The Relationship of Affective and Continuance Organizational Commitment with Correctional Staff Occupational Burnout

Occupational Burnout

Burnout is commonly mistaken as a form of stress, but although they are related, burnout and stress are different (Belcastro, Gold, & Grant, 1982). Stress has been defined in the correctional literature as “the psychological discomfort or tension which results from exposure to stressors” (Cullen, Link, Wolfe, & Frank, 1985, p. 507). Unlike burnout, which typically has negative consequences, not all stress is counterproductive. Moderate levels of stress can motivate workers and raise productivity levels by making employees more alert and focused (Hurst & Hurst, 1997). Burnout is an unwanted by-product of unmanaged stress that impedes worker productivity and organizational functioning (Garland, 2002; Morris, 1986).

Burnout has been variously defined as a person becoming exhausted, becoming worn out, or failing to cope with excessive work demands (Freudenberger, 1974), and also as a three-stage process involving stress, strain, and defensive coping (Cherniss, 1980). The most popular and empirically investigated definition comes from Maslach (1978, 1981, 1982), who depicted burnout essentially as a diminished capacity for professional caring resulting from overwhelming client interaction among human service professionals. According to Maslach and Jackson (1981, 1984), the burnout syndrome included three dimensions: emotional exhaustion, depersonalization, and feelings of reduced accomplishment. Emotional exhaustion refers to emotional drain and fatigue stemming from occupational life. Depersonalization is defined as treating clients callously and impersonally. Reduced feelings of reduced accomplishment refer to heightened feelings of ineffectiveness at work.

Burnout can cause physical health problems such as head, neck, and back pain, stomach ailments, insomnia, general fatigue, and delay in illness recovery (Belcastro et al., 1982; Elman & Dowd, 1997; Maslach, 1982; Pines, Aronson, & Kafry, 1981). Psychological problems can also arise from burnout, such as increased feelings of anger, depression, and helplessness and decreased sympathy, motivation, and self-esteem (Belcastro et al., 1982; Kamis, 1981; Maslach, 1982; Pines & Maslach, 1978). Burnout has an impact on behavior by increasing the likelihood of staff withdrawal (Maslach, 1982), oppositional posturing (Freudenberger, 1974), unnecessary risk-taking (Maslach, 1982), negative attitudes toward offender rehabilitation (Corrigan et al., 1998), and alcohol and drug abuse (Maslach, 1981).

One undesirable consequence of difficult and stressful work environments is staff burnout, and, unfortunately, the risk of burnout is greater in correctional professions than many others (Keinan & Maslach-Pines, 2007). Burnout has a negative impact on the employee, family and friends, coworkers, inmates, the organization, and even society. Burnout harms the psychological and physical health of the burned-out person (Maslach, Schaufeli, & Leiter, 2001; Neveu, 2007), and has been linked with increased probabilities of substance abuse, poorer family interactions, and withdrawal from family and friends (Garland, 2002; Maslach, 1982). Burnout can negatively affect interactions with coworkers and inmates. Burned-out correctional staff often may withdraw, leading to a strained work environment where burned-out staff treat inmates and coworkers in a callous manner (Garner, Knight, & Simpson, 2007; Whitehead, 1989). As the work performance of the burned-out person drops (Neveu, 2007), it affects coworkers by burdening them with extra work (Dignam, Barrera, & West, 1986; Garner et al., 2007) and by contagiously spreading through the organization (Maslach et al., 2001). Burnout can lead to increased absenteeism and turnover as escape mechanisms (Carlson & Thomas, 2006; Neveu, 2007). In an era where rising costs, shrinking budgets, and personnel shortages are common, ensuring that correctional staff do not suffer burnout is increasingly important.

Affective and Continuance Commitments

Mowday, Porter, and Steers (1982) defined organizational commitment as “the relative strength of an individual’s identification with and involvement in a particular organization” (p. 27). Organizational commitment reflects one’s bond to an entire organization, not just a specific department or job (Allen & Meyer, 1990; Garland, McCarty, & Zhao, 2009). According to Lincoln and Kalleberg (1990), committed workers develop moral stakes in an organization that transcend attachments formed through pay and intrinsic rewards. Benefits from high levels of organizational commitment in corrections include increased prosocial behavior at work and lower levels of absenteeism and voluntary turnover (Camp, 1994; Culliver, Sigler, & McNeely, 1991; Stohr, Self, & Lovrich, 1992). Affective and continuance commitments are placed at opposite ends of an attitudinal-behavior continuum. Affective commitment lies at the attitudinal end, whereas continuance commitment rests at the behavioral end.

