Work-Related Well-Being in the Zimbabwean Banking Sector: A Job Demands-Resources Perspective.

Background: Research on the impact of psychosocial risks on well-being at work remains scarce in low- and middle-income countries, especially in the banking sector. This study sought to examine the relationships between job demands, job resources, well-being, job satisfaction, and work engagement in the Zimbabwean banking sector.
Methods: An online survey was administered to 259 employees from five banks. Hierarchical multiple regression tested the relationships between job demands (quantitative demands, emotional demands, work pace, and work-family conflict), job resources (possibilities for development, social support from colleagues and supervisors, quality of leadership, and influence at work), well-being, work engagement, and job satisfaction. Interactions between all variables were tested.
Results: Job demands were negatively related to well-being, work engagement, and job satisfaction. Job resources had positive relationships with the same. Work pace had positive relationships with well-being and work engagement. Influence at work moderated the relationship between emotional demands and work pace with well-being. Possibilities for development moderated the relationship between work-family conflict and well-being. Work-family conflict moderated the relationship between social support from colleagues and job satisfaction. Emotional demands, work pace, and quantitative demands moderated the relationship between influence at work with job satisfaction and work engagement.
Conclusion: Job demands should be reduced where possible in order to enhance employee well-being, work engagement, and job satisfaction. The job resources that should be availed to facilitate a positive psychosocial work environment in the banking sector include social support from supervisors, influence at work, and possibilities for development.

Introduction

Psychosocial factors are important antecedents of job-related well-being. They are related to both negative facets (e.g., work-related stress) and positive facets of well-being at work (e.g., job satisfaction and work engagement) [1,2]. Research on psychosocial factors and well-being at work remains scarce in low- and middle-income countries (e.g., [3,4]). Most available studies have focused on teachers and lecturers reporting a high level of work-related stress. Common stressors identified in the education sector included high workload, meeting targets, and long working hours [5]. A more recent study found that management employees reported high levels of occupational stress across sectors in Zimbabwe [6].

Both Javaid et al. (2018) and Kortum and Leka (2014) have highlighted the need for a comprehensive perspective, underpinned by theory to address psychosocial hazards, and well-being at work in low- and middle-income countries [3,4]. Furthermore, research on psychosocial factors and their effects on well-being may be better understood using a sectorial approach since different sectors are exposed to different working conditions [3]. For example, mining equipment, ambient conditions, work demands, and control were found to be significant predictors of quality of life and general well-being in the mining industry in Ghana [7]. On the other hand, some of the major stressors that were found in the police force in Trinidad and Tobago included obstacles to effective policing, power, resources of offenders, and threat of corruption. Their responses to these stressors included aggressive policing and avoidance of some duties [8]. High job demands like workload and work-life imbalance resulted in burnout which negatively affected employee well-being in pharmaceutical companies in Karachi, Pakistan [9]. This suggests that different sectors may have varying levels and types of psychosocial risks [3]. This study aims to contribute to sectorial research by focusing on the banking sector in Zimbabwe.

The banking sector of Zimbabwe has grown and includes 19 banking institutions, comprised of 13 commercial banks, five building societies, and one savings bank [10]. While some improvements were observed in the sector's stability, issues such as liquidity, foreign currency shortages, and trade policy uncertainty still represent concerns [11]. Nevertheless, the sector has a very important role to play in economic development. For example, it facilitates trade and movement of funds from savers to investors which enhance gross domestic product growth [12].

The intermediation function is facilitated when banks are profitable, making the banks' performance critical for economic growth [12], which is even more important in a low- and middle-income country context. Therefore, the well-being of bank employees must be prioritized to enable them to achieve their organizational objectives. Furthermore, several changes have affected the banking sector in Zimbabwe recently [13]. These include new information communication technologies, including Internet banking and artificial intelligence [13]. The increase in mobile and Internet banking is convenient and enhances financial inclusion, but bank employees are under pressure to cope with new and advanced technology [14]. Additionally, the economic environment forced some banks into cost cutting, resulting in retrenchments and potentially increasing workload for the remaining staff [15]. These changes influence working conditions and may result in psychosocial risks which can cause stress and affect work outcomes [16]. These observations support the case for determining those psychosocial work environment factors that are related to job-related well-being of bank employees in Zimbabwe, as they may face both unique and similar demands with other bank employees globally [17].

