughly equal

Table 1. Descriptive statistics for observed variables

Sample Mean or %

Counselor Demographics

Women 61.2%

Nonwhite 27.4%

Age 45.4 years (sd ¼ 10.9)

Education and Credentials

Less than College 22.5%

College Degree 34.9%

Master’s or higher 42.6%

Certified 56.1%

In recovery 51.6%

Tenure at Center 5.1 years (sd ¼ 5.3)

Job Characteristics

Hours per Week 38.5 hours (sd ¼ 11.5)

% Time in Direct Care 47.3% (sd ¼ 16.8)

Small Caseload ( < 10) 42.8%

Medium Caseload (11–20) 24.8%

Large Caseload ( > 20) 28.6%

Annual Salary $25,000–$35,000

Public-Sector Program 52.4%

90 L. J. Ducharme et al.

groups of counselors by caseload size: small caseload (ten or fewer patients, used as the reference category), medium caseload size (11–20 patients) and large caseloads (21 or more patients).

Three constructed variables are used to measure the extent of decision latitude and fairness perceived by counselors in their work setting. Job autonomy was measured using three items adapted from Pritchard and Karasek (1973). These items asked counselors the extent to which they had sufficient authority and decision latitude to perform their job tasks. Distributive justice was measured using four items asking counselors to rate the extent to which they per- ceived the distribution of workload, compensation, and fringe bene- fits to be fair. Procedural justice was measured using eight items asking counselors to rate the extent to which management involves employees in decision-making processes, accepts feedback, and explains decisions in an adequate and timely manner. Both sets of organizational justice items were adapted from workplace research by Niehoff and Moorman (1993). Consistent with prior organiza- tional research using these items, we expect that counselors reporting greater levels of autonomy, distributive justice, and procedural justice will be less likely to report emotional exhaustion or to be planning to quit their jobs.

Coworker Support Eight items adapted from Tang (1999) were used to measure co- worker support. Counselors were asked to rate their agreement (on a seven-point scale) with statements about their coworkers’ helpfulness, knowledge, and creativity in problem-solving. These items generally tap instrumental and informational support, and include such state- ments as ‘‘I find my coworkers helpful when I encounter difficulties with my work,’’ and ‘‘My coworkers and I are able to come up with creative ideas when we face tough problems.’’

Outcome Variables The emotional exhaustion component of the Maslach Burnout Inven- tory (MBI; Maslach and Jackson 1986) was used to measure the first of two key outcome variables in these analyses. The MBI is a well- validated and widely used measure of ‘‘burnout’’ among human ser- vice professionals, and the emotional exhaustion component is clearly differentiated from other constructs measured by the full inventory (Maslach and Jackson 1981, 1986). The exhaustion component was used in this survey because it is the component of burnout with the most direct and consequential health implications (Cherniss 1980; Jayaratne et al. 1986; Pines and Maslach 1978).

Exhaustion and Turnover Intention in Human Service 91

Counselors’ intent to quit was measured using three items by Walsh, Ashford, and Hill (1985). These items asked counselors to rate their agreement (on a seven-point scale) with statements about the extent to which they are thinking about leaving their job or actively looking for work elsewhere. Measuring intention to quit rather than retrospectively examining quitters allows us to under- stand factors that are most salient during this decision-making pro- cess. In studies that have examined both intentions and actual behavior, intent to quit emerges as the single strongest predictor of actual turnover (Alexander et al. 1998; Hendrix et al. 1999).

Control Variables Several control variables are included in these analyses to account for systematic differences that may be introduced by counselors’ demo- graphic and background characteristics. First, our models control for gender (1 ¼ women, 0 ¼ men), race (1 ¼ nonwhite, 0 ¼ white), and age. Several additional variables control for counselors’ creden- tials and marketability, which may influence their turnover inten- tions. Education is measured using three dichotomous variables: less than a college degree (reference category), college degree, and master’s degree or higher. Because other analyses of this dataset indi- cate that counselors in recovery from their own addictions face sys- tematic disadvantages in terms of compensation, we control for recovery status (1 ¼ in recovery, 0 ¼ not in recovery). Certification status is measured using an additional variable (1 ¼ certified addic- tions counselor, 0 ¼ no certification). Annual salary is measured as a categorical variable ranging from 1 (under $15,000 per year) to 9 (more than $50,000 per year). Responses to the salary question were recoded to the category midpoints for analysis. Finally, because the responding counselors were drawn from two separate samples of treatment organizations, we include a dichotomous variable to con- trol for the sample from which they were drawn (1 ¼ publicly funded facilities, 0 ¼ privately funded). Summary statistics for all observed variables are shown in Table 1.