Affective commitment is a psychological attachment formed through feelings of loyalty and pride toward an organization (Jaunch, Glueck, & Osborn, 1978; O’Reilly & Chatman, 1986). This type of commitment requires a person to accept and internalize the core goals and values of his or her employing organization (Mowday et al., 1982). According to resource exchange theory, affective commitment is a conscious decision influenced by the organization’s treatment of the employee (McElroy, Morrow, & Wardlow, 1999; Thibault & Kelley, 1959). Resource exchange theory suggests that stronger bonds of affective commitment are likely to form when organizations demonstrate greater respect and concern for employees and their work needs. Weaker affective bonds will develop when employees are treated negatively and their needs and contributions are underappreciated (McElroy et al., 1999).

Continuance commitment is generated from making physical, cognitive, and emotional investments in an organization. This type of commitment is rooted in Becker’s (1960) side-bet theory. Becker’s theory argued that the cumulative effect of sunken costs in an organization, such as receiving high pay due to a long tenure or ascending to a coveted supervisory position through seniority, makes leaving the organization increasingly difficult. In other words, the costs of leaving the organization are high considering the energy and time spent in the organization and the benefits received from those investments (Hrebiniak & Alutto, 1972; Jaros, Jermier, Koehler, & Sincich, 1993). The employee then commits to an organization to avoid losing the benefits gained from his or her investments.

One clear omission in the correctional literature is the absence of research examining the influence of organizational commitment on Maslach’s three dimensions of burnout. If burnout is identified as a result of affective and continuance commitments, many new avenues can be explored for reducing burnout, given that an expanding body of literature is gradually illustrating how organizational commitment can benefit prisons, or, conversely how a lack of organizational commitment can impair them (Lambert, Hogan, & Jiang, 2008; Moore & Sales, 2005). Widespread burnout can produce a potentially devastating impact on a prison organization and its staff, making the study of commitment’s impact on burnout all the more important.

High levels of affective commitment may be able to reduce the degree of burnout experienced by prison staff, whereas high levels of continuance commitment may facilitate the burnout process. Resource exchange theory maintains that employees will develop greater affective commitment to the extent they feel valued and appreciated by the organization, and identify with the organizational goals (McElroy et al., 1999; Thibault & Kelley, 1959). This type of commitment is likely to translate into a more positive outlook toward one’s job as a whole and the occupational responsibilities and duties performed. Workers in such a state of mind should be more apt to keep the demands and challenges of prison work in perspective with broader professional aspirations and organizational objectives, making them less likely to succumb to exhaustion, to depersonalize offenders, and to feel diminished occupational worth.

In contrast, continuance commitment results from the cumulative effect of making investments which offer some kind of personal benefit, which, in turn, strongly influences a worker’s decision to remain with an organization even if he or she does not like working for the organization (Becker, 1960; Jaros et al., 1993). A potential downside of continuance commitment is that an overconcentration on the benefits gained from working in corrections may overshadow the identification of intrinsic meaning and satisfaction from the work itself. Over time, a limited appreciation of the value of correctional work and one’s individual contributions can produce a wearing down effect and simultaneously diminish a sense of caring for offenders (Cherniss, 1980).

For this study, six hypotheses were proposed.

Method

Participants

All the available staff at a Midwestern private prison were provided a survey packet. ¹ The correctional facility was a maximum security level closed facility administered by a private company contracted by the state government. At the time of the survey, the prison housed approximately 450 offenders aged 14 to 19, who had been sentenced as adult offenders after being transferred from juvenile to adult court. Once an offender reached the age of 20, he was transferred to the state’s adult prison system. All the offenders were serving long sentences for violent crimes. Nevertheless, one of the primary goals of the private prison was to provide treatment and education for the inmates. The other primary goals of prison were to provide a safe, humane, and secure environment for inmates and staff. While it was a private prison, the staff were required to complete: 1) the same new employee training academy; and 2) the same annual refresher training that was required of the staff in the public prison system in the state. The private prison was located in a rural area of the state, and the staff, on average, were paid a higher wage than many of the employees in this area. At the time of the survey, the private prison had been in operation for less than 5 years, and, as such, the tenure of the employees was low.

