Eugene A. Paoline, III Department of Criminal Justice, University of Central Florida, Orlando, Florida, USA Eric G. Lambert Department of Legal Studies, University of Mississippi, University, Mississippi, USA Nancy L. Hogan School of Criminal Justice, Ferris State University, Big Rapids, Michigan, USA
Despite increased representation in correctional work settings, women still encounter obstacles in thismale-dominated occupation—obstacles that have the potential to affect their levels of job stress andjob satisfaction. Although gender-based differences in job stress and job satisfaction have been analyzedin several prison settings, much less work has been conducted in the often neglected correctionalarena of jails. The current study fills this empirical void by examining jail staff at a large countycorrectional system in Orlando, Florida. Ordinary least squares regression analysis indicated thatdifferent facets of the work environment differentially affected the job stress and job satisfaction of419 women and 493 men working in a large urban jail system. Specifically, role ambiguity, perceiveddangerousness, coworker relations, input into decision making, and administrative support had largereffects on job stress for women compared to men. For job satisfaction, the only workplace variable tohave a gendered effect was administrative support, which also had a greater effect for women thanmen. The findings reveal gender-related differences, especially in terms of job stress.
The number of women employed in institutional corrections has increased dramatically during thepast 50 years. Although men still make up the majority of correctional staff, the number of womenhas moved from very few to a substantial proportion (Lambert, Hogan, Altheimer, & Wareham,2010). For example, approximately 40% of correctional staff were women in 2006, which is aremarkable increase from 7% in the late 1970s (MTC Institute, 2008). Although a more diverseworkforce has been a positive development, working in male-dominated correctional facilities hasnot been an easy undertaking for many women. As Griffin (2006) described, correctional institutionstend to be highly masculine workplaces. Griffin, Armstrong, and Hepburn (2005) furthernoted that being a man in highly masculinized environments is often seen as a positive attribute.Furthermore, not all male correctional staff have welcomed women joining the correctional staff ranks as equals (Farkas, 1999; Griffin, 2006; Tewksbury & Collins, 2006; Zupan, 1986). Femalestaff have had to fight for their right to work in correctional facilities, combating stereotypicalattitudes that they would not be able to handle inmates or protect either themselves or theircoworkers if the need arose (Hogan, Lambert, Hepburn, Burton, & Cullen, 2004; Pollock,1995; Zupan, 1992). In addition, some female staff must deal with unfavorable treatment inthe workplace, particularly in the form of sexual harassment (Britton, 1997; Savicki, Cooley,& Gjvesvold, 2003). As Pogrebin and Poole (1997) contended, some correctional staff viewfemale staff negatively whether they do their jobs well or poorly. Overall, many women face a‘‘damned if you do and damned if you don’t’’ response from some of their male colleagues.In light of the obstacles women face in correctional occupational environments and the importantrole that staff play in the operation of secure, safe, and humane facilities, a growing body ofliterature has explored how male and female staff perceive the workplace and whether men andwomen are affected differently by workplace variables. One comparatively recent approach hasbeen to examine whether the same workplace variables have different effects between men andwomen on salient work outcomes, such as job stress and job satisfaction (Griffin, 2006). Theapproach to see whether workplace variables have gendered effects postulates that men andwomen may be differentially influenced by the various aspects of the correctional workplacebecause of their differing experiences, perceptions, expectations, and needs. This approach recognizesthe ‘‘mediating effect of gender on the influence of multiple work environment factors’’ bytrying to uncover the specific predictors that lead to differences (or similarities) between male andfemale officers (Griffin, 2006, p. 5). In other words, even for the same outcome, such as job stressor job satisfaction, some workplace variables may be more important in influencing the outcomearea for women than men (or vice versa).The current study attempts to expand the literature by using such an approach for job stress andjob satisfaction among women and men working in a large urban jail. Not only is there a need formore testing of possible gender differences between female and male correctional staff, but themajority of the past research has focused exclusively on staff working in prisons. In the currentstudy, several dimensions of the correctional work environment were explored to determinewhether their effects on job stress and job satisfaction varied between male and female jail staff.As relatively unstressed and satisfied jail staff are critical for the proper functioning of jails, so toois the need for research on possible differences in the antecedents of job stress and job satisfactionbetween women and men. This is especially true given the increased proportion of womenworking in jails.
