R o l e c o n f l i c t s , e m o t i o n a l
e x h a u s t i o n a n d h e a l t h
p r o b l e m s : a s t u d y o f p o l i c e
o f f i c e r s i n T a i w a n
Chia-huei Wu*,†
Institute of Business and Management, National Chiao Tung University, Hsinchu, Taiwan
* Correspondence to: Chia-huei Wu, 4F, No.147, Sec. 1, Xinguang Rd., Wenshan District, Taipei City 116, Taiwan.
† E-mail: b88207071@ntu.edu.tw
Stress and Health 25: 259–265 (2009)
Published online 8 January 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/smi.1245
Received 14 July 2008; Accepted 17 November 2008
Summary
This study focused on the role confl icts faced by police offi cers in Taiwan and investigated the relationship between role confl icts, emotional exhaustion and health problems based on the job demands and resources model (JD-R model). Firstly, in addition to work–family confl ict, an inter-role confl ict, this study identifi ed an intrasender inter-role confl ict for police offi cers called uncompro-mising-friendly confl ict. Secondly, according to the JD-R model, it was hypothesized that emotional exhaustion has a complete mediation effect on the relationships between role confl icts and health problems. The survey data in Huang’s study were analysed. A total of 672 police offi cers at Kaohsiung, Taiwan provided completed data for analysis. Correlation analysis showed that role confl icts, emotional exhaustion and health problems were positively correlated. Results of the structural equation model further suggested that the positive effect of role confl icts on health problems was partially mediated by emotional exhaustion. Copyright © 2009 John Wiley & Sons, Ltd.
Key Words
role confl ict; exhaustion; burnout; job demands; police
Introduction
Policing is one of the most stressful jobs. Police work is usually associated with high demand, low control and frequent contact with the public, job characteristics that are often referred to as ‘high-stress’ (Sauter & Murphy, 1995). As indicated by Collins and Gibbs (2003), police offi cers usually expose themselves to confrontation, violence,
traumatic incidents, human misery and even the possibility of being seriously injured or killed. In addition to physical danger, several organiza-tional factors also impose stress on them, such as shift work, work overload and organization structure (e.g. authoritarian) (Golembiewski & Kim, 1990). Studies on self-perceived occupa-tional stress for police offi cers also indicated that most stressors are related to organization and management factors, such as perceptions of staff shortages, inadequate resources, poor manage-ment, reorganization, bureaucratic interference, long work hours and social undermining (Biggam, Power, MacDonald, Carcary, & Moodie, 1997; Brown & Campbell, 1990; Collins & Gibbs,
2003; Duffy, Ganster, & Pagon, 2002; Kop, Euwema, & Schaufeli, 1999). Therefore, it is not surprising that previous studies have indicated that police offi cers have poor quality of life and higher depression rates (Chen et al., 2006) and that organization-related stressors can lead to burnout or other psychological illness among police offi cers (e.g. Collins & Gibbs, 2003; Gaines & Jermier, 1983; Gershon, Lin, & Li, 2002; Golembiewski & Kim, 1990).
However, in addition to the stressors coming from organizational factors, police offi cers also face role confl icts when doing their duties, another stressor in their police lives. As role theory posits, persons are members of social positions and hold expectations for their own behaviours and those of other persons (Biddle, 1986). It is particularly true for the position of police offi cer. Police offi cer is an important social position that corresponds to a set of expectations, such as maintaining public order, investigating crimes, serving the community, keeping the peace through enforce-ment of laws and societal norms, and ensuring the rights of all to liberty, equality and justice, and these duties can result in role confl ict for police offi cers (Regoli & Poole, 1980). Police offi -cers should be worthy of the public trust because they are people of integrity. Thus, they should behave in a professional manner, and their conduct must always be exemplary to maintain a position of respect in the community. However, the role of police offi cer may confl ict with other roles that a person plays. The purpose of this twofold study was aimed at discussing the role confl icts faced by police offi cers and investigating the relationship of role confl icts to burnout and health problems among police offi cers.