Two hundred twenty staff were employed at the prison. Some staff were absent from the institution for various reasons (e.g., vacation leave, sick leave, training, administrative leave, and so forth) during the week that survey packets were distributed. As such, only 200 staff members received the survey packet. This packet contained a cover letter, the survey instrument, a stamped return envelope, and a bifurcated raffle ticket. The cover letter explained the nature of the study, that participation was completely voluntary, that no responses would be seen by other prison staff (i.e., results would be confidential), and, as no identification information was being requested or to be provided, that all responses would be anonymous. The cover letter explained that a raffle would be held and completion of the survey was not a requirement to be part of the raffle. The instructions indicated that the respondent should retain one half of the raffle ticket and return the other half with the survey, regardless of whether the survey was completed or not, in the stamped, pre-addressed envelope to the research team or to a locked box in main prison area (only the research team had a key for the lock). The returned tickets were removed and separated from the surveys, so that it would be impossible to link surveys with specific staff. About a month after the survey packets were distributed and returned, a drawing of raffle tickets was held at a staff event, and individuals with a winning raffle ticket were awarded a cash prize ranging from US$50 to US$150. Unclaimed prizes were donated to the private prison’s staff association. A total of 160 usable surveys were returned, ² representing a response rate of 80% based upon the 200 staff members who were provided the survey packet. Except upper administration, all staff employed at the prison were surveyed. Overall, as indicated in the note to Table 1, the participants appeared to be demographically representative of the staff at the private prison.

OPEN IN VIEWER

Table 1:

Descriptive Statistics

Variable	Description	Minimum	Maximum	Med	M	SD
Position	38% non-CO, 62% CO	0	1	1	0.62	0.49
Supervisory status	79% non-supervisor, 21% supervisor of other staff	0	1	0	0.21	0.41
Gender	41% female, 59% male	0	1	1	0.59	0.49
Age	Measured in continuous years	19	68	33	35.77	10.82
Tenure	Measured in continuous months employed at the private prison	1	53	17	20.64	13.84
Educational level	53% had no college degree, 47% had a college degree	0	1	0	0.47	0.50
Race	21% non-White, 79% White	0	1	1	0.79	0.40
Average daily inmate contact	Less than an hour (17%), 1 to 2 hr (6%), 3 to 4 hr (9%), 5 to 6 hr (19%), 7 or more hours (49%)	1	5	4	3.78	1.51
Affective commitment	Six-item additive index, α = .88	10	26	19	18.14	4.00
Continuance commitment	Four-item additive index, α = .77	4	20	12	11.84	3.88
Burnout-emotional exhaustion	Three-item additive index, α = .85	3	14	6	7.08	2.40
Burnout-depersonalization	Four-item additive index, α = .74	4	19	9	9.41	2.93
Burnout-reduced accomplishment	Six-item additive index, α = .72	7	23	13	13.20	3.10

Note. The total number of participants with completed surveys was 160. In terms of position, 62% were correctional officers (coded 1), 9% were custody supervisors, 6% worked in the business office, 4% worked in education, 3% were unit management staff, 3% worked in the medical, and 13% worked in other areas (all no CO positions coded 0). In terms of supervisory status, 21% indicated that they were supervisors of other staff at the prison (coded 1) or was not (coded 0). Gender was coded as women = 0 and men = 1. In terms of highest educational level, 6% indicated that they had a high school degree or GED, 47% had some college but no degree, 24% had an associate’s degree, 16% had a bachelor’s degree, and 7% had a graduate or professional degree. Educational level was measured as a variable representing whether the respondent had earned a college degree (coded as 1) or not (coded as 0). In terms of race/ethnicity, 79% marked they were White (coded 1), 11% Black, 2% Hispanic, 3% Native American, and 4% other (non-White coded 0). Institutional records indicated that, at the time of the survey, approximately 81% of the all employees at the prison were White and 61% were male. In addition, the median age range of all staff at the facility about 34, the average tenure was about 20 months, and about 60% held a custody position. No educational information about the entire staff complement could be obtained. GED = general education development diploma; Med = median value; CO = correctional officer; α = Cronbach’s alpha value.