Analyses

The proposed model (depicted in Figure 1) was estimated using Mplus version 3.01, a structural equation modeling software package (Muthen and Muthen 2004). This software allows for the estimation of a measurement model and a structural model of hypothesized rela- tionships. The measurement model constructs latent variables from the shared variance between individual items. The primary advantage

92 L. J. Ducharme et al.

of latent variables is that error components from the items are par- celed out, resulting in an unobserved measure that is more valid and reliable. Mplus estimates the structural model of hypothesized relationships between latent variables and single item indicators, including direct as well as indirect effects. This software produces measures of overall model fit, estimates of the hypothesized associa- tions (unstandardized and standardized coefficients, standard errors, and t-tests), and measures of the proportion of variance explained for each dependent variable.

RESULTS

Preliminary confirmatory factor analyses (not shown) demonstrated that all indicators loaded significantly on their respective factors; however, the initial measures of model fit were below desired thresh- olds (Hu and Bentler 1999), so the revised model includes several pairs of correlated residuals between individual items. It is important to note that these pairs of correlated residuals are all within con- structs; no residuals were correlated that crossed constructs. The final confirmatory factor model (shown in Table 2) fit the data well across several measures. The CFI was .958, the root mean square error of

Figure 1. Hypothesized relationship between work characteristics, co-

worker support, emotional exhaustion, and intent to quit.

Exhaustion and Turnover Intention in Human Service 93

Table 2. Results of confirmatory factor analysis (latent variable constructs)

Factor Loadings

Autonomy:

I have sufficient authority to fulfill my job responsibilities. .832

I have enough freedom over how I do my job. .896

I have enough authority to make decisions necessary to provide

quality treatment.

.920

Procedural Justice:

Management makes sure employee concerns are heard before

decisions are made.

.864

Job decisions are applied consistently across all affected employees. .826

Employees are allowed to challenge=appeal job decisions made by

managers.a .754

Employees are involved in making decisions about how work is done.a .784

I trust that my supervisor is completely honest with me.b .777

I trust that my supervisor will share important information with me.b .778

When decisions are made, all affected people are asked for their ideas. .870

Management clarifies decisions and provides additional information

when asked.

.857

Distributive Justice:

The amount of pay employees receive is distributed fairly. .740

Employees receive an amount of fringe benefits that is fair. .621

The workload at this treatment center is fairly distributed. .774

The overall rewards received are fairly distributed. .902

Coworker Support:

My coworkers are very helpful when I encounter difficulties with my

work.c,d .779

In this treatment center, people show little interest in each other’s

work.

.568

I find my coworkers very helpful in sharing knowledge and

information.c,e

.763

My coworkers and I are able to come up with creative ideas to face

problems.

.743

In this center there are many employees with strong knowledge and

skills.f .700

My coworkers impress me with their innovative ideas and

resourcefulness.e,f,g

.775

My coworkers help others turn ideas into action and reality.d,g .829

When I encounter a problem, I usually seek help from my coworkers. .572

Emotional Exhaustion:

I am emotionally drained from my work.h,i .641

I feel fatigued when I get up and have to face another day on the job.h .772

Working with people all day is really a strain for me.j .639

I feel burned out from my work. .860

Working directly with people puts too much stress on me.j .645

I feel frustrated by my job. .784

I feel used up at the end of the workday.i,k .761

(Continued)

94 L. J. Ducharme et al.

approximation (RMSEA) was .033, and the standardized root mean square residual (SRMR) was .033. Hu and Bentler (1999) argue that a CFI > .95, RMSEA < .06, and SRMR < .08 are indicators of good model fit.

The structural model of emotional exhaustion and turnover inten- tion estimated associations between the work characteristics and exhaustion, direct effects of work characteristics on turnover inten- tion, and indirect associations between the work-related measures and turnover intention via exhaustion. Listwise deletion reduced the effective sample size to 1,869 cases. Importantly, no single item accounted for a disproportionate number of missing cases, although a total of 168 cases were lost due to one or more missing items in the procedural or distributive justice constructs. Significance tests (not shown) indicated that counselors with missing data on any of the independent variables were not significantly different in their level of reported exhaustion or turnover intention compared to counselors retained in the final models.