The research on differences between women and men correctional staff can be grouped into twocategories (Lambert, Altheimer, & Hogan, 2010). The first line of research has examined whetherthere are differences between women and men in given outcome areas, such as job stress, job satisfaction,job burnout, and perceptions of work supervisors. The results of such inquiries haverevealed mixed findings. For example, some studies have illuminated gender differences in levelsof job stress and job satisfaction (Carlson, Anson, & Thomas, 2003; Cheeseman Dial, Downey, &Goodlin, 2010; Garland, McCarty, & Zhao, 2009; Tewksbury & Collins, 2006; Wright & Saylor,1991), whereas other research has reported no differences on these outcome areas (Blevins, Cullen, Frank, Sundt, & Holmes, 2006; Castle, 2008; Griffin, 2006; Triplett, Mullings, & Scarborough,1996). A criticism of this type of research is that only including a dichotomous measure forgender fails to take into account how different workplace areas can affect the same outcome indifferent ways. Although female and male correctional staff may perceive things the same wayand have similar levels of whatever outcome is being examined, they may place different valueand importance on different aspects of the workplace that contribute to the outcome. For example,although female and male correctional staff may report similar levels of job stress, the reasons forbeing stressed may differ by gender. For example (and hypothetically speaking due to lack ofempirical research), female staff may be stressed by harassment, a work scheduling conflictkeeping them away from home, and a lack of supervisory support, whereas male staff may bestressed because of dissatisfaction with pay and a lack of promotional opportunities.The second category of research focuses on the extent to which workplace variables havedifferent effects on outcome areas for women and men working in correctional facilities. Thisapproach postulates that although the outcome level may be similar for women and men, the variablesthat explain the outcome area may vary by gender (Griffin, 2006; Lambert, Altheimer, &Hogan, 2010). Research has found empirical support for this approach. For example, in a studyof federal prison staff, Britton (1997) found that supervision played a greater role in shaping thejob satisfaction of women than men. Also, Triplett, Mullings, and Scarborough (1999) reportedthat work–family conflict was only a predictor of job stress for female correctional staff, and roleoverload was only a significant predictor of male job stress among southwestern prison staff.Among southwestern jail staff, Griffin (2001) observed that the fear of being victimized wasinversely linked to job satisfaction for women but not for men. In addition, Savicki et al.(2003) reported that harassment was linked with lower levels of organizational commitmentand higher levels of job burnout for women but not for men among northwestern prison staff.Griffin (2005), in a study of Arizona prison staff, found that a perceived lack of employmentopportunities was a significant predictor of continuance commitment (i.e., a desire to remain withthe agency) for men, whereas being the primary income earner at home was a significant predictorfor women. Griffin (2006) reported that, also among Arizona prison staff, quality of supervisionand organizational support for equal treatment policies both had significant effects on job stressfor men but not women. Utilizing data from staff at a midwestern prison, Lambert, Altheimer, andHogan (2010) found that work–family conflict was positively associated with job stress and negativelyassociated with both job satisfaction and organizational commitment among women but notmen. Moreover, they reported that role ambiguity, role overload, and perceived dangerousnessinfluenced levels of job stress for men but not women, and role conflict and role ambiguity influencedlevels of job satisfaction for men but not for women. It is important to note that although theaforementioned studies found gender differences, the vast majority of them also found many similaritiesin the effects of workplace variables on outcome areas for both women and men. Overall,the findings suggest that women and men may respond differently to some correctional workplacefactors and similarly to others.
Although the literature supports the contention that there are gender differences for someworkplace variables, two other conclusions can also be drawn. First, almost all of the researchin this area to date has focused on prison staff. In fact, of the previously mentioned studies, onlyone focused on jail staff. The gendered effects of workplace attributes could vary by type offacility, or gender differences and similarities found in past prison staff studies might also applyto jail staff. In order to answer this question, more research focusing on jail staff is needed. The second conclusion is that researchers are just beginning to scratch the surface in their inquiriesof whether workplace variables vary in predicting salient outcomes for female and malecorrectional staff. Although several different workplace factors have been studied to determinewhether there are gendered effects, many other variables remain unexplored. This study, therefore,was undertaken to explore whether there were gender differences on the effects of role ambiguity,role conflict, perceived dangerousness of the job, pay perceptions, coworker relations, input intodecision making, job variety, views on inmate control, and administrative support on job stress andjob satisfaction among staff at a large southern jail with a large degree of gender diversity.Job stress and job satisfaction are salient outcome areas for employees in general, andespecially those in correctional organizations. In the correctional literature, job stress is seen asfeelings of work-related tension, anxiety, frustration, and worry (Cullen, Link, Wolfe, & Frank,1985; Triplett et al., 1999). It generally results from workplace factors that are a strain or negativestimuli for staff members (Griffin, 2006). In the long run, excessive stress has negative outcomesfor staff as well as for the overall correctional setting. Job stress has been linked with higher healthproblems, substance abuse, relationship problems, suicide, burnout, absenteeism, and turnoveramong correctional staff (Cheek, 1984; Cheek & Miller, 1983; Dowden & Tellier, 2004; Lambert,Edwards, Camp, & Saylor, 2005; Mitchell, MacKenzie, Styve, & Gover, 2000; Slate & Vogel,1997).