Results

The descriptive statistics for the variables are presented in Table 1. The median tenure at the private prison was low because the prison had been in operation for only 5 years prior at the time of the survey. The mean scores for the organizational commitment and burnout scales indicated that the average respondent was in the middle range of the scale; however, these composite scales are not exact replicas of normed scales, so no judgment can be made regarding the commitment and burnout of respondents in comparison with other occupational populations. There appeared to be significant variation in both the dependent and independent variables (i.e., none was a constant). Based upon the skewness and kurtosis statistics, no problem existed with the shape or peak of the variable distributions. Cronbach’s alpha is a measure of internal reliability for variables representing latent concepts using multiple items. It ranges from 0 to 1.00, and higher numbers indicate higher internal reliability of the measure. The indexes’ Cronbach’s alphas were greater than .70, which indicates good internal reliability (Neal, 1996).

Pearson’s correlation coefficients are presented in Table 2. The bivariate results supported each of the hypotheses—affective commitment had a negative relationship with each of the dimensions of burnout and continuance commitment had a positive relationship with each—but the purposes of the bivariate analysis were primarily to ensure that the data did not have too much covariability for the regression models and to allow for future meta-analysis of the data.

OPEN IN VIEWER

Table 2:

Correlation Matrix for Study Variables

Variables	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Position	1.00
2. Supervisory	−.47**	1.00
3. Gender	.28**	.12	1.00
4. Age	−.53**	.16*	−.17*	1.00
5. Tenure	−.20**	.39**	.00	.07	1.00
6. Educ level	−.29**	.19*	−.13	.16*	.12	1.00
7. Race	−.02	.04	.08	−.05	.06	.08	1.00
8. Avg contact	.34**	−.19*	.06	−.09	−.01	−.24**	.00	1.00
9. Affect comm	−.24**	.23**	.05	.17*	−.04	.07	.00	−.01	1.00
10. Cont comm	.16*	−.10	−.06	−.15	.12	−.01	−.08	.04	−.39**	1.00
11. Emotional	.09	−.16*	−.05	−.06	.16	.01	−.03	−.01	−.43**	.49**	1.00
12. Deperson	.32**	−.10	.08	−.32**	.10	−.04	−.03	.10	−.40**	.37**	.46**	1.00
13. Reduce accom	.18*	−.24**	−.12	−.08	−.02	−.10	.00	−.14	−.30**	.27**	.38**	.44**	1.00

Note. See Table 1 for a description of the variables and how they were coded. Supervisory = supervisory status; educ level = educational level; avg contact = average daily inmate contact; affect comm = affective commitment; cont comm = continuance commitment; emotional = the burnout dimension of emotional exhaustion; deperson = the burnout dimension of depersonalization; and reduce accom = the burnout dimension of reduced feelings of accomplishment. The number of total participants in the bivariate analysis was 160.

*

p ≤ .05. **p ≤ .01.

The final component of the analytic strategy involved ordinary least squares (OLS) regression analyses. Three OLS regression equations were estimated with emotional exhaustion, depersonalization, and feelings of reduced accomplishment as the dependent variables. The independent variables in the three OLS regression equations were affective commitment, continuance commitment, position, supervisory status, gender, age, tenure, educational level, race, and average daily inmate contact.

Based upon the correlation matrix presented in Table 2, collinearity (i.e., a high overlap between variables represented by a correlation higher than .7) was not a problem among the independent variables (Tabachnick & Fidell, 2013). The variance inflation factor and tolerance statistics indicated no problem of multicollinearity among the independent variables. High multicollinearity occurs when an independent variable shares a very large part of its variance with the other independent variables in the regression equation, and this hinders the estimation of the effects on the dependent variable by the independent variables in question (Tabachnick & Fidell, 2013). For the emotional exhaustion equation, the variance inflation factor scores ranged from 1.03 to 2.28 and tolerance values ranged from .44 to .97. For the depersonalization equation, the variance inflation factor scores ranged from 1.03 to 2.27 and tolerance values ranged from .44 to .97. For the reduced accomplishment equation, the variance inflation factor scores ranged from 1.02 to 2.29 and tolerance values ranged from .44 to .96. Variance factor scores above 5 or tolerance values below .20 indicate a problem with multicollinearity among the independent variables (Tabachnick & Fidell, 2013). Multicollinearity was not a problem.

