Journal of Nursing Management, 2014, 22, 159–169
Effects of perceived workplace politics in hospitals on nurses’ behavioural intentions in Ghana ROGER A. ATINGA B A , M P h i l 1, KWAME A. DOMFEH ABUOSI B A , M B A 1 and GLADYS DZANSI B A , M P h i l 4
MPA, PhD
2
, ESINAM KAYI
BA, MPhil
3
, AARON
1
Lecturer, 2Associate Professor, 3Research and Teaching Assistant, Department of Public Administration and Health Services Management, University of Ghana Business School, and 4Lecturer, School of Nursing, University of Ghana, Legon, Accra, Ghana
Correspondence Roger A. Atinga Department of Public Administration and Health Services Management University of Ghana Business School Accra Ghana E-mail: [email protected]
(2014) Journal of Nursing Management 22, 159–169. Effects of perceived workplace politics in hospitals on nurses’ behavioural intentions in Ghana
ATINGA R.A., DOMFEH K.A., KAYI E., ABUOSI A. & DZANSI G.
Aim To examine the effects of perceived workplace politics in hospitals on nurses’ job satisfaction, commitment, exit intention, job neglect, absenteeism and performance. Background One of the factors contributing to nurses’ poor advancement in clinical practice is the existence of petty politics, which has given rise to competing self-interest. However, little evidence exists to inform policy direction on the implication of politics on nurses’ behaviour. Method A total of 610 nurses comprising associate and nursing professionals completed a structured questionnaire modelled on workplace politics and its outcome variables. Descriptive statistics and mean comparisons were used to analyse data. A multivariate regression model was computed to examine association between perceived politics and nurses’ behavioural intentions. Result Perceived politics potentially leads to decline in job satisfaction, commitment and work performance. However, perceived workplace politics is associated with high intention to leave, negligent behaviour and absenteeism. Conclusion Measures aimed at improving nursing management and health-care delivery should be directed at minimising the use of politics to promote selfinterest. Implication for nursing management Evidence-based best practices in nursing management centred on the creation of an enabling environment for nurses to participate in decision-making should be given critical attention. Keywords: commitment, Ghana, job satisfaction, nurses, perceived politics, performance Accepted for publication: 4 August 2013
Introduction The organisational context in which nurses practice is important in explaining the nature and quality of care administered to patients (Aiken & Patrician 2000, Aiken et al. 2011). Nurses would prefer to work in a very conducive setting that enables them to provide DOI: 10.1111/jonm.12178 ª 2013 John Wiley & Sons Ltd
adequate health care to patients. Unfortunately, nurses in hospitals are often confronted with stress and burnout in the execution of their duties (Sagie & Krausz 2003). These stressful conditions are complicated by the poor interpersonal relationships in the nursing profession (Duddle & Boughton 2007), the high rate of bullying (Hutchinson et al. 2006, Katrinli et al. 159
R A Atinga et al.
2010) and the current infiltration of politics into the profession (Des Jardin 2001). Politics in the nursing profession arises from the fact that nurses represent the largest employee group in hospitals and health systems (World Health Organisation 2007) and this has generated incentives for competing interests and power play within the ranks of the profession. Globally, there is insufficient health workforce to address the health needs of populations, but nurses’ density per 1000 population appears to be low when compared with physicians, except in a few countries such as China, Bangladesh, Lebanon and Columbia (World Health Organisation 2007). In Ghana, nurses constitute about 55% of the total number of health workers directly providing health care to patients (Ministry of Health 2011). Institutional figures from primary, secondary and tertiary facilities also show that nurse numbers dominate all categories of health workers in the country. Nurses concentration in health facilities potentially increases the phenomenon of both vertical and horizontal politics where power and influence are used to determine ‘who gets what, when and how’ (Lasswell 1958). Organisational politics in this context delineates behaviour strategically designed to maximise the selfinterest of individuals or groups in the workplace (Vigoda-Gadot & Drory 2006, Chang et al. 2009), which conflicts with the collective organisational goals as well with the personal interest of other individuals (Vigoda-Gadot & Kapun 2005). It has also been conceptualised as the use of power and influence among individuals outside the formal organisational processes and procedures (Kurchner-Hawkins et al. 2006). Political behaviours in organisational settings is in part linked to the view that organisations are inherently political arenas insofar as individuals work as a group, each staking out claims to benefit at the expense of others (Drory & Vigoda-Gadot 2010, Gotsis & Kortezi 2010). Politics is thus a continuous process that takes place within the context of what an individual gets at a particular point in time and the means used to achieve that. Workplace politics has the potential of scaling back individual efforts. If, for example, the more qualified employee such as a nurse perceives that politics is used to cheat him or her out of a deserved opportunity, that nurse may exhibit negative reactions such as dissatisfaction, apathy, anxiety and turnover (Kacmar et al. 1999). Generally workplace politics gives rise to many reactions (Silvester 2008), including a predisposition to be dissatisfied with work, exhibit low level of commitment, neglect vital tasks or put up poor performance. Other studies have documented the detrimental effects 160
of workplace politics to include undesirable relationship among individuals, feelings of a sense of guilt, hatred, lack of credibility and absenteeism (Ferris et al. 1998, Vigoda 2000). Beyond these negative consequences, resignation or exit from the organisation could also be the end results of workplace politics. The higher the perception of politics in the workplace, the lower the level of justice, equity and fairness in the eyes of employees. Politics in the workplace is not new. Burgeoning literature suggests the proliferation and interest in workplace politics both actual and perceptual in public and private organisations (Huang et al. 2003, Vigoda-Gadot & Kapun 2005, Fedor et al. 2008). Despite the proliferation of research, the literature lacks comprehensive attempts at broadening our horizons on how politics manifests and distorts work processes in health care settings. Given the uniqueness of the health care industry and the pervasiveness of ethics in nursing, understanding the consequences of political behaviour among nurses is vital as such behaviours adversely affect hospital management (Ringer & Boss 2000). In addition, given that the nursing profession is built on the fundamental principles of individuality, dignity, uniqueness, human rights and responsibilities (Gallagher 2007), understanding the likely effect of workplace politics on these very principles is vital not only to inform policy direction, but also, to reduce the many pockets of tension in the nursing profession. Literature on workplace politics in Ghana is scarce and clearly lacks specific focus on health care settings. This is one of the first comprehensive researches that attempts to explore the influence of workplace politics on nurses’ behaviour in Ghanaian hospitals. Therefore findings of the study will enable nurse leaders in Ghana and elsewhere understand the social and political processes affecting nurses’ work-related attitudes and hence provide a workplace climate conducive for nurses to carry out their clinical tasks (Mok & Au-Yeung 2002). In this study, we examine workplace politics and its effect on different behavioural intentions of nurses with specific focus on job satisfaction, commitment to work, intention to leave, negligent behaviour, absenteeism and performance.