Job satisfaction is an affective response by an employee concerning his or her particularfunction in an organization (Cranny, Smith, & Stone, 1992). Simply stated, job satisfaction isthe degree to which an employee likes his or her job (Spector, 1996). Job satisfaction generallyincreases after the experience of positive workplace variables that are desired by a staff memberand decreases after exposure to strain and other negative workplace variables. Like job stress, jobsatisfaction has significant consequences for both staff members and correctional organizations.Higher levels of job satisfaction have been found to be associated with greater support forrehabilitation, greater compliance with organizational rules, and increased satisfaction with life(Cullen et al., 1985; Fox, 1982; Jurik & Halemba, 1984; Lambert, Hogan, Paoline, & Baker,2005), whereas lower levels of job satisfaction have been linked to burnout, absenteeism, andturnover (Byrd, Cochran, Silverman, & Blount, 2000; Griffin, Hogan, Lambert, Tucker, & Baker,2010; Lambert, Edwards, et al., 2005; Lambert & Hogan, 2009; Mitchell et al., 2000).Based on the premise that workplace variables may differ in their effect for women and menand the central features of several prior correctional studies, the effects of workplace variables ofrole ambiguity, role conflict, perceived dangerousness of the job, instrumental communication,pay perceptions, coworker relations, input into decision making, job variety, views on inmatecontrol, and administrative support on job stress and job satisfaction were hypothesized to varybetween women and men working in a jail. Role ambiguity occurs when staff are unclear of whatis expected of them as well as receive unclear or no information about carrying out the duties andresponsibilities of a given position or job (Triplett et al., 1999). Role conflict arises when a staffmember is given orders or directions that conflict with one another (Triplett et al., 1999).Perceived dangerousness of the job refers to a perception that the job is viewed as being a riskto the employee’s safety and well-being (Hartley, Davila, Marquart, & Mullings, 2013). Instrumentalcommunication is the perception by staff that the level of information they receive in orderfor them to function effectively at work is adequate (Lambert & Paoline, 2012). Pay perceptionsrefers to views by staff that the pay is fair and that they are satisfied with it (Lambert, Reynolds,Paoline, & Watkins, 2004). Coworker relations focuses on perceptions of having positive interactions with and support from coworkers (Lambert & Paoline, 2012). Input into decisionmaking deals with the perception that staff are allowed a voice in salient organizational issuesand decisions (Slate & Vogel, 1997). Job variety is the degree of variation in the job (Lambert& Paoline, 2012). Views on inmate control focuses on the perception of the level of control ofinmates in the work area of a staff member (Castle & Martin, 2006; Paoline & Lambert,2012). Administrative support is the degree to which a staff member feels that he or she is supportedby the administration of the facility (Griffin, 2006). To the degree that these workplacevariables have similar effects for women and men on job stress and job satisfaction, the contentionthat workplace variables have different effects on female and male jail staff is not supported.Conversely, if differences are observed, it suggests that women and men place different levelsof importance on various aspects of the jail work environment in shaping their job stress andjob satisfaction.