The results for the three OLS regression equations are reported in Table 3. For the emotional exhaustion regression equation, the adjusted R-squared value was .29, which means that the independent variables explained approximately 29% of the observed variance in the emotional exhaustion index. Affective commitment had a significant negative association with emotional exhaustion, meaning that staff with higher levels of this type of commitment were less likely to be emotionally exhausted. Continuance commitment was positively related to emotional exhaustion, meaning that higher levels of this form of commitment were associated with greater emotional exhaustion. Tenure was the only control variable significantly related to emotional exhaustion. The direction of the coefficient indicates that longer tenure was associated with higher levels of emotional exhaustion. By examining the standardized regression coefficients (i.e., the values in the β column in Table 3), the size of the association can be estimated. Continuance commitment had the largest-sized effect, followed by affective commitment. Tenure had the smallest-sized effect, almost half of that of either of the forms of commitment.

OPEN IN VIEWER

Table 3:

The Effects of Affective and Continuance Commitments on Burnout.

Independent Variable	Burnout Emotional Exhaustion			Burnout Depersonalization			Burnout Reduced Accomplishment
	B	SE	β	B	SE	β	B	SE	β
Position	−0.27	0.50	−.06	1.01	0.63	.17	1.09	0.70	.17
Supervisory status	−0.95	0.51	−.16	0.33	0.64	.05	−1.12	0.71	−.15
Gender	0.13	0.36	.27	0.16	0.46	.03	−0.77	0.51	−.12
Age	0.01	0.02	.02	−0.04	0.02	−.17*	0.02	0.02	.06
Tenure	0.03	0.01	.15*	0.02	0.02	.08	0.01	0.02	.05
Educational level	0.07	0.34	.02	0.33	0.43	.06	−0.64	0.48	−.10
Race	−0.02	0.40	−.01	−0.24	0.51	−.03	0.28	0.56	.04
Average daily inmate contact	−0.06	0.12	−.04	0.07	0.15	.04	−0.50	0.16	−.24**
Affective commitment	−0.16	0.04	−.26**	−0.19	0.06	−.26**	−0.12	0.06	−.16*
Continuance commitment	0.23	0.05	.37**	0.16	0.06	.21**	0.14	0.06	.17*
F Value	7.42**			6.13**			4.03**
Adjusted R2	.29			.25			.17

Note. See Table 1 for a description of the variables and how they were coded. The number of participants used in the multivariate analysis was 160. B = the unstandardized regression slope; SE = the standard error of the slope; β = the standardized regression slope.

*

p ≤ .05. **p ≤ .01.

The second model in Table 3 sets depersonalization as the dependent variable. The independent variables accounted for about 25% of the variance observed in depersonalization (i.e., adjusted R-squared was .25). Both affective and continuance commitments were statistically significant. Higher levels of depersonalization were more likely among staff who scored lower on affective commitment and higher on continuance commitment. Age was the only control variable significantly related to depersonalization. Older correctional staff were less likely to depersonalize inmates. Based on the standard regression coefficients, affective commitment had the largest-sized effect, followed by continuance commitment, and then age.

The final model in Table 3 examines feelings of reduced accomplishment as the dependent variable. The independent variables explained approximately 17% of the variance found in the reduced accomplishment index. Both forms of commitment had significant relationships with the reduced accomplishment index. Correctional staff reporting less affective commitment and greater continuance commitment were more likely to have higher scores on the reduced accomplishment index. Average daily inmate contact was the only control variable significantly related to reduced feelings of accomplishment. The more contact a staff member had each day with inmates, the less likely he or she was to report feeling ineffective. Based on the standardized regression coefficients, average daily contact with inmates had the largest-sized effect, followed by continuance commitment, and then affective commitment.

Discussion

Burnout is a serious occupational hazard that has the potential to affect prison work environments adversely. Keinan and Maslach-Pines (2007) reported that correctional employees had much higher levels of burnout than the general U.S. population, even higher than police officers. Monitoring and addressing staff burnout in prisons is especially critical when considering the severe consequences that can result from inadequate performance. Prisons house a range of different types of criminals, including some who are violent and manipulative. Breakdowns in control emanating from the decisions and actions of burned-out personnel could threaten personal safety and lead to property damage.