Materials and methods Study setting Data was collected in hospitals from two urban settings, Accra and Kumasi in the Greater Accra and ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Workplace politics in hospitals
Ashanti regions, respectively. We tailored the study towards urban centres because nurses in Ghana are concentrated in cities and towns to the disadvantage of the peri-urban and rural communities. More than 40% of nurses in Ghana work in urban boroughs, especially in Accra and Kumasi (Awofeso 2010). This has provided a platform against which politics is used as a weapon to advance the interest of some nurses.
Sample Hospitals were purposively selected based on their size, bed capacity and nursing staff numbers. We established a self-determined minimum bed capacity of 40 and nursing staff numbers of at least 100 against which to select health facilities. The purpose was to eliminate smaller health facilities with few nursing staff as anecdotal evidence suggests that workplace politics manifest strongly in large hospitals engaging the services of many nurses with mixed backgrounds. Information about hospitals bed capacity in Ghanaian hospitals is rarely available in a single report. Therefore, we used key words to search the Google/Ghana search engine for the characteristics of all hospitals, excluding health centres, clinics and health posts located within the two cities. The key words used were ‘bed’, ‘occupancy rate’, ‘size’, ‘structure’. Telephone calls were made to managers of facilities whose information was difficult to obtain online. Ultimately, all our online search results and telephone calls produced a total of 15 hospitals with bed capacities of at least 40. A list of the distribution of nurses in the 15 hospitals obtained from the Ghana Health Service revealed that 10 out of the 15 hospitals had nursing staff numbers consistent with our established criteria. The selected hospitals were classified into Level 1, Level 2 and Level 3. Level 1 comprised polyclinics that provide limited multilevel care to patients. Polyclinics usually have a maternity ward, they provide limited general inpatient services and have some specialised services such as ophthalmology and dental care. The total number of Level 1 health facilities in the sample was 3. Level 2 comprised hospitals that provide a full range of outpatient and inpatient services. These hospitals act as the first level of referrals for polyclinics, health centres and clinics. Their total number in the sample was 5. Level 3 consisted of teaching hospitals that provide multidisciplinary care. In Ghana, Level 3 hospitals are the final level of referrals in the health service delivery hierarchy because they are home to a mix of physicians with different ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
specialty areas. The total number of Level 3 hospitals in the sample was 2. Hospitals within the different classifications were stratified to facilitate the selection of sample nurses (You et al. 2013). There are two main classifications of nurses in Ghana: nursing professionals (registered general nurses) and nursing associate professionals (community health nurses and health assistants) (Ministry of Health 2011). The selection of nurses varied across the different hospital classifications. Nurses in Level 1 hospitals were sampled from the outpatient departments and all inpatient units, including male, children and maternity wards. A total of 48 nurses were sampled in each facility at Level 1. The sample scope of Level 2 facilities was wider. Nurses were selected from outpatient departments and at least four inpatient units including emergency, intensive care unit, medical and postoperative units. A total of 65 nurses were selected in each hospital classified under Level 2. On account of the size and scope of health-care delivery in Level 3 hospitals, we stratified them further into directorates, departments and units. Nurses were then selected in the departmental and unit strata. A total of 100 nurses from each hospital classified under Level 3 were selected. Detailed information about facility type and numbers of nurses selected within the units of the health facilities is given in Figure 1. Nurses in the hospitals of each classification were conveniently sampled. We compensated for this sampling strategy by selecting only nurses with at least 6 months’ working experience. This allowed for the selection of nurses with relatively good knowledge of management processes, work systems and organisational culture of the hospitals.
Questionnaire A wealth of literature demonstrates that perceived workplace politics is related to employee job satisfaction (Vigoda 2000, Witt et al. 2000), organisational commitment (Witt et al. 2000, Vigoda-Gadot & Kapun 2005), intention to leave, negligent behaviour and absenteeism (Vigoda 2001, Ram & Prabhakar 2010) as well as job performance (Kacmar et al. 1999, Vigoda 2000). Therefore, we developed a questionnaire for data collection based on these dimensions associated with workplace politics. Perception of workplace politics was measured by Vigoda’s (2000) nine-item scale. Nurses were asked to indicate the extent to which they agree or disagree with the existence of favouritism, preferential treatment, biased reward systems and unfair treatment in the 161
R A Atinga et al.