In the existing studies, work–family confl ict has been identifi ed as a role confl ict faced by police offi cers (Alexander & Walker, 1996; Burke, 1993, 1994, 1997; Burke & Deszca, 1986; Goodman, 1990). Because of the long working hours, shift work and cancelled leave, police offi -cers usually have diffi culty successfully playing the role of a husband/wife and father/mother simultaneously. This confl ict is a form of inter-role confl ict resulting from two incompatible roles from different positions (Van sell, Brief, & Schuler, 1981). For example, Alexander and Walker (1996) surveyed 409 spouses of police offi cers by questionnaire to assess the impact of police work on the family life. Their results showed that a quarter of participants claimed that their police partners were ‘considerably’ or
‘extremely’ stressed by their work. In addition, participants also reported that stress of police work impaired family relationships and other personal relationships outside the home. This stress also had a detrimental effect on marriage, spouses’ health, spouses’ relationship with chil-dren and on their spouses’ social life. This work– family confl ict causes not only problems within the family (e.g. higher divorce rate; Abdollahi, 2002) and unhealthy behaviours (e.g. drank; Alexander & Walker, 1996) but also burnout at work. In Burke’s (1993, 1994, 1997; Burke & Deszca, 1986) and Goodman’s (1990) studies, work–family confl ict has been identifi ed as a factor that had a signifi cant and positive effect in predicting burnout among police offi cers.
However, there is another role confl ict that was not identifi ed in past studies for police offi cers. This confl ict is associated with the dual image imposed on police offi cers: uncompromising and friendly. This confl ict is a form of intrasender role confl ict resulting from incompatible expectations from a single role sender (Van sell et al., 1981). Specifi cally, these two images come from different sources of expectations of police offi cers, i.e. the police and law system and public opinion. On the one hand, the expectation from the police and law system is that police offi cers maintain justice and, thus, they should be dedicated to enforcing laws and cannot let personal feelings, animosities or friendships infl uence offi cial conduct in carry-ing out their responsibilities. On the other hand, the expectation from the public is that police offi cers serve communities (see Fielding, 2005) and, thus, they should be friendly and maintain good relationships with people.
This dual expectation of police offi cers becomes especially more salient in Taiwanese society. Mon (2003) analysed interviews of 36 police offi cers and quotations from the director of the National Police Agency in Taiwan after 1987 (the year for lifting of martial law in Taiwan). The results showed that serving communities gradually has become an important police work in recent years. Moreover, Mon (2003) also conducted a survey of 40 police offi cers who were chief of a police district over four periods: 1987–1991 (10 per-sons), 1992–1996 (10 perper-sons), 1997–2001 (10 persons) and 2002–2003 (10 persons). In this survey, the 40 police offi cers were asked to iden-tify the three most important values of police work and the three most important tasks of police work during their periods of being the chief. The results showed that, prior to 1997, values and
tasks related to law enforcement were identifi ed as being more important than those related to community services. However, after 1997, values and tasks related to law enforcement and com-munity services were identifi ed as having the same level of importance. It can be seen that as Taiwan gradually became democratic after 1987 (e.g. repealing restrictions on political parties and the media and having direct legislative and presiden-tial elections), different roles were imposed upon police offi cers. This dual expectation of police offi cers imposed pressure on them as they attempted to play their roles, and police offi cers usually felt they were misunderstood by the public.
From the perspective of a job demands and resources model (JD-R model), the role confl icts faced by police offi cers can be regarded as job demands for them. Job demands are defi ned as ‘those physical, psychological, social, or organi-zational aspects of the job that require sustained physical and/or psychological (i.e., cognitive or emotional) effort and are therefore associated with certain physiological and/or psychological costs’ (Schaufeli & Bakker, 2004, p. 296). Accord-ing to this defi nition, role confl icts can be regarded as job demands that are imposed and caused by different role expectations. According to the JD-R model, when job demands require high effort, they may become stressors and elicit nega-tive consequences such as burnout (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001; Schaufeli & Bakker, 2004). Moreover, the JD-R model further proposed that burnout can lead to health problems such as psychosomatic health com-plaints including headaches, cardiovascular prob-lems and stomach aches (Schaufeli & Bakker, 2004). This phenomenon has been observed among police offi cers (e.g. Burke, 1994; Burke & Deszca, 1986; Stearns & Moore, 1993). This sequential process linking job demands with health problems via burnout was called the ener-getic process by Schaufeli and Bakker (2004), which was used to illustrate that high job demands exhaust employees’ energy back-up.