The current inquiry was based on data from a survey that was provided to all available staff atthe Orange County Corrections Department (OCCD), a large county jail complex located inOrlando, Florida. The county contains one major city and 10 other municipalities, all of whichrely on the county jail for local detention needs. At the time of the survey, the jail employedapproximately 1,500 paid staff and typically housed more than 4,000 inmates.Under the direction of an ad hoc oversight commission that was interested in exploring concernsand issues among the jail’s staff (i.e., the organizational climate), a 155-item survey wasadministered over a 5-day period. Survey items were constructed from extant research as wellas a series of focus groups (i.e., seven 2-hour meetings) with 48 OCCD employees from differentorganizational levels, facilities, and assignments. The OCCD staff were informed that the surveywas completely voluntary in nature and that the responses would be anonymous. With the consentof the jail director, jail personnel received 2 hours of overtime for participating in the survey.The administration of the survey took place across each of the three primary shifts (i.e.,6 a.m.–5:30 p.m., 3 p.m.–2:30 a.m., and 7 p.m.–6:30 a.m.), and staff were afforded the opportunityto take part in any of the survey times even if it was not their assigned shift or work day. Of the1,500 paid employees at the facility during the week the survey was administered, 1,062 staffmembers participated in the survey, which resulted in a response rate of slightly more than 70%.Respondents represented all areas of the jail facility, such as correctional officers, casemanagers, medical staff, industry staff, foodservice workers, and so on. Moreover, therespondents represented various administrative levels of the correctional facility, from line staffto supervisors and managers.Of the participants, 46% were women and 54% were men. Age was measured as an ordinalvariable, with 4% of the participants younger than 25 years old, 9% between 25 and 29, 17%between 30 and 34, 21% between 35 and 39, 19% between 40 and 44, 12% between 45 and49, 11% between 50 and 54, and 6% 55 years old or older. Median tenure in months workingat the jail was 72 and ranged from 0 to 336 (i.e., 0–28 years). In terms of education, slightlyless than 0.5% had less than a high school diploma, 21% had a high school diploma or general equivalency diploma, 43% had some college but no degree, 15% had an associate’s degree, 16%had a bachelor’s degree, 4% had a master’s degree, and slightly less than 1% had a professional orterminal degree. Moreover, 10% indicated that they were supervisors of other staff. Approximately40% of the respondents were Black, 11% Hispanic, 43% White, and 6% of another racial=ethnicgroup. Participants represented all areas of the correctional facility, with 67% indicating that theywere correctional officers (i.e., worked in a custody position).
In this study, job stress and job satisfaction represented the dependent variables. Job stress wasmeasured using six items from Crank, Regoli, Hewitt, and Culbertson (1995). The responsesto these six items were summed to form an additive index that had a Cronbach’s alpha valueof .78 for internal reliability. Job satisfaction was measured using five items from Brayfieldand Rothe (1951). The responses to the five items were summed to form an additive index thathad a Cronbach’s internal reliability value of .82. The individual survey items that were utilizedto construct the job stress and job satisfaction indices are presented in the Appendix.
For this study, the independent variables, capturing various facets of the work environment,were role ambiguity, role conflict, perceived dangerousness of the job, instrumental communication,pay perceptions, coworker relations, input into decision making, job variety, inmatecontrol views, and administrative support. The individual survey items used to measure these conceptsare also presented in the Appendix. Role ambiguity was measured by six items from Cullenet al. (1985) that were used to form a summed index of role ambiguity and had a Cronbach’s alphaof .75. Role conflict was based on a single item from Poole and Regoli (1983). An index measuringperceived dangerousness of the job was created by summing the responses to five items derivedfrom Cullen et al. The items had a Cronbach’s alpha internal reliability value of .82. The five itemsmeasuring instrumental communication items were adapted from Curry, Wakefield, Price, andMueller (1986) and were summed to form an additive index that had a Cronbach’s alpha valueof .89. Three items created from the focus group meetings were used to measure pay perceptions.The summed index had a Cronbach’s reliability coefficient of .80. Using three items based on thosefrom Mueller, Boyer, Price, and Iverson (1994), we formed an index to measure coworker relations,and it had a Cronbach’s alpha of .83. Input into decision making was measured using four itemsfrom Curry et al. The responses to these items were summed to form an index with a Cronbach’salpha coefficient of .88. Four items from Curry et al. were used to measure perceived job variety bysumming the responses to these items to form an index that had a Cronbach’s alpha value of .65.Views on inmate control were measured using five items created for the study. The responses tothese items were used to form an additive index that had a Cronbach’s internal reliability valueof .71. Using five items derived from the findings of the focus groups, we created an index foradministrative support. This index had a Cronbach’s alpha of .79. Finally, six measures of personalcharacteristics were included in the analyses: age, tenure, education, supervisory status, race, and position. These measures have been part of numerous attitudinal studies as well as works specificallygeared toward examinations of correctional job stress and satisfaction. Please see Table 1 for adescription of each of these individual measures.