Number of facilities selected at each level
Composition of units and numbers of nurses selected in the health facilities OPD (n = 36)
Level 1 n=3
Inpatient units (n = 90)
Maternity (n = 18) OPD (n = 75)
Facility type
Level 2 n=5
All inpatient units (n = 175)
Emergency, ICU, ENT (n = 75) General OPD (n = 40) Level 3 n=2
All inpatient units (n = 80) All other departments (n = 80)
workplace. The scale ranged from 1 (strongly disagree) to 5 (strongly agree). For the purpose of multivariate regression, these measures were repeated in the questionnaire and the responses aggregated to binary to reflect the percentage of nurses that rated politics perceptions as either ‘high’ or ‘low’. Job satisfaction was measured using a nine-item scale developed and tested by earlier studies (Aiken et al. 2008, Choi et al. 2011). Nurses were asked to rate their level of job satisfaction with aspects of their job, including relationship with co-workers and supervisors, salary, opportunities for promotion and managerial fairness. A tenth measure gleaned from a nursing study by Kutney-Lee et al. (2013) sought to determine nurses’ overall satisfaction with their present jobs. We relied on the scale developed by Porter et al. (1974) to determine the measures of organisational commitment. These are: the desire to remain loyal to the health facility, willingness to accept values of the health facility and willingness to give of best efforts on behalf of the health facility. We also asked the nurses to indicate the extent to which they are proud to tell others that they are employees of the hospital where they work and whether they are committed to the performance of their hospitals. The measures of intention to leave were developed according to the scale defined by Farrell and Rusbult (1992). These include intention to quit the health facility and intention to consider job offers elsewhere. In addition to these, we asked the nurses whether they would consider applying for reposting. 162
Figure 1 Number of nurses selected within the units of each hospital type. OPD, outpatient department; ICU, intensive care unit; ENT, ear, nose, throat.
Job neglect or negligent behaviour refers to a situation in which an employee remains in the organisation but demonstrates dissatisfaction through low output or misconduct. For example, a nurse may be reluctant to respond promptly to emergencies or generally fail to show an appreciable level of ingenuity and innovation. Job neglect is usually used by employees who are handicapped and prefer to sanction the organisation for injustice (Farrell & Rusbult 1992). We used a four-item scale to measure job neglect. Nurses were asked for responses to statements such as: sometimes I put in less effort at work; sometimes I postpone important duties for an unlimited period of time; and personally I do not care so much about the progress of my health facility. Absenteeism arises when a nurse exhibits behavioural changes through frequent absence from work. It was measured by a three-item scale that asked nurses to indicate the extent to which they agree or disagree that they sometimes feel like not working because others do not see their efforts. Nurses were also asked to indicate the number of times they have been absent from work outside the normal leave days and the reason for such absence. The nurse self-rating performance index (Parker & Kulik 1995) was used to examine the degree of nurses’ performance on the job. We asked the nurses to personally evaluate their performance on the job in the areas of knowledge of clinical procedures, commitment to work, communication with co-workers, quality of care provided to patients and overall job performance. ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Workplace politics in hospitals
Control variables Both theoretical and empirical studies have demonstrated how the characteristics of individuals are related to organisational behaviour. Personal variables (e.g. age, sex, educational level, ethnicity, religion and rank) were therefore included to account for likely confounding relationship between perceptions of organisational politics and the various behavioural intentions.
Data collection Data was collected between September and November 2012. All nurses who participated in the study were assured confidentiality of the information provided. The instrument developed for the study was pilot tested among nine nurses working in the typology of Level 1, 2 and 3 health facilities located in Accra and Kumasi. The pilot study was helpful as nurses identified items within the performance, absenteeism and politics perception constructs that needed clarity to enhance understanding. In addition, during the pilot study, nurses took steps to reduce information overload associated with some of the items and measures. Considering the sensitive nature of the subject matter under investigation, nurses sampled in the actual study were told to complete the questionnaire either after working hours or at home and return them to the researchers the next day. Nurses who were unable to complete the questionnaire within 24 hours were given the opportunity to complete and submit it in a sealed envelope to a designated nurse in the health facility. To encourage frankness of opinions, participation was made voluntary. This was accomplished by including a caveat on the questionnaire instructing nurses to decide whether to participate or not (You et al. 2013). We offered no incentives for nurses who opted to participate. Bedside nurses of all shifts excluding nursing managers were given the opportunity to complete the questionnaire. A total of 610 out of the 669 sample nurses completed and returned the questionnaires. This represented a response rate of 91%.
tive statistics to account for the background characteristics of participating nurses using means and standard deviations (SD). We also compared the mean difference of workplace politics and its outcome variables between the different nursing professional groups. This allowed us to determine whether nurses with a higher perception of politics had greater fall-off in job satisfaction, organisational commitment and performance but high scores for intention to leave, negligent behaviour and absenteeism. We computed multivariate regression model to examine possible associations. Perceived politics in the hospital setting was the dependent variable. In the model, we sought to examine the nature of association between perceived politics and the covariates – job satisfaction, intention to leave, negligent behaviour, absenteeism and performance. We controlled for age, educational level, ethnicity, job tenure and rank in order to check spurious relationships. Based on the literature, we hypothesised that perceived politics will be positively related to intention to leave, negligent behaviour and absenteeism but negatively related to job satisfaction, job commitment and performance after accounting for the life-course variables, hence the expected signs in Table 1. To account for clustering of nurses in the health facilities, robust procedures were used to adjust for standard errors.