Hence, the purpose of this study was to inves-tigate the relationship between role confl icts, emotional exhaustion and health problems among police offi cers based on the energetic process proposed in the JD-R model. In the research model, two role confl icts—work–family con-fl ict and uncompromising–friendly concon-fl ict—were included. Emotional exhaustion was selected as the sole indicator for burnout because: (a) it is the
central or core dimension of burnout (Gaines & Jermier, 1983; Maslach & Jackson, 1981); and (b) job demands are closely associated with the emotional exhaustion component of burnout (Demerouti et al., 2001). Finally, health problems associated with psychosomatic health complaints were also included in the model. According to the JD-R model proposed by Schaufeli and Bakker (2004), it was hypothesized that role confl icts have a direct effect on emotional exhaustion, and emotional exhaustion has a direct effect on health problems. It is a complete mediation model that links role confl icts with health problems via emo-tional exhaustion. The theoretical research model is displayed in Figure 1.
Method
Participants and procedure
In this study, the survey data from Huang’s (1999) study were analysed. Using a proportional strati-fi ed sampling technique, we selected 1100 police offi cers from a total of 2700 police offi cers at fi ve branches of the Kaohsiung City Government Police Bureau in Taiwan based on the name list in the personnel offi ce at Kaohsiung City Govern-ment Police Bureau. Questionnaires were admin-istered and completed at the selected offi ces. A total of 867 police offi cers returned the survey (return rate = 78.8 per cent). For the current study, only 672 cases (654 of them were male and 18 participants did not report their sex) were used because they provided complete data on the variables that were used in the current study. The mean age of the sample was 35.06 [range from 24 to 59 years; standard deviation (SD) = 6.79].
Instruments
Items for role confl icts. There were four response items for work–family confl ict, includ-ing: (a) The requirements of my task usually lead
Role conflicts Emotional exhaustion Health problems
to complaints from my spouse or opposite-sex friends. (b) The long working hours usually deprive me of spending time with family. (c) My family usually worries about the risks associated with police work. (d) Because of work, I cannot relax at home, which often results in complaints from my family. Participants were asked to rate these items on a fi ve-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach’s alpha value for these four items was 0.82. The average score among these items was computed as the index for work–family con-fl ict. Further, there were three response items for uncompromising–friendly confl ict, including: (a) As a police offi cer, I often get confused in playing a role of an uncompromising lawman on the one hand, and playing a role of a friendly helper on the other hand. (b) Police offi cers would like to help the public but they are often misunderstood. I have no idea what I can do. (c) Police work should be uncompromised, but people do not respect police offi cers. Participants were asked to rate these items on a fi ve-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach’s alpha value for these three items was 0.72. The average score among these items was computed as the index for uncompromising–friendly confl ict.
Items for emotional exhaustion. There were fi ve symptoms for assessing emotional exhaustion at work: (a) lethargy; (b) lassitude; (c) feeling down; (d) languor; and (e) exhaustion. Partici-pants were asked to rate these items on a fi ve-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach’s alpha value for these four items was 0.87. The average score among these items was computed as the index for emotional exhaustion.
Items for health problems. There were fi ve symptoms for health problems, including: (a)
headache; (b) stomach ache; (c) sleep problems; (d) body aches; and (e) hypertension. Participants were asked to rate these items on a fi ve-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach’s alpha value for these four items was 0.85. The average score among these items was computed as the index for health problems.
Results
Descriptive statistics analysis
Table I presents descriptive statistics of the four research variables, including means, SDs and cor-relations. From the mean score it can be seen that police offi cers in the current study face serious role confl icts. The mean scores of the work–family and uncompromising–friendly confl icts were 4 points on a fi ve-point scale. The scores of emo-tional exhaustion and health problems exceeded the mid-point on a fi ve-point scale, also implying the presence of problems of emotional exhaustion and health.