As seen in Table 3, coworker support was significantly associated with exhaustion, such that greater coworker support was associated with lower levels of counselors’ emotional exhaustion (b ¼ � .138, p < .001), net of the other variables in the model. Two of the three other workplace characteristics were also significant predictors of exhaustion. Counselors with greater job autonomy reported signifi- cantly lower exhaustion (b ¼ � .170, p < .001). Although there was not an association between procedural justice and exhaustion, coun- selors employed in centers characterized by greater distributive justice reported lower levels of emotional exhaustion (b ¼ � .286, p < .001). Interestingly, none of the workload measures was significant in this model. Neither the number of hours worked per week, the proportion

Table 2. (Continued )

Factor Loadings

I feel I’m working too hard on my job.k .707

I feel like I’m at the end of my rope. .701

Intent to Quit:

As soon as I find a better job, I will leave this treatment center.l,m .781

I am actively looking for a job at another treatment center.l,n .640

I am seriously thinking about quitting my job.m,n .661

I think I will be working for this treatment center five years from now.

(reversed)

.701

CFI ¼ .958, RMSEA ¼ .033, SRMR ¼ .033.

Note: Superscripts indicate items with correlated error terms.

Exhaustion and Turnover Intention in Human Service 95

of hours in direct care activities (as opposed to administrative or paperwork activities), nor caseload size were significantly associated with exhaustion, net of the effects of the other variables in the model.

Of the sociodemographic characteristics, several were associated with exhaustion. Older counselors (b ¼ � .062, p < .01) and counse- lors of a racial=ethnic minority background (b ¼ � .050, p < .05) reported significantly lower levels of emotional exhaustion. Regard- ing education, college graduates were no different than their non- college educated counterparts on emotional exhaustion, but Master’s-level counselors reported significantly higher levels of emotional exhaustion than the noncollege reference group (b ¼ .115, p < .001). Measures of recovery status, certification status, and the study sample (public vs. private sector) were not significant predictors of exhaustion. Together, the organizational, demographic, and coworker support measures explained 25.9% of the variance in counselor’s emotional exhaustion scores.

Table 3. Standardized structural model estimates for emotional exhaustion

and turnover intention of addiction treatment counselors (N ¼ 1,869)

Emotional Exhaustion Intent to Quit

Exhaustion — .361��

Coworker Support �.138�� �.118��

Autonomy �.170�� �.075�

Distributive Justice �.286�� �.212��

Procedural Justice .015 �.174��

Weekly Work Hours (log) .047 .011

% Time in Direct Care �.029 �.011

Medium Caseloada .082 .041

Large Caseloada .035 .093

Gender (Female) .046 �.073��

Nonwhite �.050� .032

Age �.062�� �.077��

College Degreeb �.046 �.063�

Master’s Degreeb .115�� .113��

Salary .043 �.072��

Tenure �.006 �.119��

Recovering �.032 �.045�

Certified �.017 �.003

Public Center �.014 .031

Latent Variable R2 .259 .596

Chi-Square ¼ 2983.29, df ¼ 985, p < .001 aReference category ¼ small caseload (fewer than ten patients). bReference category ¼ less than college degree. �p < .05, ��p < .01.

96 L. J. Ducharme et al.

Turning to intention to quit, all four work environment measures were directly associated with turnover intention, net of work stres- sors, sociodemographic characteristics, and exhaustion. The signifi- cant negative association between coworker support and intention to quit (b ¼ � .118, p < .001) was such that counselors reporting greater support from coworkers were significantly less likely to be intending to seek employment elsewhere. Job autonomy was also negatively associated with turnover intention (b ¼ � .075, p < .001). Both justice-related measures were negatively associated with counse- lors’ intentions to quit. Greater procedural justice (b ¼ � .174, p < .001) and greater distributive justice (b ¼ � .212, p < .001) were both directly associated with lower levels of turnover intention. As before, none of the workload measures was directly associated with turnover intention, although having the largest caseloads approached significance (p < .10).