Therefore, using the job demands-resources model (JD-R) [18] as the theoretical framework, this study aims to explore the relationship between psychosocial factors, well-being, job satisfaction, and work engagement in the banking sector of Zimbabwe. In addition, interactions of job demands and job resources were investigated to enhance understanding of how they influence these outcomes. The study responds to the call for more research on the psychosocial work environment in low- and middle-income countries given the current gaps that have been highlighted [1] and seeks to provide further insight into the relationship between psychosocial factors and various facets of well-being at work.

The JD-R model divides psychosocial factors into two broad categories of job demands and job resources [18]. Job demands refer to the physical, organizational, psychological, and social characteristics of a job that require sustained physical and/or psychological effort which are associated with some physiological and psychological costs [18]. Job resources are the physical, social, psychological, or organizational aspects of a job that assist in goal achievement and reduce job demands and their associated costs. According to the model, job demands may buffer the positive effects of job resources, and job resources may buffer the negative effects of job demands [19]. Bakker and Demerouti (2018) argue that studies should include job demands and resources that make sense in the particular work context [19].

The job demands variables included in this study were quantitative demands, emotional demands, work pace, and work–family conflict. These demands are more likely to be relevant in the banking sector, which has faced considerable changes globally [17]. For example, the pace of work has increased in order to meet targets, and this has been associated with increased stress levels in bank employees [20]. Additionally, there are potential issues to do with work–family conflict since employees may work long hours in order to meet production targets, especially following an economic crisis [21]. The job resources that were chosen for the current study were quality of leadership, social support from supervisors, social support from colleagues, possibilities for development, and influence at work. For example, lack of support from superiors was associated with stress in bank employees in Quetta city [22].

The evidence regarding interactions between job demands and job resources with well-being and engagement is still weak [23]. There is still need to investigate and appreciate these relationships and contribute to the empirical evidence regarding interactions of job resources and demands with outcomes, especially in low- and middle-income county contexts.

Based on these earlier studies, the following hypotheses were formulated:

Job demands are negatively related to employee well-being.

Job demands are negatively related to work engagement.

Job demands are negatively related to job satisfaction.

Job resources are positively related to employee well-being.

Job resources are positively related to work engagement.

Job resources are positively related to job satisfaction.

The relationship between job demands and employee well-being will be weaker when job resources are high rather than low.

The relationship between job resources and organizational health outcomes (job satisfaction and work engagement) will be weaker when job demands are high rather than low.

Materials and methods

Participants and procedure

Ethical approval was granted by the Social Research Ethics Committee of the University conducting this research. We approached 10 out of 19 banking institutions in order to get a balance between the 13 commercial banks, five building societies, and one savings bank [10]. In addition, we wanted to obtain representation from both international and local commercial banks as these may have different cultures and resources [24]. The final participation depended on the positive response received from the banks. A letter was sent to the human resources managers of the ten banks explaining the research purpose, ethical considerations, and seeking permission to conduct the study in their organization. Information about the study was disseminated to all staff in five banks with 259 employees providing their consent and completing the online questionnaire (response rate of 35.53%). A sample of 145 employees was considered to be a minimum for this study. This was based on a medium effect size of .30 and power of 80% and a significance level of .005 [25]. The sample size was calculated using the formular 50+8 m where m is the number of predictors in the current study [26]. Therefore, the sample of 259 bank staff that was obtained was deemed sufficient for analysis.

Measures

Demographic characteristics

The questionnaire included questions on gender, age, educational qualifications, and tenure. Age was measured in years with five ranges from less than 30 up to 60 years. Gender was measured categorically as one male and two females. Tenure was measured in years and months as a continuous variable. Highest educational qualification responses ranged from the General Certificate of Education; ordinary level-1 to PhD-7 as shown in Table 1. Tenure was measured in years and months as a continuous variable.