Furthermore, all four of these variables were highly correlated. Firstly, the two kinds of con-fl icts were highly correlated (r = 0.75, p < 0.001). Secondly, both confl icts were also positively related to emotional exhaustion and health prob-lems (r = 0.50 to 0.57, ps < 0.001). Finally, emo-tional exhaustion and health problems were also highly correlated (r = 0.72, p < 0.001).
Structural equation modelling
The proposed model, depicted in Figure 1, was analysed using AMOS 7.0 (Chicago: SPSS) (Arbuckle, 2006) with a maximum likelihood estimation procedure. Because the two kinds of confl icts were highly correlated, a latent factor
Table I. Descriptive statistics of the research variables (n = 672).
Research variables Mean SD 1 2 3
Work–family confl ict 4.01 0.79 — — —
Uncompromising–friendly confl ict 4.03 0.76 0.75*** — —
Emotional exhaustion 3.44 0.90 0.57*** 0.50*** —
Health problems 3.59 0.85 0.56*** 0.51*** 0.72***
*** p < 0.001. SD: standard deviation.
called role confl ict was extracted to infl uence the two confl icts. Variance of the role confl ict factor was set as 1. The role confl ict factor had a direct effect on emotional exhaustion, and emotional exhaustion had a direct effect on health problems. Although all parameters estimated in the model were signifi cant, the χ2 value of this model was 56.4 with 2 degrees of freedom (df) (p < 0.05), rejecting the model. Other fi t indices were not satisfactory as well (Tucker Lewis Index (TLI) = 0.88, Comparative Fit Index (CFI) = 0.96, Root Mean Square Error of Approximation (RMSEA) = 0.20, 90 per cent confi dence interval (CI) RMSEA = 0.16–0.25, Root Mean Square Residual (RMR) = 0.45).
Furthermore, the direct effect of role confl ict factors on health problems was added to loosen the complete mediation hypothesis on emotional exhaustion. The result of the analysis is presented in Figure 2, in which the standardized estimates of path loadings are displayed. The χ2 value of this model was 0.3 with 1 df (p > 0.05), supporting the model. Other fi t indices were satisfactory as well (TLI = 1.00, CFI = 1.00, RMSEA = 0.00, 90 per cent CI RMSEA = 0.00–0.08, RMR = 0.01). This fi nding showed that emotional exhaustion did not have a complete mediation effect on the relation-ship between role confl ict and health problems.
Discussion
This study focused on the role confl icts faced by police offi cers in Taiwan and investigated the rela-tionship between role confl icts, emotional exhaus-tion and health problems based on the energetic process proposed in the JD-R model (Schaufeli & Bakker, 2004). In brief, correlation analysis showed that role confl icts had positive correlations to
emotional exhaustion and health problems. Emo-tional exhaustion also had a positive correlation to health problems. The result of the structural equa-tion model further suggested that the positive effect of role confl icts on health problems was partially mediated by emotional exhaustion.
There were two main contributions in this study. Firstly, previous studies on role confl ict of police offi cers usually focus on work–family confl ict (e.g. Alexander & Walker, 1996; Burke, 1993, 1994, 1997; Burke & Deszca, 1986; Goodman, 1990), an inter-role confl ict that results from the imposition of two incompatible roles on a single person (Van sell et al., 1981). However, there were other types of role confl icts faced by police offi cers. This study indicated an intrasender role confl ict that results from incompatible expectations for police offi cers (Van sell et al., 1981), that is, the dual image of being uncompromising and friendly. This intra-sender role confl ict has become more salient recently in Taiwan. From Mon’s (2003) analysis, it can be seen that as Taiwan became more democratic, dif-ferent tasks and expectations were imposed upon police offi cers, which made the intrasender role confl ict more serious for police offi cers in Taiwan. This intrasender role confl ict was not secondary to the work–family confl ict; as indicated in the current results, the mean scores of work–family and uncom-promising–friendly confl icts were 4 points on a fi ve-point scale, suggesting that uncompromising– friendly confl ict was as serious as the work–family confl ict faced by police offi cers in Taiwan. There-fore, the fi rst contribution of this study was to indi-cate another form of role confl ict faced by police offi cers and empirically examine its effect on emo-tional exhaustion and health problems.