Results Majority of the nurses in Level 2 and 3 health facilities were registered general nurses compared with Level 1 which was dominated by community health nurses (Figure 2). This is not surprising given that community health nurses and health assistants are trained to provide first-level care. As shown in Table 2, many of the nurses were female (n = 333, 54.6%). The mean age of the respondents was 33 years. 30.0% of the nurses sampled completed nursing training schools, 32.1% graduated from community health nursing schools, 16.7% completed Table 1 Variables and their expected signs
Data analysis
Variable
Data was analysed with the aid of SPSS v16 (SPSS Inc., Chicago, IL, USA). We first present a histogram illustrating the distribution of community health nurses, health assistants and registered general nurses who participated in the study. We then computed descrip-
Job satisfaction Commitment to hospital Intention to leave Negligent behaviour Absenteeism Performance
ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Number of items 10 8 4 4 3 5
Expected sign
+ + +
163
R A Atinga et al.
Figure 2 Distribution of nurses by health facility type.
health assistant training schools and 21.1% graduated from universities. The nurses sampled possessed sufficient work experience (mean = 8.7) and were found to be mostly senior staff nurses (n = 194, 31.8%). In terms of ethnicity, majority were Akans (n = 242, 39.7%) followed by Ewes (31.8%) and Gas (14.6%), while all other tribes comprised the least number (n = 85, 13.9%).
As can be seeen in Table 3, health assistants had high mean scores (mean = 4.59, SD = 1.655) on politics perception. A similar view was shared by community health nurses. This suggests that nursing associate professionals were likely to be experiencing a distinct disadvantage compared with their counterparts within the nursing professional ranks and this was more likely to affect stability of their work in the future (McHugh et al. 2011). Job satisfaction was perceived to be fairly high for community health nurses (mean = 3.83, SD = 1.689) and registered general nurses (mean = 3.17, SD = 1.540) compared with health assistants (mean = 2.71, SD = 1.608). The results further demonstrate that community health nurses are more inclined to be committed to work than health assistants and registered nurses. Significant difference was also observed in turnover intentions among the nurses. The propensity to exit from the present work setting was fairly high for health assistants and registered nurses but low for community health nurses (mean = 2.55, SD = 1.451). Nearly all
Table 2 Nurses characteristics Characteristic Sex Male Female Female Age (years) 20–24 25–29 30–34 35–39 40–44 45+ Educational level Health assistant training Community health training Nursing training University Job tenure (years) 1–5 5–10 11–15 15+ Ethnicity Ewe Akan Ga Others Rank Staff nurse Senior staff nurse Nursing officer Senior nursing officer Principal nursing officer
n (%)
Mean (SD)
277 (45.4) 333 (54.6)
1.55 (0.498)
18 146 217 108 93 28
(3.0) (23.9) (35.6) (17.7) (15.2) (4.6)
102 196 183 129
(16.7) (32.1) (30.0) (21.1)
225 188 96 101
(36.9) (30.8) (15.7) (16.6)
194 242 89 85
(31.8) (39.7) (14.6) (13.9)
180 194 87 57 92
(29.5) (31.8) (14.3) (9.3) (15.1)
33 (5.793)
Median
Minimum
Maximum
32.0
21
49
8.7
1
25
2.49 (1.223)
2.12 (1.084)
2.85 (1.084)
2.49 (1.391)
Ranks of nursing professionals and nursing associate professionals were aggregated into a single composite unit. For example, community health staff nurse, health assistants staff nurse and registered general staff nurses were aggregated into ‘staff nurse’. The same applies to the other rank categories.
164
ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Workplace politics in hospitals
Table 3 Mean differences of perceived politics variables between the nursing professional groups Nursing professional group, mean (SD) Community health nurse Perception of politics Job satisfaction Commitment to work Intention to leave Negligent behaviour Absence from work Self-performance rating
3.64 3.83 3.38 2.55 2.46 2.46 3.30
Health assistant
(1.607) (1.689) (1.576) (1.451) (1.472) (1.644) (1.322)
4.59 2.71 2.47 3.00 1.82 2.65 3.41
(1.655) (1.608) (1.546) (1.335) (1.103) (1.419) (1.254)
Registered general nurse 3.11 3.17 2.23 3.07 2.15 2.30 3.42
(1.599) (1.540) (1.504) (1.579) (1.536) (1.498) (1.276)
All, mean (SD) 3.65 3.16 2.32 2.89 2.20 2.41 3.38
(1.609) (1.614) (1.535) (1.516) (1.465) (1.536) (1.286)
P 0.001 0.003 0.005 0.001 0.022 0.001 0.001
Job satisfaction was measured on a 5-point Likert scale that ranged from 1 (very unsatisfied) to 5 (satisfied). The scale for all other variables ranged from 1 (strongly disagree) to 5 (strongly agree).