Table 1

Sample demographic characteristics

Variable	Frequency	Percentage
Gender
Min-max	1-2
Mean (SD)	1.4 (.491)
Female	102	39.4
Male	154	59.5
Missing	3	1.2
Age
Min-max	1-4
Mean (SD)	2.4 (819)
Under 30	34	13.1
30-39	108	41.7
40-49	94	36.6
50-59	21	8.1
	2	.8
Educational qualifications
Min-max	2-8
Mean (SD)	5.97 (1.242)
Advanced level	3	1.2
Professional certificate	16	6.2
National diploma	20	7.7
Higher national diploma	13	5.0
Undergraduate degree	106	40.9
Master's degree	96	37.1
PhD	4	1.5
	1	.4
Tenure
Min-max	1-7
Mean (SD)	2.60 (1.258)
0-5	64	24.7
6-10	47	18.1
11-15	104	40.2
16-20	26	10.0
21-25	11	4.2
26-30	5	1.9
Above 30	2	.8

Job demands and job resources

The Copenhagen Psychosocial questionnaire (COPSOQ 11), medium version for professionals was used to assess job demands and job resources [27]. The majority of questions have five response alternatives from “always” to “hardly ever” or from “to a very large extent” to “to a very small extent”. The scoring is from 0–100. In terms of job demands, an example item for quantitative demands is “Do you get behind with your work?”. For job resources, an example item for possibilities for development is “Do you have the possibility of learning new things through your work?”. The reliability coefficients (Cronbach's alphas) are shown in Table 1.

Employee well-being

The General Well-Being Questionnaire (GWBQ) worn-out scale was used to measure employee well-being, which has been used as an indicator of symptoms of work-related stress [28]. Responses range from “never” to “all the time” on a five-point Likert scale ranging from 0–4 with total scores from 0 to 48. An example of the items is, “Over the last six months, how often have you become easily annoyed or irritated?”. The Cronbach's alpha for the scale was .92.

Work engagement

Work engagement was assessed using the Utrecht Work Engagement Scale (UWES-9) [29]. An example of the items is “When I get up in the morning, I feel like going to work”. Likert-type response options were from never (0) to always (6). The scale is made up of three parts measuring dedication, absorption, and vigor. A reliability of .89 was obtained in this study (see Table 1).

Job satisfaction

Job satisfaction was measured using a single item, “How satisfied are you with your job?”. A single-item measure was used considering the length of the study questionnaire. Previous research has supported the validity of this single-item measure [30]. Responses range from 1 “not at all satisfied” to 5 “extremely satisfied”. Higher scores indicate more satisfaction.

Data analysis

The initial phase of analysis involved principal component analysis, confirmatory factor analysis, and descriptive statistics. Secondly, hierarchical multiple regression was conducted to test the hypotheses using the IBM SPSS statistical package, version 24. Confirmatory factor analysis was conducted using AMOS version 24.

Principal component analysis

All items loaded on the expected components but the quality of leadership and social support from supervisors' items loaded on component 1. The two scales touch on some aspects of leadership and we decided to use them as originally intended. Unexpectedly, two factors were extracted for the UWES-9 scale. All vigor items loaded on component 1, but dedication items 2 and 3 and absorption items 1 and 2 loaded on component 2 together with the remaining items. Given these results, the work engagement scale was used as a total scale in this study [29].

Hierarchical multiple regression

Hierarchical linear multiple regression was used because participants were drawn from different banks and there were several predictors. Furthermore, the hypotheses were based on the job demands-resources theory; therefore, the hierarchical structure was predetermined by theory, previous research, and the nature of the study questions [31]. We controlled for demographic variables as they could potentially confound the dependent variables in Model 1. Model 2 is comprised of demographic variables and individual job demands. Model 3 is comprised of demographic variables, individual job demands, and individual job resources. It was important to assess the relationship between all study variables with aspects of well-being and determine their unique contributions, rendering methods like the stepwise inappropriate as they statistically determine variables that should remain in the model [32].

Moderation analyses were performed by hierarchical multiple regression using SPSS 24 with PROCESS version 3 [31]. The predictor and moderator variables were first centered on the mean because it makes interpretation easier [31]. The product term which depicts the interaction was obtained by multiplying the moderator and the predictor, i.e., individual job resources and individual job demands. The interaction term must be significant in order to determine a moderation effect [33]. We used one standard deviation (SD) below the mean as the low level, mean as the average level, and one SD above the mean as the high level and the Johnson-Neyman significance regions to determine significance levels [31].

We tested whether job resources individually moderate the effect of individual job demands on employee well-being and whether job demands individually moderate the relationship between individual job resources, job satisfaction, and work engagement. Each criterion had 20 interactions tested; hence, only significant results are reported.