Several of the sociodemographic measures were directly associated with intent to quit. Female counselors reported lower turnover inten- tion than male counselors (b ¼ � .073, p < .001). Counselors’ tenure at the center (b ¼ � .119, p < .001) and age (b ¼ � .077, p < .001) were both negatively associated with turnover intention. Counselors with higher annual earnings were significantly less likely to be intend- ing to quit their jobs (b ¼ � .047, p < .05). Education revealed a non- linear effect. Counselors holding a college degree were significantly less likely than noncollege-educated counselors to be contemplating quitting (b ¼ � .063, p < .01), whereas Master’s level counselors reported significantly greater turnover intention than counselors without a college degree (b ¼ .113, p < .001). Counselors in recovery from their own addiction reported significantly lower turnover inten- tion (b ¼ � .045, p < .05), while neither certification status nor the sample from which the treatment centers were drawn significantly predicted intention to quit.

As expected, emotional exhaustion was significantly and posi- tively associated with turnover intention. The magnitude of this association (b ¼ .361, p < .001) was considerably larger than that of the other independent variables. The model of turnover intention also hypothesized indirect effects of work on intention to quit via exhaustion. Indirect effects were calculated for the work-related measures that were significantly associated with exhaustion. Specifi- cally, the indirect effects of coworker support, job autonomy, and distributive justice on turnover intention were � .050, � .061, and � .103, respectively. The addition of the indirect effects to the direct effects of these measures yield total effects of � .167 for coworker support, � .136 for job autonomy, � .315 for distributive justice,

Exhaustion and Turnover Intention in Human Service 97

and � .169 for procedural justice on turnover intention (all signifi- cant at p < .01).

The integration of coworker support, job characteristics, and sociodemographics along with emotional exhaustion resulted in a model that explained a large amount of the variance in counselors’ intentions to quit. This model of direct and indirect effects explained 59.6% of the variance in this latent variable measuring turnover intention.

DISCUSSION

These models suggest that the receipt of instrumental and informa- tional support from coworkers has significant implications for sub- stance abuse treatment counselors’ well-being and retention. While workload had no measurable effect net of the other variables mea- sured, characteristics of the work environment significantly affected both outcomes, and coworker support was significantly and inversely related to both exhaustion and turnover intention, net of the effects of these structural variables. Coworker support reduced intent to quit both directly and indirectly via exhaustion. The persistent effects of social support in these models suggests the important role that social relationships play in protecting counselors’ own mental health, and in discouraging turnover. Moreover, because clients are at a distinct dis- advantage when their treatment is interrupted by counselor turnover, the significant effects of coworker support have nontrivial down- stream implications for client recovery and quality of care.

These analyses examined a specific sector of the service industry, namely behavioral health counseling occupations. Unlike work in much of the broader service sector (characterized by discrete transac- tions and surface acting), counseling occupations are characterized by the development and maintenance of meaningful relationships with clients in the trust-based context of the therapeutic alliance. As such, counselors make substantial investments in these relationships. The nature of addiction treatment (the industry from which this study’s counselors were drawn) is such that clients are prone to high rates of relapse, which requires counselors to repeatedly ‘‘start over’’ in establishing the basis for sustained sobriety and progress towards other treatment goals. We argue that this emotional investment in the therapeutic relationship is a source of job stress that far exceeds the structural properties of counseling occupations in their adverse effects, as manifest in emotional exhaustion. Indeed, our findings show that structural variables (number of clients, hours worked, percent of time in direct care) are far less predictive of emotional

98 L. J. Ducharme et al.

exhaustion and turnover intention relative to the interpersonal aspects of the work environment. Counselors working in settings in which the established pattern of interaction provides a sense of auto- nomy, fairness, and interpersonal support are less likely to express symptoms of emotional exhaustion, and are less likely to desire to quit their jobs. The interpersonal relationships characterizing the work environment—the milieu within which therapeutic alliances are built—are highly predictive of the well-being and stability of those who engage in counseling occupations.

These models were tested in a large sample of counselors that are representative of the clinical staff in public and private-sector addic- tion treatment programs; however, this study has several limitations that should be noted. First, all data were derived from counselor self- report items on questionnaires. Because counselors reported on working conditions, coworker relations, and their exhaustion and turnover intentions, common-method variance issues should be con- sidered. Similarly, the data are cross-sectional and, therefore, the cau- sal relationships between the variables specified here cannot be determined. Studies that can undertake longitudinal designs and incorporate objective measures of working conditions at the unit or organizational level would provide a more robust test of this model.