Table 4 Multivariate regression results of perceived politics on nurses behavioural intentions Covariate Job satisfaction Organisational commitment Intention to leave Negligent behaviour Absenteeism Performance
Coefficient
Odds ratio (95% CI)
SE
P
0.365 0.407
0.694 (0.616–0.782) 0.665 (0.586–0.756)
0.061 0.065
Effects of perceived workplace politics in hospitals on nurses’ behavioural intentions in Ghana ROGER A. ATINGA B A , M P h i l 1, KWAME A. DOMFEH ABUOSI B A , M B A 1 and GLADYS DZANSI B A , M P h i l 4
MPA, PhD
2
, ESINAM KAYI
BA, MPhil
3
, AARON
1
Lecturer, 2Associate Professor, 3Research and Teaching Assistant, Department of Public Administration and Health Services Management, University of Ghana Business School, and 4Lecturer, School of Nursing, University of Ghana, Legon, Accra, Ghana
Correspondence Roger A. Atinga Department of Public Administration and Health Services Management University of Ghana Business School Accra Ghana E-mail: [email protected]
(2014) Journal of Nursing Management 22, 159–169. Effects of perceived workplace politics in hospitals on nurses’ behavioural intentions in Ghana
ATINGA R.A., DOMFEH K.A., KAYI E., ABUOSI A. & DZANSI G.
Aim To examine the effects of perceived workplace politics in hospitals on nurses’ job satisfaction, commitment, exit intention, job neglect, absenteeism and performance. Background One of the factors contributing to nurses’ poor advancement in clinical practice is the existence of petty politics, which has given rise to competing self-interest. However, little evidence exists to inform policy direction on the implication of politics on nurses’ behaviour. Method A total of 610 nurses comprising associate and nursing professionals completed a structured questionnaire modelled on workplace politics and its outcome variables. Descriptive statistics and mean comparisons were used to analyse data. A multivariate regression model was computed to examine association between perceived politics and nurses’ behavioural intentions. Result Perceived politics potentially leads to decline in job satisfaction, commitment and work performance. However, perceived workplace politics is associated with high intention to leave, negligent behaviour and absenteeism. Conclusion Measures aimed at improving nursing management and health-care delivery should be directed at minimising the use of politics to promote selfinterest. Implication for nursing management Evidence-based best practices in nursing management centred on the creation of an enabling environment for nurses to participate in decision-making should be given critical attention. Keywords: commitment, Ghana, job satisfaction, nurses, perceived politics, performance Accepted for publication: 4 August 2013
Introduction The organisational context in which nurses practice is important in explaining the nature and quality of care administered to patients (Aiken & Patrician 2000, Aiken et al. 2011). Nurses would prefer to work in a very conducive setting that enables them to provide DOI: 10.1111/jonm.12178 ª 2013 John Wiley & Sons Ltd
adequate health care to patients. Unfortunately, nurses in hospitals are often confronted with stress and burnout in the execution of their duties (Sagie & Krausz 2003). These stressful conditions are complicated by the poor interpersonal relationships in the nursing profession (Duddle & Boughton 2007), the high rate of bullying (Hutchinson et al. 2006, Katrinli et al. 159
R A Atinga et al.
2010) and the current infiltration of politics into the profession (Des Jardin 2001). Politics in the nursing profession arises from the fact that nurses represent the largest employee group in hospitals and health systems (World Health Organisation 2007) and this has generated incentives for competing interests and power play within the ranks of the profession. Globally, there is insufficient health workforce to address the health needs of populations, but nurses’ density per 1000 population appears to be low when compared with physicians, except in a few countries such as China, Bangladesh, Lebanon and Columbia (World Health Organisation 2007). In Ghana, nurses constitute about 55% of the total number of health workers directly providing health care to patients (Ministry of Health 2011). Institutional figures from primary, secondary and tertiary facilities also show that nurse numbers dominate all categories of health workers in the country. Nurses concentration in health facilities potentially increases the phenomenon of both vertical and horizontal politics where power and influence are used to determine ‘who gets what, when and how’ (Lasswell 1958). Organisational politics in this context delineates behaviour strategically designed to maximise the selfinterest of individuals or groups in the workplace (Vigoda-Gadot & Drory 2006, Chang et al. 2009), which conflicts with the collective organisational goals as well with the personal interest of other individuals (Vigoda-Gadot & Kapun 2005). It has also been conceptualised as the use of power and influence among individuals outside the formal organisational processes and procedures (Kurchner-Hawkins et al. 2006). Political behaviours in organisational settings is in part linked to the view that organisations are inherently political arenas insofar as individuals work as a group, each staking out claims to benefit at the expense of others (Drory & Vigoda-Gadot 2010, Gotsis & Kortezi 2010). Politics is thus a continuous process that takes place within the context of what an individual gets at a particular point in time and the means used to achieve that. Workplace politics has the potential of scaling back individual efforts. If, for example, the more qualified employee such as a nurse perceives that politics is used to cheat him or her out of a deserved opportunity, that nurse may exhibit negative reactions such as dissatisfaction, apathy, anxiety and turnover (Kacmar et al. 1999). Generally workplace politics gives rise to many reactions (Silvester 2008), including a predisposition to be dissatisfied with work, exhibit low level of commitment, neglect vital tasks or put up poor performance. Other studies have documented the detrimental effects 160
of workplace politics to include undesirable relationship among individuals, feelings of a sense of guilt, hatred, lack of credibility and absenteeism (Ferris et al. 1998, Vigoda 2000). Beyond these negative consequences, resignation or exit from the organisation could also be the end results of workplace politics. The higher the perception of politics in the workplace, the lower the level of justice, equity and fairness in the eyes of employees. Politics in the workplace is not new. Burgeoning literature suggests the proliferation and interest in workplace politics both actual and perceptual in public and private organisations (Huang et al. 2003, Vigoda-Gadot & Kapun 2005, Fedor et al. 2008). Despite the proliferation of research, the literature lacks comprehensive attempts at broadening our horizons on how politics manifests and distorts work processes in health care settings. Given the uniqueness of the health care industry and the pervasiveness of ethics in nursing, understanding the consequences of political behaviour among nurses is vital as such behaviours adversely affect hospital management (Ringer & Boss 2000). In addition, given that the nursing profession is built on the fundamental principles of individuality, dignity, uniqueness, human rights and responsibilities (Gallagher 2007), understanding the likely effect of workplace politics on these very principles is vital not only to inform policy direction, but also, to reduce the many pockets of tension in the nursing profession. Literature on workplace politics in Ghana is scarce and clearly lacks specific focus on health care settings. This is one of the first comprehensive researches that attempts to explore the influence of workplace politics on nurses’ behaviour in Ghanaian hospitals. Therefore findings of the study will enable nurse leaders in Ghana and elsewhere understand the social and political processes affecting nurses’ work-related attitudes and hence provide a workplace climate conducive for nurses to carry out their clinical tasks (Mok & Au-Yeung 2002). In this study, we examine workplace politics and its effect on different behavioural intentions of nurses with specific focus on job satisfaction, commitment to work, intention to leave, negligent behaviour, absenteeism and performance.