Results

Descriptive statistics

Most participants were male (60%) and between 30 and 39 years of age. The majority of participants had an undergraduate degree (41%) and between 11 and 15-years’ tenure. Table 1 presents descriptive statistics of the participants' demographic characteristics, while Table 2 presents means and standard deviations of the study variables. Table 3 presents Cronbach's alphas and correlations between all study variables.

Table 2

Means and standard deviations of variables

Variables	No of items	Mean	Standard Deviation
Quantitative demands	4	40.77	19.44
Emotional demands	4	48.77	20.53
Work pace	3	65.64	19.05
Work–family conflict	4	47.97	27.23
Possibilities for development	4	69.55	21.86
Influence	4	41.74	21.23
Quality of leadership	4	57.17	23.65
Social support from supervisors	3	67.13	22.99
Social support from colleagues	3	65.03	19.11
Job satisfaction	1	3.32	.89
Work engagement	9	5.09	.99
General well-being	12	28.00	7.98

Table 3

Cronbach's alpha and correlations

Variables	1	2	3	4	5	6	7	8	9	10	11	12
1. Well-being	1
2. Emotional demands	.370∗∗	(.69)
3. Influence	-.141∗	.139∗	(.72)
4. Quantitative demands	.433∗∗	.436∗∗	-.152∗	(.73)
5. Work pace	.141∗	.402∗∗	.060	.324∗∗	(.65)
6. Possibilities for development	-.177∗∗	.035	.385∗∗	-.098	.131∗	(.87)
7. Colleague support	-338∗∗	-.111	.207∗∗	-.264∗∗	.030	-.261∗∗	(.65)
8. Supervisor support	-.301∗∗	-.138∗	.233∗∗	-.278∗∗	-.021	.333∗∗	.614∗∗	(.82)
9. Quality of Leadership	-.229∗∗	-.097	.286∗∗	-.286∗∗	.014	.513∗∗	.490∗∗	.712∗∗	(.85)
10. Work–family conflict	.555∗∗	.466∗∗	-.006	.501∗∗	.342∗∗	-.036	-.224∗∗	-.178∗∗	-.194∗∗	(.87)
11. Work engagement	-.438∗∗	-.059	.353∗∗	-.306∗∗	.061	.520∗∗	.327∗∗	.424∗∗	.449∗∗	-.241∗∗	(.88)
12. Job satisfaction	-.386∗∗	-.262∗∗	.266∗∗	-.351∗∗	-.021	.542∗∗	.366∗∗	.438∗∗	.576∗∗	-.300∗∗	.646∗∗

∗p < .05. ∗∗p < .01. Cronbach's alphas are presented in parentheses diagonally.

Hypothesis testing

Results from the hierarchical regression analysis are shown in Table 4. As anticipated, job demands variables had positive relationships with poor employee well-being except for work pace which had a negative relationship with poor well-being (β = .155, p < .01). Social support from supervisors was the only job resource that was significantly negatively related to employee well-being (β = - .163, p < .05).

Table 4

Hierarchical regression results: Job demands, job resources, employee well-being, work engagement, and job satisfaction

	Job satisfaction		Employee well-being		Work engagement
Predictors	ΔR2	Β	ΔR2	Β	ΔR2	Β
Model1	.024		.022		0.35
Education		-.108		.061		-.130∗
Gender		- .013		.121		-.005
Tenure		.079		-.001		.042∗
Age		.071		-.069		.134
Model 2	.187∗		.373∗		.136∗
Education		-.058		-.007		-.097
Gender		-.070		.170∗∗		-.057
Tenure		.050		.051		-.051
Age		-.078		-.082		.094
Quantitative demands		-.282∗		.221∗		-.310∗∗
Work–family conflict		-.168		.424∗		-.168∗
Work pace		.173∗∗		-.151∗∗		.202∗∗
Emotional demands		-.143		.171∗∗		.065
Model 3	.301∗		.050∗		.259∗
Education		-.011		-.017		-.084
Gender		-.077		.155∗∗		-.050
Tenure		.050		.073		.027
Age		.086		-.081		.085
Quantitative demands		-.119∗		.153∗		-.162∗
Work–family conflict		-.096		.409∗		-.147∗
Work pace		.083		-.111∗		.112∗∗
Emotional demands		-.198∗∗		.171∗∗		.025
Possibilities for development		.293∗		-.083∗		.338∗
Leadership quality		.354∗		.113∗		.029
Social support from supervisors		-.003		-.105		.213∗∗
Influence at work		.030		-.063		.134∗
Social support from colleagues		.022-		-.105		-.027

∗p < 0.001. ∗∗p < 0.01. ∗p < 0.05.