In addition, this dataset offers no measures of client caseload com- position at the counselor level. Other studies have found that counse- lors with proportionately more ‘‘difficult’’ clients—that is, those with more health issues (such as clients with HIV=AIDS) or greater service resource needs (such as those with cooccurring mental illness)— experience burnout and turnover more frequently than other counse- lors (Acker 1999; Shoptaw et al. 2000). Had caseload data been avail- able, we might have found that counselors with more demanding clients experienced significantly higher levels of exhaustion and=or turnover intention. At the same time, however, it is difficult to pre- sume whether instrumental social support would have played any less an important role in predicting exhaustion or intent to quit in the presence of such data.

Finally, while we were concerned with specifying the role of coworker support, the nature of the addiction counseling field sug- gests a need to examine the potentially important role of supervis- ory support. Clinical supervision in this field extends beyond the more purely administrative or management role performed by supervisors in other settings and, therefore, may have specific impli- cations for burnout and turnover. Clinical supervision is critical to the socialization of new counselors, and is essential in ensuring the fidelity of implementation of various psychosocial therapies (Powell

Exhaustion and Turnover Intention in Human Service 99

1993). Clinical supervision should assist counselors in learning how to handle challenging clients, and in that manner buffer counselors against burnout (Culbreth 1999; Kennard et al. 1987). Given this unique role, future studies on counseling occupations should seek to specify the role of clinical supervisors in counselors’ work experi- ence and whether this type of support yields additive effects beyond coworker support for counselors’ job satisfaction, retention, and well-being.

While addiction and mental health counseling represents a speci- alty field within the much broader healthcare and human services are- nas, these data, nevertheless, provide a useful lens through which to view the interplay of job demands, work environment, social support, and job-related affect. On the one hand, the availability of nationally- representative data from a specific human service occupation allows for the identification of variables that are key predictors of job reten- tion and well-being. On the other hand, replication of such models in other settings is warranted to identify whether the role of job demands and coworker support are unique in their combined influ- ence on this segment of the healthcare delivery system, and what if any differences are seen across occupational categories and industries in the service sector.

REFERENCES

Acker, G. M. 1999. ‘‘The Impact of Clients’ Mental Illness on Social Workers’ Job

Satisfaction and Burnout.’’ Health & Social Work 24:112–120.

Alexander J. A., R. Lichtenstein, H. J. Oh, and E. Ullman.1998. ‘‘A Causal Model

of Voluntary Turnover Among Nursing Personnel in Long-term Psychiatric

Settings.’’ Research in Nursing and Health 21:415–427.

Arches, J. 1991. ‘‘Social Structure, Burnout, and Job Satisfaction.’’ Social Work

36:193–172.

Bennett L., M. Kelaher, and M. Ross. 1994. ‘‘Quality of Life in Health Care Profes-

sionals: Burnout and Its Associated Factors in HIV=AIDS-related Care.’’

Psychology & Health 9:273–283.

Blankertz, L. E. and S. E. Robinson. 1997. ‘‘Turnover Intentions of Community

Mental Health Workers in Psychosocial Rehabilitation Services.’’ Community

Mental Health Journal 33:517–529.

Bloom, J. R., J. A. Alexander, and B. A. Nuchols. 1992. ‘‘The Effect of the Social

Organization of Work on the Voluntary Turnover Rate of Hospital Nurses in

the U.S.’’ Social Science and Medicine 34:1413–1424.

Bowden, G. 1994. ‘‘Work Stress, Burnout, and Coping: A Review and an Empirical

Study of Staff in Supported Housing.’’ Clinical Psychology and Psychotherapy

1:219–232.

100 L. J. Ducharme et al.

Brown, N. C., B. J. Prashanthan, and M. Abbott. 2003. ‘‘Personality, Social Sup-

port, and Burnout Among Human Service Professionals in India.’’ Journal of

Community & Applied Social Psychology 13:320–324.

Budrionis, M. M., W. L. Cunningham, and K. A. Pratt. 1987. ‘‘Factors Related to

Burnout Among Alcoholism Counselors.’’ Alcohol Health and Research World

11:82–85.

Bulan, H. F., R. Erickson, and A. Wharton. 1997. ‘‘Doing for Others on the Job:

The Affective Requirements of Service Work, Gender, and Emotional Well-

being.’’ Social Problems 44:235–256.

Bureau of Labor Statistics. 2004. Job Openings and Labor Turnover Survey (JOLTS),

Summary Data for 2003. Retrieved December 14, 2004 from http://www.bls.

gov/jlt.