Materials and methods Study setting Data was collected in hospitals from two urban settings, Accra and Kumasi in the Greater Accra and ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Workplace politics in hospitals
Ashanti regions, respectively. We tailored the study towards urban centres because nurses in Ghana are concentrated in cities and towns to the disadvantage of the peri-urban and rural communities. More than 40% of nurses in Ghana work in urban boroughs, especially in Accra and Kumasi (Awofeso 2010). This has provided a platform against which politics is used as a weapon to advance the interest of some nurses.
Sample Hospitals were purposively selected based on their size, bed capacity and nursing staff numbers. We established a self-determined minimum bed capacity of 40 and nursing staff numbers of at least 100 against which to select health facilities. The purpose was to eliminate smaller health facilities with few nursing staff as anecdotal evidence suggests that workplace politics manifest strongly in large hospitals engaging the services of many nurses with mixed backgrounds. Information about hospitals bed capacity in Ghanaian hospitals is rarely available in a single report. Therefore, we used key words to search the Google/Ghana search engine for the characteristics of all hospitals, excluding health centres, clinics and health posts located within the two cities. The key words used were ‘bed’, ‘occupancy rate’, ‘size’, ‘structure’. Telephone calls were made to managers of facilities whose information was difficult to obtain online. Ultimately, all our online search results and telephone calls produced a total of 15 hospitals with bed capacities of at least 40. A list of the distribution of nurses in the 15 hospitals obtained from the Ghana Health Service revealed that 10 out of the 15 hospitals had nursing staff numbers consistent with our established criteria. The selected hospitals were classified into Level 1, Level 2 and Level 3. Level 1 comprised polyclinics that provide limited multilevel care to patients. Polyclinics usually have a maternity ward, they provide limited general inpatient services and have some specialised services such as ophthalmology and dental care. The total number of Level 1 health facilities in the sample was 3. Level 2 comprised hospitals that provide a full range of outpatient and inpatient services. These hospitals act as the first level of referrals for polyclinics, health centres and clinics. Their total number in the sample was 5. Level 3 consisted of teaching hospitals that provide multidisciplinary care. In Ghana, Level 3 hospitals are the final level of referrals in the health service delivery hierarchy because they are home to a mix of physicians with different ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
specialty areas. The total number of Level 3 hospitals in the sample was 2. Hospitals within the different classifications were stratified to facilitate the selection of sample nurses (You et al. 2013). There are two main classifications of nurses in Ghana: nursing professionals (registered general nurses) and nursing associate professionals (community health nurses and health assistants) (Ministry of Health 2011). The selection of nurses varied across the different hospital classifications. Nurses in Level 1 hospitals were sampled from the outpatient departments and all inpatient units, including male, children and maternity wards. A total of 48 nurses were sampled in each facility at Level 1. The sample scope of Level 2 facilities was wider. Nurses were selected from outpatient departments and at least four inpatient units including emergency, intensive care unit, medical and postoperative units. A total of 65 nurses were selected in each hospital classified under Level 2. On account of the size and scope of health-care delivery in Level 3 hospitals, we stratified them further into directorates, departments and units. Nurses were then selected in the departmental and unit strata. A total of 100 nurses from each hospital classified under Level 3 were selected. Detailed information about facility type and numbers of nurses selected within the units of the health facilities is given in Figure 1. Nurses in the hospitals of each classification were conveniently sampled. We compensated for this sampling strategy by selecting only nurses with at least 6 months’ working experience. This allowed for the selection of nurses with relatively good knowledge of management processes, work systems and organisational culture of the hospitals.
Questionnaire A wealth of literature demonstrates that perceived workplace politics is related to employee job satisfaction (Vigoda 2000, Witt et al. 2000), organisational commitment (Witt et al. 2000, Vigoda-Gadot & Kapun 2005), intention to leave, negligent behaviour and absenteeism (Vigoda 2001, Ram & Prabhakar 2010) as well as job performance (Kacmar et al. 1999, Vigoda 2000). Therefore, we developed a questionnaire for data collection based on these dimensions associated with workplace politics. Perception of workplace politics was measured by Vigoda’s (2000) nine-item scale. Nurses were asked to indicate the extent to which they agree or disagree with the existence of favouritism, preferential treatment, biased reward systems and unfair treatment in the 161
R A Atinga et al.