However, the author also acknowledges a lim-itation in the generalizability of the current results to other nations because this study was only
Role conflicts W/F e1 0.91*** U/F e2 0.82*** Health problems d2 Emotional exhaustion d1 0.28*** 0.54*** 0.62***
Figure 2. The research model with standardized path coeffi cients. *** p < 0.001. U/F: uncompromising–friendly confl ict; W/F: work–family confl ict. d1 = disturbance of emotional exhaustion; d2 = disturbance of health problems; e1 = error of work-family confl ict; e2 = error of uncompromising-friendly confl ict.
conducted in Taiwan. In other nations, the intra-sender role confl ict (uncompromising–friendly confl ict) identifi ed in this study may not be a serious problem for police offi cers. However, police offi cers in other nations may face other role confl icts stemming from the values and expecta-tions of police offi cers in those naexpecta-tions. Thus, this study does not claim that the uncompromising– friendly confl ict is a stressor for all police offi cers in the world but rather would like to indicate that in addition to the work–family confl ict, there are other kinds of role confl icts faced by police offi cers. These role confl icts should not be neglected in investigating work stress among police offi cers.
Furthermore, the second contribution of this study was to examine the relationship between role confl icts, emotional exhaustion and health problems based on the energetic process in the JD-R model (Schaufeli & Bakker, 2004). On the one hand, the JD-R model provided a theoretical framework to explain the relationship between role confl icts, emotional exhaustion and health problems. On the other hand, this study also empirically examined the energetic process in the JD-R model on another occupational sample, the police offi cers. However, there were several dif-ferences between Schaufeli and Bakker’s (2004) work and the present study that should be addressed. According to Schaufeli and Bakker’s (2004) theoretical analysis and empirical fi ndings on four different occupational samples, it was hypothesized that the effect of job demands on health problems was fully mediated by burnout. However, in the current study, the effect of role confl icts (a form of job demands) on health prob-lems was not completely mediated by emotional exhaustion, the indicator of burnout. This alter-native fi nding may have resulted from the use of different measurements in the current study from those used in Schaufeli and Bakker’s (2004) study.
Specifi cally, the job demands measured by Schaufeli and Bakker (2004) were quantitative (i.e. workload) and qualitative (i.e. emotional) demands. However, the job demands measured in the current study were role confl icts faced by participants. In other words, the contents of job demands were different in these two studies. In addition, Schaufeli and Bakker (2004) used the Maslach Burnout Inventory-General Survey to assess burnout from three perspectives: exhaus-tion, cynicism and professional effi cacy. High scores on exhaustion and cynicism and low scores
on professional effi cacy are indicative of burnout. The current study only used emotional exhaus-tion as an indicator of burnout, not covering other components of burnout. Therefore, the dif-fering results may stem from different job demands and the lack of a comprehensive indicator of burnout. This was a limitation of this study testing the energetic process in the JD-R model.
Generally, this study focused on the role con-fl icts faced by police offi cers in Taiwan and identifi ed an intrasender role confl ict (uncom-promising–friendly confl ict) that was not investi-gated in previous studies of police offi cers. Furthermore, the energetic process in the JD-R model (Schaufeli & Bakker, 2004) was applied as a framework to investigate the relationship between role confl icts, emotional exhaustion and health problems. Although there are some limita-tions, this study does contribute to the body of police studies.
References
Abdollahi, M.K. (2002). Understanding police stress research.
Journal of Forensic Psychology practice, 2, 1–24.
Alexander, D.A., & Walker, L.G. (1996). The perceived impact of police work on police offi cers’ spouses and fam-ilies. Stress Medicine, 12, 239–246.
Arbuckle, J.L. 2006. Amos (Version 7.0) [Computer Program]. Chicago: SPSS.
Biddle, B.J. (1986). Recent development in role theory. Annual
Review of Sociology, 12, 67–92.