After listwise deletion of missing values on the variables, there were 912 participants, of which419 (46%) were women and 493 (54%) were men. The descriptive statistics for the study variablesare presented in Table 1. An examination of this table reveals considerable variation in boththe dependent and independent variables. In terms of respondents, the majority were non-White,worked in custody, and were not supervisors of other employees. In addition, the typical respondentwas in his or her late 30s to early 40s, had taken some college courses (but had no degree),and had been at his or her current position for approximately 6 to 8 years. Across the attitudinal indices, the Cronbach’s alpha values (as a measure of reliability) were all above the generallyaccepted minimum level of .60 (Gronlund, 1981). In addition, a principal-axis factor analysisfor each latent variable was conducted in which the items for each index were examined to ensurethat a single factor was extracted for each latent concept. The results (not reported) indicated thateach index was in fact measuring a single factor.
Because the focus of the study was to determine whether work environment variables had differenteffects on job stress and satisfaction between female and male jail staff, separate analyseswere conducted across each gender. Pearson correlations for female jail staff are presented inTable 2. The only personal characteristic that had a statistically significant correlation with jobstress among female staff was position. That is, female correctional officers were more likely toreport job stress compared to female staff in other positions. However, all of the work environmentvariables had significant correlations with job stress. More specifically, increases in role ambiguity,role conflict, and perceived dangerousness of the job were all associated with higher levelsof stress for women working in jail settings. In addition, increases in instrumental communication,pay perceptions, coworker relations, input into decision making, job variety, views on inmate control,and administrative support were all related to reductions in self-reported job stress.
In terms of job satisfaction, many of the same significant correlates existed, although in theopposite direction. There were also more statistically significant correlations for the personalcharacteristics. For example, older female staff on average reported higher satisfaction with theirjobs compared to their younger counterparts. In addition, female supervisors were in generalmore satisfied than nonsupervisory female staff. White female jail staff tended to report higherjob satisfaction than did non-White female jail staff. With respect to the work environment variables,all except role conflict had significant correlations with job satisfaction among the femalejail staff. More specifically, role ambiguity and perceived dangerousness of the job both hadnegative corrections, whereas instrumental communication, pay perceptions, coworker relations,input into decision making, views on inmate control, and administrative support had positiveassociations with job satisfaction.
Next, the correlations for male jail staff are presented in Table 3. Among the personalcharacteristics, tenure and supervisory status had statistically significant correlations with jobstress. That is, men in supervisory positions reported lower stress from the job compared to thosein nonsupervisory roles, while more experienced staff were more stressed on the job. All of thework environment measures had significant correlations. Role ambiguity, role conflict, andperceived dangerousness were all correlated with higher job stress among male jail staff.Conversely, instrumental communication, pay perceptions, coworker relations, input intodecision making, job variety, inmate control views, and administration support each had a negativeassociation with job stress.
In terms of male job satisfaction, there were two statistical associations with the personalcharacteristics. More specifically, older male staff reported higher levels of job satisfaction thantheir younger counterparts, and male supervisors were more satisfied with their job than theirnonsupervisory male peers. As with job stress, all of the work environment variables had a statisticallysignificant correlation with job satisfaction. That is, role ambiguity, role conflict, andperceived dangerousness of the job were all negatively correlated with job satisfaction, whereasinstrumental communication, pay perceptions, coworker relations, input into decision making,job variety, views on inmate control, and administrative support were positively connected tosatisfaction with the job.
Ordinary least squares (OLS) regression was used to estimate the effects of the work environmentvariables on job stress and job satisfaction for male and female jail staff (i.e., separateanalytical models) while controlling for the shared effects of age, tenure, education, supervisorystatus, race, and position. The results are presented in Table 4. Although not reported, the varianceinflation factor scores and tolerance values did not indicate a problem with collinearity ormulticollinearity for any of the regression equations. In addition, issues with outliers, influentialcases, normality, linearity and homoscedasticity of residuals, and independence of errors in theregression analysis were tested, with no concerns.
In the OLS job stress regression model for women jail staff, the R2 value was .47, indicatingthat the independent variables explained approximately 47% of the variance in the dependentvariable. Among the personal characteristics, tenure, supervisory status, and position all had asignificant (independent) association with the job stress measure. That is, increases in tenurewere associated with higher levels of job stress. Also, supervisors on average reported morestress from work than did nonsupervisory jail staff. It is interesting that, in terms of position,female correctional officers reported less stress from work than did their female counterpartsin other areas. Among the work environment variables, role ambiguity, perceived dangerousnessof the job, coworker relations, input into decision making, and administrative support had significantassociations with job stress. Increases in role ambiguity and perceived dangerousness of the job were associated with increased job stress for female staff. In contrast, increases inpositive views of coworker relations, input into decision making, and administrative supportwere related to lower levels of stress from the job. Based on the standardized regression coefficients,perceived dangerousness of the job had the greatest effect, followed closely by role ambiguity.Supervisory status had the smallest effect on job stress among the significant variables.