The primary goal of the present study was to identify whether affective and/or continuance commitment contributed to prison staff burnout. The study adds to the literature as it examines not only emotional exhaustion but also depersonalization and reduced personal accomplishment. As noted earlier, the latter two dimensions could have more severe consequences for a prison organization than emotional exhaustion.

All six hypotheses were supported by the multiple regression analysis. Higher levels of affective commitment were associated with less emotional exhaustion, depersonalization, and reductions in feelings of accomplishment (Hypotheses 1-3). In contrast, high levels of continuance commitment were associated with greater emotional exhaustion, depersonalization, and reductions in feelings of accomplishment (Hypotheses 4-6). In other words, correctional staff who believe that they have significant investments that they feel would be lost if they were to separate from that organization were more likely to experience emotional drain and fatigue, to treat coworkers and inmates in an impersonal and callous manner, and to perceive that they are not making a positive impact. In contrast, when employees form deep affective bonds with their organization they are less likely to feel drained, less likely to treat others at work as objects, and less likely to feel ineffective. These findings call attention to the possibility of reducing severe occupational hazards like workers who depersonalize inmates and feel little occupational worth by targeting the affective and continuance commitments of staff.

One strategy for reducing correctional staff burnout validated by this study is to increase levels of affective commitment. Several studies focusing on non-criminal justice organizations have reported that efforts to increase employee’s affective commitment can reduce stress and burnout (Getahun, Sims, & Hummer, 2008; Somers, 2009). Correctional organizations can likely increase the affective commitment of their employees through efforts to improve person–organization fit. For example, evidence suggests that when employees’ values and goals are consistent with those of the organization, their level of organizational commitment tends to increase (Lambert, Altheimer, Hogan, & Barton-Bellessa, 2011). Conversely, incongruence between worker-organizational values and goals can result in dissonance and the weakening of affective bonds (O’Reilly, Chatman, & Caldwell, 1991). Thus, by providing prospective employees with realistic job previews, accurate job descriptions, and clear organizational mission statements, applicants can better assess the quality of their fit with the job and the organization. Developing and using psychometric tests to identify applicants most likely to bond affectively with an organization has also been done (Meyer & Allen, 1997).

Evidence suggests that providing organizational support significantly increases affective commitment (Griffin & Hepburn, 2005). This finding is consistent with Eisenberger, Huntington, Hutchison, and Sowa’s (1986) argument that organizations wanting affectively committed employees must demonstrate their own commitment by providing a supportive work environment. Prison administrators can demonstrate organizational support by treating employees fairly, increasing instrumental communication, finding opportunities for employee participation in decision-making, and demonstrating strong leadership skills (Griffin, 2006; Griffin & Hepburn, 2005; Lambert et al., 2008). Prison officials are also encouraged to provide ongoing training and take time to ensure that quality performance appraisals and feedback mechanisms are in place (Lambert, 2004).

Offering suggestions for reducing continuance commitment is more difficult. Correctional work environment studies suggest that continuance commitment will not be altered by variables linked to higher affective commitment, such as increasing employee input in organizational decisions, strengthening perceptions of fairness, offering greater promotional opportunities, increasing job autonomy, or building collaborative working relationships (Lambert et al., 2008; Moore & Sales, 2005). Job stress reduction programs are the only type of intervention with empirical backing for lowering levels of continuance commitment among prison staff (Lambert et al., 2008). Another avenue for scholars to explore for continuance commitment might not be so much what causes it to emerge among current prison employees, but rather who is most susceptible to high levels of continuance commitment within prison staff applicant pools. McGregor (1960) proposed that many workers simply dislike the thought of work, lack ambition, avoid responsibility, and seek job security more than anything else. Lombardo (1981) noted that a strong attraction of prison employment has been job security and the expectation of steady pay (Lombardo, 1981). The types of people identified by McGregor and Lombardo are presumably more likely to develop higher levels of continuance commitment. Correctional administrators need to be aware that building affective commitment may be more important in the long run than would be building continuance commitment.