As expected, quantitative demands and emotional demands had significant negative relationships with job satisfaction, (β = -.119, p < .05) and (β = -.198, p < .01), respectively, while possibilities for development and quality of leadership were positively associated with job satisfaction, (β = .293, p < .001), (β = .354, p < .001), respectively.

Job demands had significant negative relationships with work engagement except for emotional demands which were not significantly related to work engagement. Interestingly, work pace had a positive relationship with work engagement, (β = .112, p < .05). Job resources were positively related to work engagement except for quality of leadership and social support from colleagues which were not significantly related to work engagement.

Interaction results

Interactions between job demands, job resources, and employee well-being

Influence at work significantly moderated the relationship between emotional demands and work pace with employee well-being [t (259) = -2.8589, p = .0046, ΔR = .0256] and [t (259) = -2.317, p = .0213, ΔR = .0198], respectively. Emotional demands explained an additional 2.56% of the variance in employee well-being while work pace explained 1.98% of variance. Possibilities for development moderated the relationship between work–family conflict and employee well-being [t (259) = 2.514, p = .0125, ΔR = .0162] explaining 1.62% of variance.

Interactions between job demands, job resources, and job satisfaction

The interaction term between social support from colleagues and work–family conflict was significant [t (256) = 2.191, p = .0294, ΔR = .0153], explaining an additional 1.53% of variance in job satisfaction. Furthermore, emotional demands moderated the relationship between influence at work and job satisfaction, [t (257) = 3.1358, p = .0019, ΔR = .0313], explaining an additional 3.13% of variance. The interaction of quantitative demands and influence at work was significant [t (257) = 2.758, p = .0062 ΔR = .0242], explaining an additional 2.42% of variance in job satisfaction. Work pace moderated the relationship between influence at work and job satisfaction, [t (257) = 2.275, p = .0238 ΔR = .0186], explaining an additional 1.86% of variance. Interestingly, the highest levels of job satisfaction were experienced when both work pace and influence at work were high. The lowest level of job satisfaction occurred when work pace was high and influence at work was low.

Interactions between job demands, job resources, and work engagement

The interaction term between emotional demands and influence at work was significant [t (256) = 2.235, p = 0263, ΔR = .0168], explaining an additional 1.68% of variance in work engagement. The interaction of quantitative demands and influence at work was significant [t (256) = 3.286, p = .0012, ΔR = .0333]. It explained an additional 3.33% of the variance in work engagement. The interaction between influence at work and work pace was significant [t (256) = 2.3842, p = .0179, ΔR = .0193], explaining an additional 1.93% of variance. The highest levels of work engagement were experienced when both work pace and influence at work were high.

Discussion

The hypotheses that job demands will be negatively related to job satisfaction, while job resources will be positively related to job satisfaction were partially supported. Unexpectedly, work pace and work–family conflict were not significantly related to job satisfaction. In contrast, Diana et al. (2020) found a negative relationship between work–family conflict and job satisfaction at a bank in Indonesia [34]. Bank employees are susceptible to emotional demands due to their constant interaction with customers [35]. They are not expected to express their emotions but to maintain professionalism at all times. Possibilities for development and quality of leadership had significant relationships with job satisfaction. This is consistent with previous research, for example, transformational leadership positively influenced job satisfaction in some commercial banks in Ghana [36]. Furthermore, bank employees seem to value development opportunities as the nature of their jobs requires continuous learning in order to keep up with constant changes and new challenges [36]. Training enhanced job satisfaction together with engagement and motivation in the banking sector of Tangail, Bangladesh [37]. Quality of leadership may also influence the availability of opportunities for staff development.

This research supports the health impairment process since job demands were the most important predictors of employee well-being. This might explain why most job resources did not significantly influence employee well-being. There is very limited research on the relationship between job demands and resources on well-being in the banking sector in low- and middle-income countries [4]. However, others have investigated the relationship between only job demands and stress. For example, demands including workload and long working hours were associated with stress in some banks in Ghana [38]. Continuous exposure to job demands without sufficient job resources will eventually affect health through various mechanisms, for example, work-related stress [39]. Social support from supervisors was the only job resource that influenced well-being positively, consistent with previous research [22].