Cherniss, C. 1980. Professional Burnout in Human Service Organizations. New York:

Praeger.

Connors, G. J., K. M. Carroll, C. C. DiClemente, R. Longabaugh, and D. M. Donovan.

1997. ‘‘The Therapeutic Alliance and Its Relationship to Alcoholism Treatment

Participation and Outcome.’’ Journal of Consulting and Clinical Psychology

65:588–598.

Culbreth, J. R. 1999. ‘‘Clinical Supervision of Substance Abuse Counselors: Cur-

rent and Preferred Practices.’’ Journal of Addictions and Offender Counseling

20:15–25.

Drake, B. and G. N. Yadama. 1996. ‘‘A Structural Equation Model of Burnout and

Job Exit Among Child Protective Service Workers.’’ Social Work Research

20:179–187.

Elman, B. D. and E. T. Dowd. 1997. ‘‘Correlates of Burnout in Inpatient Sub-

stance Abuse Treatment Therapists.’’ Journal of Addictions & Offender Counseling

17:56–66.

Gallon, S. L., R. M. Gabriel, and J. R. W. Knudsen. 2003. ‘‘The Toughest Job You’ll

Ever Love: A Pacific Northwest Treatment Workforce Survey.’’ Journal

of Substance Abuse 24:183–196.

Garman, A. N., P. W. Corrigan, and S. Morris. 2002. ‘‘Staff Burnout and Patient

Satisfaction: Evidence of Relationships at the Care Unit Level.’’ Journal of Occu-

pational Health Psychology 7:235–241.

Geurts, S., W. Schaufeli, and D. DeJonge. 1998. ‘‘Burnout and Intention to Leave

Among Mental Health Care Professionals: A Social Psychological Approach.’’

Journal of Social and Clinical Psychology 17:341–362.

Glass, D. C. and J. D. McKnight. 1996. ‘‘Perceived Control, Depressive Symptoma-

tology, and Professional Burnout: A Review of the Evidence.’’ Psychology &

Health 11:23–48.

Gray, A. M. and V. I. Phillips. 1994. ‘‘Turnover, Age, and Length of Service: A Com-

parison of Nurses and Other Staff in the National Health Service.’’ Journal of

Advanced Nursing 19:819–827.

Griffeth, R. W., P. W. Hom, and S. Gaertner. 2000. ‘‘A Meta-analysis of Antece-

dents and Correlates of Employee Turnover: Update, Moderator Tests, and

Research Implications for the Next Millennium.’’ Journal of Management

26:463–488.

Exhaustion and Turnover Intention in Human Service 101

Hendrix, W. H., T. Robins, J. Miller, and T. P. Summers. 1999. ‘‘Effects of

Procedural Justice and Distributive Justice on Factors Predictive of Turnover.’’

Journal of Social Behavior and Personality 13:611–632.

Himle, D. P., S. Jayaratne, and P. Thyness. 1989. ‘‘The Effects of Emotional Support

on Burnout, Work Stress and Mental Health Among Norwegian and American

Social Workers.’’ Journal of Social Service Research 12:27–45.

Hochschild, A. R. 1983. The Managed Heart: The Commercialization of Human Feel-

ing. Berkeley: University of California Press.

Hser, Y. I. 1995. ‘‘Drug Treatment Counselor Practices and Effectiveness: An

Examination of Literature and Relevant Issues in a Multilevel Framework.’’

Evaluation Review 19:389–408.

Hu, L. T. and P. M. Bentler. 1999. ‘‘Cutoff Criteria for Fit Indices in Covariance

Structure Analysis: Conventional Criteria Versus New Alternatives.’’ Structural

Equation Modeling 6:1–55.

Jackson, S. E., R. L. Schwab, and R. S. Schuler. 1986. ‘‘Toward an Understanding of

the Burnout Phenomenon.’’ Journal of Applied Psychology 71:630–640.

Jamal, M. 1990. ‘‘Relationship of Job Stress and Type-a Behavior to Employees’ Job

Satisfaction, Organizational Commitment, Psychosomatic Health Problems, and

Turnover Motivation.’’ Human Relations 43:727–738.

Janssen, P., J. DeJonge, and A. B. Bakker. 1999. ‘‘Specific Determinants of Intrinsic

Work Motivation, Burnout, and Turnover Intentions: A Study Among Nurses.’’