As with all research, the current study had limitations. First, this was a single study. A single study cannot provide definitive conclusions regarding the relationship between the various forms of organizational commitment and the multiple dimensions of burnout. The prison staff studied here worked at a high-security, private facility that housed juveniles convicted in adult court, which differs from many prisons. One of the goals of the private prison was to provide education and treatment for the inmates. It could be working with offenders aged 14 to 19 in a more rehabilitative environment, could be more stressful than found in other prisons, especially for staff who took the job to help others. Freudenberger (1980) postulated that individuals who were highly idealistic and wished to change the lives of others for the better were much more likely to experience burnout than staff members who had more realistic expectations about the job. In addition, working with juvenile inmates could be more stressful than working with adult inmates.

It is unclear if working in a private or juvenile prison is more or less straining than working in a public or adult prison. Private prisons may have greater flexibility than public prisons to implement interventions to develop commitment among staff and to deal with negative stimuli. However, the benefits and promotional opportunities may be greater in a public prison system when there was only one private prison in the entire state. Moreover, the private prison studied had been in operation only 5 years prior to the survey. It takes time to build organizational commitment (Mowday et al., 1982). In addition, burnout takes time to develop and the negative stimuli of a workplace are often cumulative over time (Maslach, 1982). It is possible the effects of the findings might vary across different types of facilities and the longer a person has worked at a prison. This means that the current findings may not be generalized to other types of correctional institutions. In sum, the relationship between affective and continuance commitments on the dimensions of burnout could be situational and contextual, and, as such, might vary between public and private prisons, between adult and juvenile facilities, by security level, and by region. Studies at other prisons as well as other types of correctional organizations (e.g. jails, probation agencies) are needed to determine whether these results can be replicated and whether the current results can be generalized to other types of correctional facilities.

A major limitation was that the current study relied on items adapted from the federal Prison Social Climate Survey, which measured the three dimensions of burnout using items somewhat similar to the Maslach Burnout Inventory, but not the actual Maslach Burnout Inventory itself (Maslach & Jackson, 1981; Wright & Saylor, 1991). In addition, rather than using the seven-point frequency response scale (i.e., never to always), the present study utilized a five-point Likert-type response scale (i.e., strongly disagree to strongly agree). This makes comparing the current findings with results from other studies that utilized the Maslach Burnout Inventory more difficult. The authors recommend that future research in this area use the Maslach Burnout Inventory, so the level of burnout and its associations with workplace variables can be compared with other studies not only on correctional staff but other occupational fields. The Maslach Burnout Inventory is a widely used measure of the three dimensions of burnout and is viewed as both a valid and reliable measure (Langballe, Falkum, Innstrand, & Aasland, 2006; Maslach et al., 1982).

Another shortcoming was the use of cross-sectional data. One-shot empirical studies do not allow for clear tests of temporality or causality. For example, the present study cannot rule out the possibility that the burnout of prison staff actually precedes either type of organizational commitment in time. Longitudinal studies are needed to examine the chronological order of the organizational commitment–burnout relationship. Furthermore, future studies should examine the process from workplace stressors (i.e., workplace factors, such as role conflict, dangerousness of the job, or poor supervision) to job stress to burnout to determine if the two forms of commitment mediate/moderate the effects of the workplace on correctional staff. This was beyond the scope of the current study. Additional research is also necessary to explore the relationship of other factors with burnout among correctional staff. Importantly, the amount of variance explained across the equations differed noticeably for the three dimensions of burnout. Approximately, 29%, 25%, and 17% of the variance in emotional exhaustion, depersonalization, and feelings of ineffectiveness were explained by variables in this study. This means that other variables accounted for the remaining variance of each of the burnout measures. Future research is needed to determine other possible correlates of correctional staff burnout. Without more solid knowledge in this area, making burnout prevention recommendations to correctional administrators will continue to be haphazard.

In closing, burnout can be devastating not only for the staff member but also for coworkers, inmates, and the correctional organization itself. In this study, the associations between two forms of organizational commitment and three dimensions of correctional staff burnout were examined. Both affective and continuance commitments had statistically significant relationships with burnout. These are two areas in which administrators should focus their efforts to reduce burnout among the staff at correctional facilities. The authors hope that this study will spur more research on how organizational commitments are associated with staff burnout.