As expected, job demands had negative relationships with work engagement except for emotional demands which did not have a significant relationship with work engagement. In contrast, others reported a negative relationship between work engagement and emotional demands [40]. Unexpectedly, work pace had a positive relationship with work engagement suggesting that it may have been perceived as a challenge demand.

Social support from supervisors and influence at work had positive relationships with work engagement. Similarly, a study of six large banks in Pakistan found both supervisor and colleague support to be important resources for enhancing work engagement [41]. It is important to note that the direct relationships between most of the study variables have been established mostly in developed countries, but the same knowledge is sparse in African countries like Zimbabwe, hence, the investigations in this study.

Interactions

Influence at work could have moderated the impact of emotional demands and work pace on well-being because it involves control. When one has control, they can work up to a level that they are comfortable with beyond which they can negotiate a reduction of the workload or pace. Lack of control and supervisory support have been found to negatively affect employee health [16,21]. Consistent with previous studies, this study highlighted that high job resources and low job demands result in better well-being and that an increase in work–family conflict is bad for well-being [42].

Influence at work and social support from colleagues positively influenced job satisfaction consistently with other research [43]. Social support from colleagues may enhance job satisfaction by providing assistance on tasks [44]. Job satisfaction was highest when both work pace and influence at work were high, rendering support for the boosting hypothesis. Work pace might have been perceived as a challenging job demand which positively influenced job satisfaction. A challenging job has been associated with job satisfaction [45]. Creating a good work-life balance enhances job satisfaction [46].

The results imply that reducing job demands may increase job satisfaction in some cases depending on the particular job demand or job resource. The results also show that when work pace was low and influence at work was high job satisfaction was not at its best, so reducing job demands may not work in all situations, sometimes, it may reduce the challenge in the job [47].

Work engagement was highest when influence at work was high and emotional demands or work pace were high. Work engagement was also highest when influence at work was high and quantitative demands were low. Similarly, Kenyi and John (2020) found that job resources were positively related to work engagement while job demands had a negative influence on the relationship between job resources and work engagement [48]. Increasing structural job resources and decreasing hindering job demands did not predict work engagement, but increasing social job resources and increasing challenging job demands positively influenced work engagement [47]. Furthermore, a negative relationship between emotional demands and work engagement has been reported [40], suggesting that other factors may be at play, which may need further investigation.

Limitations and suggestions for future studies

The use of a cross-sectional design does not establish causal relationships, and future studies may consider using longitudinal studies to determine the nature of the relationships over time [16]. Self-report measures that were used are prone to common method biases like socially desirable responding, which may threaten reliability and validity [7]. Self-report measures were considered because of the nature of the research purpose, which involved exploring individual perceptions. In addition, other factors may influence psychosocial factors and health. For example, personal resources like self-efficacy enable workers to cope with job demands effectively and can be included in future studies [19].

Study implications

This study provided additional support for the validity of the JD-R theory in an African country. It supported that prolonged exposure to work demands without job resources influences health negatively. Work–family conflict had the strongest relationship with employee well-being. This has implications in terms of work design and organization. Policies that prevent work–family conflict and consider flexible working where possible can be developed. Some job demands should be reduced where feasible to promote well-being. However, this may not be practical in some situations. In those circumstances, relevant job resources should be increased, for example, influence at work can moderate certain job demands as discussed.

The results support the importance of investing in staff development. Furthermore, interventions can be designed around providing influence at work which can enhance job satisfaction and work engagement when working at a high work pace in banks. It is important to determine the specific job resources that work well with certain job demands. It is also important to acknowledge that some demands, like emotional demands from customers, can be difficult to reduce, and appropriate resources should be provided in these cases. It is hoped that occupational psychologists, human resource managers, and policy-makers can take advantage of these results to enhance employee attitudes and well-being through action and that further research will build on this study.

This research contributed to the scarce literature on the relationship between psychosocial risks, employee, and job well-being in low- and middle-income countries and has shed light on important job demands and resources and their interactions to consider when developing interventions in the banking sector.

Funding

No Funding.

Conflicts of interest

None.