Biggam, F.H., Power, K.G., MacDonald, R.R., Carcary, W. B., & Moodie, E. (1997). Self-perceived occupational stress and distress in a Scottish police force. Work and Stress, 11, 118–133.
Brown, J.M., & Campbell, E.A. (1990). Source of occupa-tional stress in the police. Work and Stress, 4, 305– 318.
Burke, R.J. (1993). Work-family stress, confl ict, coping and burnout in police offi cers. Stress Medicine, 9, 171–180. Burke, R.J. (1994). Stressful events, work-family confl ict,
coping, psychological burnout, and well-being among police offi cers. Psychological Reports, 75, 787–800. Burke, R.J. (1997). Toward an understanding of
psychologi-cal burnout among police offi cers. International Journal of
Stress Management, 4, 3–27.
Burke, R.J., & Deszca, G. (1986).Correlates of psychological burnout phases among police offi cers. Human Relations,
39, 487–502.
Chen, H.C., Chou, F.H., Chen, M.C., Su, S.F., Wang, S.Y., Feng, W.W., Chen, P.C., Lai, J.Y., Chao, S.S., Yang, S.L., Tsai, T.C., Tsai, K.Y., Lin, K.S., Lee, C.Y., Wu, H.C. (2006). A survey of quality of life and depression for police offi cers in Kaohsiung, Taiwan. Quality of Life Research,
15, 925–932.
Collins, P.A., & Gibbs, A.C.C. (2003). Stress in police offi -cers: A study of the origins, prevalence and severity of stress-related symptoms within a county police force.
Demerouti, E., Bakker, A.B., Nachreiner, F., & Schaufeli, W.B. (2001). The job demands-resources model of burnout.
The Journal of Applied Psychology, 86, 499–512.
Duffy, M.K., Ganster, D.C., & Pagon, M. (2002). Social undermining in the workplace. Academy of Management
Journal, 45, 331–351.
Fielding, N.G. (2005). Concepts and theory in community policing. Howard Journal of Criminal Justice, 44, 460– 472.
Gaines, J., & Jermier, J.M. (1983). Emotional exhaustion in a high stress organization. Academy of Management
Journal, 26, 567–586.
Gershon, R.R., Lin, S., & Li, X.B. (2002). Work stress in aging police offi cers. Journal of Occupational and
Environ-mental Medicine, 44, 160–167.
Golembiewski, R.T., & Kim, B. (1990). Burnout in police work: stressors, strain, and the phase model. Police Studies,
14, 74–80.
Goodman, A.M. (1990). A model for police offi cer burnout.
Journal of Business and Psychology, 5, 85–99.
Huang, H.J. (1999). Role-Confl ict Problem of Policemen in
Taiwan, R.O.C. National Science Council project report
(NSC 88-2416-H-110-029).
Kop, N., Euwema, M.C., & Schaufeli, W.B. (1999). Burnout, job stress and violent behavior among Dutch police offi cers.
Work and Stress, 13, 326–340.
Maslach, C., & Jackson, S.E. (1981). The measurement of experienced burnout. Journal of Occupational Behaviour,
2, 99–113.
Mon, W.T. (2003). The police and democratic society—An empirical study on police role. The Chinese Public
Admin-istration Review, 12, 1–42.
Regoli, R., & Poole, E. (1980). Police professionalism and role confl ict: A comparison of rural and urban departments.
Human Relations, 33, 241–252.
Sauter, S.L., & Murphy, L.R. (1995). Organizational risk
factors for job stress. Washington, DC: American
Psycho-logical Association.
Schaufeli, W.B., Bakker, A.B. (2004). Job demands, job resources, and their relationship with burnout and engage-ment: A multi-sample study. Journal of Organizational
Behavior, 25, 293–315.
Stearns, G.M., & Moore, R.J. (1993). The physical and psy-chological correlates of job burnout in the Royal Canadian Mounted Police. Canadian Journal of Criminology, 35, 127–148.
Van sell, M., Brief, A.P., & Schuler, R.S. (1981). Role confl ict and role ambiguity: Integration of the literature and directions for future research. Human Relations, 34, 43–71.