With regard to the job satisfaction model for women, approximately 48% of the variance wasexplained by the independent variables. The only personal characteristic to have a significantassociation with job satisfaction in the regression equation was race. In general, White femalestaff reported higher levels of satisfaction from work than non-White females. Among the workenvironment variables, coworker relations, input into decision making, job variety, and administrativesupport all had significant associations with job satisfaction. That is, increases in each ofthese characteristics were related to greater satisfaction from the job. Based on the standardizedcoefficients, job variety had the largest impact, whereas race had the smallest statistical associationwith job satisfaction.
In the OLS job stress regression model for men, the independent variables accounted forabout 44% of the variance found in the dependent variable. Tenure had a positive associationwith job stress, meaning that the longer a male staff member worked at the jail, the higherthe level of stress from work reported. Position also had a significant effect, with correctionalofficers reporting typically less job stress. Role conflict and perceived dangerousness of thejob both had significant positive associations with job stress. Increases in perceptions of positivecoworker relations were related to decreased job stress. All of the remaining independent variableshad nonsignificant associations with job stress. Looking at the standardized regressioncoefficient values, we find that perceived dangerousness of the job had the largest effect,whereas role conflict had the smallest statistical effect.
Among male jail staff members, the independent variables explained approximately 48% of thevariance in the job satisfaction model. The only personal characteristic to have a significant relationshipwas age. Older male staff tended to report higher levels of job satisfaction. Among the workenvironment variables, instrumental communication, coworker relations, input into decision making,and job variety were significant predictors. Increases in each of these work environment measureswere associated with greater satisfaction from the job. Job variety was the strongest predictor,having more than twice the effect on job satisfaction as the other significant variables.
The results from Table 4 suggest that there may be gender differences across the explanatoryfactors of job stress and job satisfaction among the jail staff in this study. In order to determinewhether the observed gender differences were statistically significant, we conducted the equalityof regression coefficients test (Paternoster, Brame, Mazerolle, & Piquero, 1998). The null hypothesiswith this test is that the difference between the regression coefficients is zero, and anyobserved differences are the result of random chance. The alternative hypothesis is that there isa significant difference in the regression coefficient of a particular variable between femaleand male jail staff. The results of the equality of regression coefficients test are reported in Table 5.
For job stress, the equality of regression coefficients test was not calculated for age, education,race, instrumental communication, pay perceptions, job variety, or inmate control because of thelack of significant results in the OLS regression equation for job stress. The equality of regressioncoefficients test results indicated that the size of the regression slopes for tenure, supervisory status,and role conflict were not significantly different between female and male staff. The results ofthe equality of regression coefficients test did, however, indicate several effects (i.e., significant differences). Although position, perceived dangerousness of the job, and relations with coworkerswere significant predictors of job stress for both women and men, the size of the effects differedby gender. That is, female correctional officers reported greater stress from the job than did menholding the same position. Also, perceived dangerousness of the job had a slightly greater positiveeffect on job stress for women than it did for their male counterparts. Conversely, relations withcoworkers had a slightly stronger negative association with job stress for men than it did forwomen. The equality of regression coefficients test also indicated that the significant effectsfor women but not men for role ambiguity, input into decision making, and administrative supporton job stress were probably not due to random chance. In other words, these variables hadsignificant relationships with job stress for women and not for men.
For job satisfaction, the equality of regression coefficients test was not calculated for tenure,education, supervisory status, position, role ambiguity, role conflict, perceived dangerousness ofthe job, pay perceptions, or inmate control views. Specifically, although age and instrumentalcommunication were statistically significant predictors of job satisfaction for men in the OLSregression analysis, and coworker relations, input into decision making, and job variety were significantpredictors of job satisfaction for both women and men (see Table 4), the equality ofregression coefficients test indicated that most of the effects of the independent variables did not differ by gender (i.e., they had similarly sized, nonsignificant effects for both groups). That is,even though the OLS regression analysis revealed that age and instrumental communication hadsignificant relationships with job satisfaction for men but not women, the equality of regressioncoefficients test indicated that the difference across gender for these two variables was not statisticallysignificant. In other words, there was no difference in the effects of age and instrumentalcommunication on job satisfaction by gender. Finally, the equality of regression coefficients testdid indicate that there was a significant difference in the size of the regression coefficients forwomen and men for race and administrative support. For women, White staff reported greaterjob satisfaction than did non-White staff. For men, race had a nonsignificant association withjob satisfaction. For women, administrative support had a positive effect on job satisfaction,whereas for men it had a nonsignificant effect.