Number of facilities selected at each level
Composition of units and numbers of nurses selected in the health facilities OPD (n = 36)
Level 1 n=3
Inpatient units (n = 90)
Maternity (n = 18) OPD (n = 75)
Facility type
Level 2 n=5
All inpatient units (n = 175)
Emergency, ICU, ENT (n = 75) General OPD (n = 40) Level 3 n=2
All inpatient units (n = 80) All other departments (n = 80)
workplace. The scale ranged from 1 (strongly disagree) to 5 (strongly agree). For the purpose of multivariate regression, these measures were repeated in the questionnaire and the responses aggregated to binary to reflect the percentage of nurses that rated politics perceptions as either ‘high’ or ‘low’. Job satisfaction was measured using a nine-item scale developed and tested by earlier studies (Aiken et al. 2008, Choi et al. 2011). Nurses were asked to rate their level of job satisfaction with aspects of their job, including relationship with co-workers and supervisors, salary, opportunities for promotion and managerial fairness. A tenth measure gleaned from a nursing study by Kutney-Lee et al. (2013) sought to determine nurses’ overall satisfaction with their present jobs. We relied on the scale developed by Porter et al. (1974) to determine the measures of organisational commitment. These are: the desire to remain loyal to the health facility, willingness to accept values of the health facility and willingness to give of best efforts on behalf of the health facility. We also asked the nurses to indicate the extent to which they are proud to tell others that they are employees of the hospital where they work and whether they are committed to the performance of their hospitals. The measures of intention to leave were developed according to the scale defined by Farrell and Rusbult (1992). These include intention to quit the health facility and intention to consider job offers elsewhere. In addition to these, we asked the nurses whether they would consider applying for reposting. 162
Figure 1 Number of nurses selected within the units of each hospital type. OPD, outpatient department; ICU, intensive care unit; ENT, ear, nose, throat.
Job neglect or negligent behaviour refers to a situation in which an employee remains in the organisation but demonstrates dissatisfaction through low output or misconduct. For example, a nurse may be reluctant to respond promptly to emergencies or generally fail to show an appreciable level of ingenuity and innovation. Job neglect is usually used by employees who are handicapped and prefer to sanction the organisation for injustice (Farrell & Rusbult 1992). We used a four-item scale to measure job neglect. Nurses were asked for responses to statements such as: sometimes I put in less effort at work; sometimes I postpone important duties for an unlimited period of time; and personally I do not care so much about the progress of my health facility. Absenteeism arises when a nurse exhibits behavioural changes through frequent absence from work. It was measured by a three-item scale that asked nurses to indicate the extent to which they agree or disagree that they sometimes feel like not working because others do not see their efforts. Nurses were also asked to indicate the number of times they have been absent from work outside the normal leave days and the reason for such absence. The nurse self-rating performance index (Parker & Kulik 1995) was used to examine the degree of nurses’ performance on the job. We asked the nurses to personally evaluate their performance on the job in the areas of knowledge of clinical procedures, commitment to work, communication with co-workers, quality of care provided to patients and overall job performance. ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Workplace politics in hospitals
Control variables Both theoretical and empirical studies have demonstrated how the characteristics of individuals are related to organisational behaviour. Personal variables (e.g. age, sex, educational level, ethnicity, religion and rank) were therefore included to account for likely confounding relationship between perceptions of organisational politics and the various behavioural intentions.
Data collection Data was collected between September and November 2012. All nurses who participated in the study were assured confidentiality of the information provided. The instrument developed for the study was pilot tested among nine nurses working in the typology of Level 1, 2 and 3 health facilities located in Accra and Kumasi. The pilot study was helpful as nurses identified items within the performance, absenteeism and politics perception constructs that needed clarity to enhance understanding. In addition, during the pilot study, nurses took steps to reduce information overload associated with some of the items and measures. Considering the sensitive nature of the subject matter under investigation, nurses sampled in the actual study were told to complete the questionnaire either after working hours or at home and return them to the researchers the next day. Nurses who were unable to complete the questionnaire within 24 hours were given the opportunity to complete and submit it in a sealed envelope to a designated nurse in the health facility. To encourage frankness of opinions, participation was made voluntary. This was accomplished by including a caveat on the questionnaire instructing nurses to decide whether to participate or not (You et al. 2013). We offered no incentives for nurses who opted to participate. Bedside nurses of all shifts excluding nursing managers were given the opportunity to complete the questionnaire. A total of 610 out of the 669 sample nurses completed and returned the questionnaires. This represented a response rate of 91%.
tive statistics to account for the background characteristics of participating nurses using means and standard deviations (SD). We also compared the mean difference of workplace politics and its outcome variables between the different nursing professional groups. This allowed us to determine whether nurses with a higher perception of politics had greater fall-off in job satisfaction, organisational commitment and performance but high scores for intention to leave, negligent behaviour and absenteeism. We computed multivariate regression model to examine possible associations. Perceived politics in the hospital setting was the dependent variable. In the model, we sought to examine the nature of association between perceived politics and the covariates – job satisfaction, intention to leave, negligent behaviour, absenteeism and performance. We controlled for age, educational level, ethnicity, job tenure and rank in order to check spurious relationships. Based on the literature, we hypothesised that perceived politics will be positively related to intention to leave, negligent behaviour and absenteeism but negatively related to job satisfaction, job commitment and performance after accounting for the life-course variables, hence the expected signs in Table 1. To account for clustering of nurses in the health facilities, robust procedures were used to adjust for standard errors.