Journal of Advanced Nursing 29:1360–1369.

Jayaratne, S. and W. A. Chess. 1984. ‘‘The Effects of Emotional Support on

Perceived Stress and Strain.’’ Journal of Applied Behavioral Science 20:141–153.

Jayaratne, S., W. A. Chess, and D. A. Kunkel. 1986. ‘‘Burnout: Its Impact on Child

Welfare Workers and Their Spouses.’’ Social Work 31:53–59.

Jinnett, K. and J. A. Alexander. 1999. ‘‘The Influence of Organizational Context on

Quitting Intention: An Examination of Treatment Staff in Long-term Mental

Health Care Settings.’’ Research on Aging 21:176–204.

Johnson, P .J. and A. Rubin. 1983. ‘‘Case Management in Mental Health: A Social

Work Domain?’’ Social Work 28:49–55.

Karasek, R. 1979. ‘‘Job Demands, Job Decision Latitude, and Mental Strain:

Implications for Job Redesign.’’ Administrative Science Quarterly 24:285–308.

Kennard, B. D., S. M. Stewart, and M. R. Cluck. 1987. ‘‘The Supervision Relation-

ship: Variables Contributing to Positive Versus Negative Experiences.’’

Professional Psychology: Research and Practice 18:172–175.

Knudsen, H. K., L. J. Ducharme, and P. M. Roman. 2006. ‘‘Counselor Emotional

Exhaustion and Turnover Intention in Therapeutic Communities.’’ Journal of

Substance Abuse Treatment 31:173–180.

Knudsen, H. K., J. A. Johnson, and P. M. Roman. 2003. ‘‘Retaining Counseling

Staff at Substance Abuse Treatment Centers: Effects of Management Practices.’’

Journal of Substance Abuse Treatment 24:129–135.

Koeske, G. F. and R. D. Koeske. 1989. ‘‘Workload and Burnout: Can Social Support

and Perceived Accomplishment Help?’’ Social Work 34:243–244.

Laundergan, J. C., D. Flynn, and J. D. Gaboury. 1986. ‘‘An Alcohol and Drug

Counselor Training Program: Hazelden Foundation’s Trainee Characteristics

and Outcomes.’’ Journal of Drug Education 16:167–179.

102 L. J. Ducharme et al.

Layne, C. M., T. H. Hohenshil, and K. Singh. 2004. ‘‘The Relationship of

Occupational Stress, Psychological Strain, and Coping Resources to Turnover

Intentions of Rehabilitation Counselors.’’ Rehabilitation Counseling Bulletin

48:19–30.

LeCroy, C. W. and M. R. Rank. 1987. ‘‘Factors Associated with Burnout in the

Social Services: An Exploratory Study.’’ Journal of Social Service Research

10:23–39.

Lee, R. T. and B. E. Ashforth. 1993. ‘‘A Further Examination of Managerial

Burnout: Toward an Integrated Model. Journal of Organizational Behavior

14:3–20.

Lichtenstein, R., J. A. Alexander, J. F. McCarthy, and R. Wells. 2004. ‘‘Status Dif-

ferences in Cross-functional Teams: Effects on Individual Member Participation,

Job Satisfaction, and Intent to Quit. Journal of Health and Social Behavior

45:322–335.

Lum, L., J. Kervin, K. Clark, F. Reid, and W. Sirola. 1998. ‘‘Explaining Nursing

Turnover Intent: Job Satisfaction, Pay Satisfaction, or Organizational Commit-

ment?’’ Journal of Organizational Behavior 19:305–320.

Maslach, C. 1978. ‘‘The Client Role in Staff Burn-out.’’ Journal of Social Issues

34:111–124.

Maslach, C. and S. E. Jackson. 1981. ‘‘The Measurement of Experienced Burnout.’’

Journal of Occupational Behavior 2:99–113.

———. Maslach Burnout Inventory Manual, 2nd edition. Palo Alto, CA: Consulting

Psychologists Press.

McLellan, A. T., D. Carise, and H. D. Kleber. 2003. ‘‘Can the National Addiction

Treatment Infrastructure Support the Public’s Demand for Quality Care?’’ Jour-

nal of Substance Abuse Treatment 25:117–121.

Morris, J. A. and D. C. Feldman. 1996. ‘‘The Dimensions, Antecedents, and

Consequences of Emotional Labor.’’ Academy of Management Review 21:

986–1010.