DISCUSSION AND CONCLUSION
The aim of the current study was to explore whether the effects of selected workplace variablesdiffered in predicting the job stress and job satisfaction levels of female and male jail staff. Toachieve this aim, we ran separate regression analyses for women and men, which revealed anumber of interesting findings, some of which supported the contention that workplace variableswould have different effects on job stress and job satisfaction among women and men in a jailsetting and some of which did not. In terms of similarities (i.e., not supporting the postulationthat the effects of workplace variables would differ by gender), the most powerful predictorsof job stress and job satisfaction, dangerousness and job variety, respectively, were the samefor women and men. That is, both women and men with higher perceptions of dangerousnessreported greater levels of job stress, and those with more job variety reported that they weremore satisfied. These preliminary regression analyses also illustrated that both genders withmore positive perceptions of coworker relations were less stressed and more satisfied at work.Finally, both women and men who reported greater input into decision making expressed higherlevels of job satisfaction. Although such findings do not generally support the contention thatworkplace variables have differing effects for female versus male jail staff, practitioners(especially in management positions) interested in staff stress and job satisfaction would benefitfrom an understanding of how they can possible affect such dimensions of their jails. In terms ofwhat matters, to reduce job stress, jail management should focus on reducing the level of dangerthat staff perceive while also promoting the role=importance of the immediate workgroup (i.e.,coworker relations). Likewise, efforts to increase job satisfaction among women and men wouldbenefit from allowing staff more variety in their jobs and increased input into decision makingand promoting the role=importance of the immediate workgroup. Such themes are all consistentwith a supportive approach to managing personnel (McGregor, 1990).
Although similarities were found among jail staff, results also point to a number of genderrelatedstatistical differences, especially with respect to job stress. The findings from the equalityof regression coefficients test illustrated that although both women and men who perceived higherlevels of danger in the work environment reported greater levels of stress, the effect was morepronounced for women. Similarly, both women and men who perceived higher levels of coworkerrelations reported less job stress, but the effect was greater for men. In addition, more inputinto decision making and administrative support resulted in less stress for women but not men, whereas higher levels of role ambiguity led to increased job stress for women but not men.Gender-specific statistically significant results for job satisfaction were far fewer. Similar tothe results found for job stress, women placed a greater importance on administrative support.Overall, such gendered results suggest that administrators, especially those concerned with occupationalstress and job satisfaction, should be aware that features of the jail work environmentdifferentially impact female and male staff. Administrators need to make supervisors aware thatworkplace factors can affect male and female staff differently, thus allowing supervisors to tailortheir responses in varied ways to different groups of staff. In addition, part of initial and refreshertraining could cover the issue of how workplace factors affect female and male staff differently aswell as cover various coping mechanisms. Furthermore, peer mentoring programs could be institutedin which same-sex staff are paired together in order to help staff both avoid stressfulworkplace stimuli and effectively cope with stressful situations. In the proposed mentoringprogram, senior staff who have a positive work history would be trained and paired with new staffmembers. Finally, focus groups based on gender could be set up to allow staff to provideinformation on different aspects of their jobs that cause them to feel stress.
The current study is not without its limitations. Similar to other correctional studies conductedin prisons (Britton, 1997; Griffin, 2005, 2006; Savicki et al., 2003; Triplett et al., 1999), ourquantitative approach focused on perceptual components of the work environment. Althoughthis is useful for statistically isolating specific components of the work environment for whichmen and women differ in their perceptions of workplace factors and how they are associatedwith job stress and job satisfaction, it does not fully inform about possible gender-related divergentexperiences, expectations, and needs. A theoretical gendered model of why workplace variableswould differ in their effects on work outcomes, such as job stress and job satisfaction,among correctional staff needs to be constructed. Such a theoretical construction would helpfocus future testing on how various dimensions of the correctional work environment affect maleand female staff. Developing and testing a gendered model of correctional staff would assist inobtaining a more comprehensive understanding of how such perceptions are acquired and whytheir effects differ between female and male staff. Furthermore, future work complementing thatwhich was presented here would benefit from incorporating qualitative approaches that couldcapture informal socialization patterns and behavioral interactions among jail staff that couldnot be gleaned with a structured survey instrument. In this way, observations and open-endedextended interviewing techniques could be used to shed light on the ways in which experiences,expectations, and needs are shaped and developed for men and women.