Results Majority of the nurses in Level 2 and 3 health facilities were registered general nurses compared with Level 1 which was dominated by community health nurses (Figure 2). This is not surprising given that community health nurses and health assistants are trained to provide first-level care. As shown in Table 2, many of the nurses were female (n = 333, 54.6%). The mean age of the respondents was 33 years. 30.0% of the nurses sampled completed nursing training schools, 32.1% graduated from community health nursing schools, 16.7% completed Table 1 Variables and their expected signs
Data analysis
Variable
Data was analysed with the aid of SPSS v16 (SPSS Inc., Chicago, IL, USA). We first present a histogram illustrating the distribution of community health nurses, health assistants and registered general nurses who participated in the study. We then computed descrip-
Job satisfaction Commitment to hospital Intention to leave Negligent behaviour Absenteeism Performance
ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Number of items 10 8 4 4 3 5
Expected sign
+ + +
163
R A Atinga et al.
Figure 2 Distribution of nurses by health facility type.
health assistant training schools and 21.1% graduated from universities. The nurses sampled possessed sufficient work experience (mean = 8.7) and were found to be mostly senior staff nurses (n = 194, 31.8%). In terms of ethnicity, majority were Akans (n = 242, 39.7%) followed by Ewes (31.8%) and Gas (14.6%), while all other tribes comprised the least number (n = 85, 13.9%).
As can be seeen in Table 3, health assistants had high mean scores (mean = 4.59, SD = 1.655) on politics perception. A similar view was shared by community health nurses. This suggests that nursing associate professionals were likely to be experiencing a distinct disadvantage compared with their counterparts within the nursing professional ranks and this was more likely to affect stability of their work in the future (McHugh et al. 2011). Job satisfaction was perceived to be fairly high for community health nurses (mean = 3.83, SD = 1.689) and registered general nurses (mean = 3.17, SD = 1.540) compared with health assistants (mean = 2.71, SD = 1.608). The results further demonstrate that community health nurses are more inclined to be committed to work than health assistants and registered nurses. Significant difference was also observed in turnover intentions among the nurses. The propensity to exit from the present work setting was fairly high for health assistants and registered nurses but low for community health nurses (mean = 2.55, SD = 1.451). Nearly all
Table 2 Nurses characteristics Characteristic Sex Male Female Female Age (years) 20–24 25–29 30–34 35–39 40–44 45+ Educational level Health assistant training Community health training Nursing training University Job tenure (years) 1–5 5–10 11–15 15+ Ethnicity Ewe Akan Ga Others Rank Staff nurse Senior staff nurse Nursing officer Senior nursing officer Principal nursing officer
n (%)
Mean (SD)
277 (45.4) 333 (54.6)
1.55 (0.498)
18 146 217 108 93 28
(3.0) (23.9) (35.6) (17.7) (15.2) (4.6)
102 196 183 129
(16.7) (32.1) (30.0) (21.1)
225 188 96 101
(36.9) (30.8) (15.7) (16.6)
194 242 89 85
(31.8) (39.7) (14.6) (13.9)
180 194 87 57 92
(29.5) (31.8) (14.3) (9.3) (15.1)
33 (5.793)
Median
Minimum
Maximum
32.0
21
49
8.7
1
25
2.49 (1.223)
2.12 (1.084)
2.85 (1.084)
2.49 (1.391)
Ranks of nursing professionals and nursing associate professionals were aggregated into a single composite unit. For example, community health staff nurse, health assistants staff nurse and registered general staff nurses were aggregated into ‘staff nurse’. The same applies to the other rank categories.
164
ª 2013 John Wiley & Sons Ltd Journal of Nursing Management, 2014, 22, 159–169
Workplace politics in hospitals
Table 3 Mean differences of perceived politics variables between the nursing professional groups Nursing professional group, mean (SD) Community health nurse Perception of politics Job satisfaction Commitment to work Intention to leave Negligent behaviour Absence from work Self-performance rating
3.64 3.83 3.38 2.55 2.46 2.46 3.30
Health assistant
(1.607) (1.689) (1.576) (1.451) (1.472) (1.644) (1.322)
4.59 2.71 2.47 3.00 1.82 2.65 3.41
(1.655) (1.608) (1.546) (1.335) (1.103) (1.419) (1.254)
Registered general nurse 3.11 3.17 2.23 3.07 2.15 2.30 3.42
(1.599) (1.540) (1.504) (1.579) (1.536) (1.498) (1.276)
All, mean (SD) 3.65 3.16 2.32 2.89 2.20 2.41 3.38
(1.609) (1.614) (1.535) (1.516) (1.465) (1.536) (1.286)
P 0.001 0.003 0.005 0.001 0.022 0.001 0.001
Job satisfaction was measured on a 5-point Likert scale that ranged from 1 (very unsatisfied) to 5 (satisfied). The scale for all other variables ranged from 1 (strongly disagree) to 5 (strongly agree).
Table 4 Multivariate regression results of perceived politics on nurses behavioural intentions Covariate Job satisfaction Organisational commitment Intention to leave Negligent behaviour Absenteeism Performance
Coefficient
Odds ratio (95% CI)
SE
P
0.365 0.407
0.694 (0.616–0.782) 0.665 (0.586–0.756)
0.061 0.065