Mulvey, K. P., S. Hubbard, and S. Hayashi. 2003. ‘‘A National Study of the Sub-

stance Abuse Treatment Workforce.’’ Journal of Substance Abuse Treatment

24:51–57.

Munn, E. K., C. E. Barber, and J. J. Fritz. 1996. ‘‘Factors Affecting the Professional

Well-being of Child-life Specialists.’’ Children’s Health Care 25:71–91.

Muthen, L. K. and B. O. Muthen. 2004. Mplus User’s Guide, 3rd edition. Los

Angeles: Muthen & Muthen.

Niehoff, B. P. and R. H. Moorman. 1993. ‘‘Justice as a Mediator of the Relationship

Between Methods of Monitoring and Organizational Citizenship Behavior.’’

Academy of Management Journal 36:527–556.

Niehoff, M. S. 1984. ‘‘Burnout and Alcoholic Treatment Counselors.’’ Counseling

and Values 29:67–69.

Pines, A. and C. Maslach. 1978. ‘‘Characteristics of staff burnout in mental health

settings.’’ Hospital and Community Psychiatry 29:233–237.

Poulin, J. and C. Walter. 1993. ‘‘Social Workers Burnout: A Longitudinal Study.’’

Social Work Research & Abstracts 29:5–11.

Powell, D. J. 1993. Clinical Supervision in Alcohol and Drug Abuse Counseling.

New York: Lexington Books.

Exhaustion and Turnover Intention in Human Service 103

Powell, M. J. and R. O. York. 1992. ‘‘Turnover in County Public Welfare Agencies.’’

Journal of Applied Social Sciences 16:111–127.

Pritchard, R. D. and B. W. Karasek. 1973. ‘‘The Effects of Organizational Climate

on Managerial Job Performance and Job Satisfaction.’’ Organizational Behavior

and Human Performance 9:126–146.

Rubin, A. 1984. ‘‘Community-based Care of the Mentally Ill: A Research Review.’’

Health & Social Work 9:165–177.

Schaefer, J. A. and R. H. Moos. 1996. ‘‘Effects of Work Stressors and Work Climate

on Long-term Care Staff ’s Job Morale and Functioning.’’ Research in Nursing

and Health 19:63–73.

Schaufeli, W. B. and D. Enzmann. 1998. The Burnout Companion to Study and

Practice: A Critical Analysis. London: Taylor & Francis.

Shinn, M., M. Rosario, H. Morch, and D. E. Chestnut. 1984. ‘‘Coping with Job

Stress and Burnout in the Human Services.’’ Journal of Personality and Social

Psychology 46:864–876.

Shoptaw, S., J. A. Stein, and R. A. Rawson. 2000. ‘‘Burnout in Substance Abuse

Counselors: Impact of Environment, Attitudes, and Clients with HIV.’’ Journal

of Substance Abuse Treatment 19:117–126.

Siefert, K., S. Jayaratne, and W. A. Chess. 1991. ‘‘Job Satisfaction, Burnout, and

Turnover in Health Care Social Workers.’’ Health and Social Work 16:193–202.

Somers, M. 1995. ‘‘Organizational Commitment, Turnover, and Absenteeism:

An Examination of Direct and Interaction Effects.’’ Journal of Organizational

Behavior 16:49–58.

Tang, H. K. 1999. ‘‘An Inventory of Organizational Innovativeness.’’ Technovation

19:41–51.

Vahey, D. C., L. H. Aiken, D. M. Sloane, S. P. Clarke, and D. Vargas. 2004. ‘‘Nurse

Burnout and Patient Satisfaction.’’ Medical Care 42(Suppl. 2):57–66.

Vinokur-Kaplan, D., S. Jayaratne, and W. A. Chess. 1994. ‘‘Job Satisfaction and

Retention of Social Workers in Public Agencies, Nonprofit Agencies, and Private

Practice: The Impact of Workplace Conditions and Motivators.’’ Administration

in Social Work 18:93–121.

Walsh, J. P., S. J. Ashford, and T. E. Hill. 1985. ‘‘Feedback Obstruction: The

Influence of Information Environment on Employee Turnover Intentions.’’

Human Relations 38:23–46.

Wharton, A. S. 1999. ‘‘The Psychosocial Consequences of Emotional Labor.’’

Annals of the American Academy of Political and Social Science 561:158–176.

104 L. J. Ducharme et al.