In addition, the results presented here are from staff at a single jail organization. Although thefindings are likely to generalize well to other large county jail facilities, they might not beequally applied to smaller ones in which the dynamics of job stress and satisfaction might differ.Thus, populations at other jails of varying size and location should be examined to determinewhether the results can be replicated. Furthermore, additional research is needed not only at jailsbut at other facilities (e.g., juvenile facilities and prisons of varying levels of custody). Genderedeffects of workplace variables could vary by type of institution. Currently, there is not enoughcumulative empirical knowledge to determine whether the gendered effects of workplace variablesare universal or whether they are situational across different correctional settings.
Also, although we were encouraged by the large amount of variance explained for job stressand satisfaction across both men and women, slightly more than 50% of the variance was notaccounted for by the variables in the regression analyses. Future studies should work toward identifying additional workplace variables to assess whether they have gendered effects on jobstress and job satisfaction. For example, future researchers may wish to include dimensions ofthe workplace such as role overload (being asked to do too much), quality of supervision, andwork–family conflict (in which the spillover of the domains of work and home causes conflictfor a person).
Finally, this study looked at two outcome areas, job stress and job satisfaction. Although theseare important correctional dimensions, there are others that should be examined for gendereffects. Other salient outcome areas include organizational commitment, job involvement,burnout, absenteeism, turnover intent=turnover, performance, and organizational citizenshipbehaviors (going beyond what is expected at work). It would also be interesting to test the genderedapproach model at facilities that employ differing proportions of women. The sample herewas a nearly equal gender split. As has been noted by researchers, when women are a smallminority of the workforce, they can face greater tokenism compared to men (Kanter, 1977).Such tokenism results in greater barriers and obstacles for minority employees (Matthews,Monk-Turner, & Sumter, 2010). When the number of women working in a correctional facilityis higher, critical mass can be reached in which greater equality and meaningfulness in the organizationcan be reached. Thus, gender differences might vary by the proportion of women workingin the given institution. As the number of women working in jails and prisons increases, it is verypossible that the results from studies will vary as well. Clearly, there is a need for much moreresearch on the gendered effects of the correctional workplace as well as construction and testingof gendered models of why workplace variables may or may not differ in explaining salientoutcomes of female and male correctional staff.
In closing, staff are one of the most important resources for any jail. In light of theirimportance, there has been growth in the number of studies that have examined how workplacefactors affect staff. The vast majority of these studies have focused on prison staff, even thoughjail staff play a vital role in the field of corrections. Moreover, many past studies have assumedthat women and men are affected in similar manners by workplace variables. This may not betrue, as the effects of different workplace variables may vary for women and men because oftheir divergent experiences, expectations, values, and needs. This study examined the effectsof different aspects of the workplace on the job stress and job satisfaction of women and menworking in a large southern urban jail. Mixed results were found regarding support for thepostulation that workplace variables differ in their effects on women and men working in a jailsetting. In some ways, the workplace variables tested in this study had similar effects on womenand men, particularly for the outcome area of job satisfaction. However, there were differencesin the effects of a number of workplace variables, especially for job stress, among the womenand men in this study. With growing gender equality in the workforce of jails, understandingthe similarities and differences of women and men in terms of how workplace factors influencejob stress and job satisfaction is paramount. Satisfied and unstressed staff are needed for jails tomeet their goals and objectives. In an era of increasing inmate populations, rising costs,shrinking budgets, and personnel shortages, lowering job stress and increasing job satisfactionis crucial. In order to do this, empirical information is needed. Without this information, neitherscholars nor administrators will be able to understand how workplace variables may or may nothave gendered effects on staff. The findings of this study support the contention for continuedresearch on jail staff, and if there are differences for women and men, not only jail staff but staffin other types of correctional organizations.
Eric Lambert and Eugene Paoline contributed equally to the article and are listed in randomorder. In addition, we thank the editor and the anonymous reviewers for their comments andsuggestions to improve the article.