Social well-being, job satisfaction, organisational citizenship behaviour and intentions to leave in a utility organisation

social elements of work, and the organisation in which employees are their well-being, and retention secure Social well-being can be linked to the self-determination theory Orientation: Employee social well-being is likely to influence individual and organisational outcomes, especially in African countries where a high premium is often placed on one’s personhood being rooted in one’s relations with others. Research purpose: This study investigated the associations between social well-being, job satisfaction, organisational citizenship behaviour and intentions to leave in a South African utility organisation. Motivation for the study: Given the history of relationships amongst diverse people in South Africa, social well-being seems to be a critical component of the overall well-being of employees. However, few studies in South Africa have focused on social well-being in organisational contexts. Research approach/design and method: A cross-sectional survey design was used, targeting permanent employees in a South African utility organisation. Consenting participants ( N = 403) completed previously validated measures of social well-being, job satisfaction, organisational citizenship behaviour and intentions to leave. Structural equation modelling was performed to test hypotheses. Main findings: Social well-being was positively associated with job satisfaction and organisational citizenship behaviour and negatively associated with intentions to leave. Social well-being indirectly affected organisational citizenship behaviour and intention to leave through job (dis)satisfaction. Practical/managerial implications: Managers and human resources practitioners are alerted to practical ways of sustaining employees’ social well-being such as by implementing tailor-made policies that support social aspects of well-being and by ensuring the alignment of well-being programmes with changing circumstances in the modern world of work. Originality/value-add: This study illuminated social well-being associations with selected outcomes in a developing African country workplace context. Social acceptance (the acceptance of the diversity of colleagues), actualisation (the belief in one’s organisation, team and colleagues’ potential), coherence (the belief that one’s organisation and social relations at work are both meaningful and comprehensible), contribution (the belief that one’s daily work tasks add value to one’s team, department, and organisation) and integration (the belief that one experiences a sense of communal connectedness and belongingness).


Introduction
Workplace well-being is increasingly becoming critical for both individuals and organisations (Abe, Fields, & Abe, 2016;Moller & Rothmann, 2019). According to the International Labour Organization, the concept of workplace well-being refers to a wide range of aspects of working life, including the quality and safety of the working environment, how employees view their workplaces, the climate at work and how they are organised (ILO, 2017). Workplace well-being is essential because it has an impact on individual and organisational outcomes (Kowalski, Loretto, & Redman, 2017;Litchfield, Cooper, Hancock, & Watt, 2016). Notably, a range of variables referred to in the definition of workplace well-being include social elements of work, and the organisation in which employees are embedded affects their well-being, satisfaction, performance and retention (Keyes, 1998;Redelinghuys, Rothmann, & Botha, 2019). (Deci & Ryan, 2011), with its conceptualisation of relatedness as a psychological need of people. Relatedness refers to individuals' need to be loved and cared for by others and to feel connected with others. It is satisfied when individuals develop close relationships with others and feel a sense of communion with them (Deci & Ryan, 2011). The best relationships are those in which true relatedness and mutuality are experienced.
However, only a few studies (e.g. Moller & Rothmann, 2019;Redelinghuys et al., 2019) have been carried out using a multidimensional model of social well-being in South African work contexts. Using the MHC (Keyes, 2005) to measure emotional, psychological and social well-being, Moller and Rothmann (2019) showed that a large sample of managers scored the lowest on social well-being in four different latent well-being profiles. More specifically, they found that social actualisation (i.e. the feeling that the world is becoming a better place to function in) and social coherence (grasping how societies work) were relatively low compared with other elements of well-being. Moreover, research on personality functioning (Nel et al., 2011) suggested that the socialrelational aspects of personality were strongly present in 11 language groups in South Africa.
A number of Southern African studies have focused on social aspects of work-related well-being. For example, Janik and Rothmann (2015) investigated the effects of co-worker and manager relations in organisations. Redelinghuys et al. (2019) investigated the validity of a measure of workplace flourishing (which included social well-being). However, the latter study focused on the three components of well-being rather than exclusively on social well-being. Moreover, outcomes of interest were either different or not combined in the same context as proposed in this study (i.e. a utility organisation in South Africa). Therefore, more research is needed to understand social well-being in work and organisational settings in South Africa (Keyes, 1998;Rothmann, 2013).
Individuals' social well-being relies on positive interpersonal connections to accomplish individual and organisational work goals (Rosales, 2016). Different models of employeremployee relationships confirm that positive social environments in the workplace are beneficial to sustain employee well-being (Daniels, Watson, & Gedikli, 2017;Keyes, 1998;Rosales, 2016). Research by Deci and Ryan (2011) revealed that social context (e.g. feedback, communication and rewards) influenced the satisfaction of individuals' psychological needs (including relatedness). Therefore, the social context will affect employees' social well-being. For this reason, social well-being will likely influence individual and organisational outcomes. It is, thus, of particular interest to human resource practitioners and work and organisational psychologists to develop a deeper understanding of the linkages between social well-being and individual and organisational outcomes, mainly because these linkages are under-researched in developing country contexts and, specifically, in the utility industry (Daniels et al., 2017).
The utility industry is instrumental in the development and maintenance of the South African economy. However, the utility in the study, which supplies 95% of electricity to South Africa, is plagued with skills loss and has encountered daunting challenges, especially since 2005, evident in electricity supply shortages, frequent load-shedding, financial instability (including allegations of state capture), structural problems, poor operational performance, a deteriorating reputation and the loss of many valuable employees with critical skills (Govender, 2017;Ratshomo & Nembahe, 2019). The well-being and morale of employees are affected by the environment in which they find themselves. Masilela (2018) found that social well-being and stress in the utility could be linked to social relationships in the entity. Stress and low morale of employees affected their job satisfaction, organisational citizenship behaviour and retention (Masilela, 2018). Estimations of employee turnover showed that the utility remained at risk of losing valuable human resources (I. Venter, personal communication, April 7, 2019). The utility industry had lost 2110 employees since 2019 (R. Mey, personal communication, September 21, 2021). These statistics make it evident that the utility organisation should investigate employees' social well-being and the association thereof with job satisfaction, organisational citizenship behaviour and intention to leave (Keim, Landis, Pierce, & Earnest, 2014).
Organisations cannot afford to ignore the importance of social well-being and more research on this aspect is needed (Boreham, Povey, & Tomaszewski, 2016;Daniels et al., 2017;De Simone, 2014;Gandy, Harrison, & Gold, 2018). For instance, De Simone (2014) highlighted the need to study social well-being and its spillover effects on critical outcomes such as job satisfaction, organisational citizenship behaviour and intention to leave. Negative social experiences affect individuals' emotions and functioning, such as satisfaction with work conditions, decision-making quality and inclinations to withdraw. However, social experiences also affect relations with colleagues and management, affecting individual and organisational performance. Moreover, the five-factor model of social well-being (Keyes, 1998;Rothmann et al., 2019) has not been well-researched in organisationsparticularly in non-Western cultural contexts. Given that social embeddedness is considered vital to an African sociocultural orientation, even more so than individual personhood (Khumalo, Ejoke, Oppong Asante, & Rugira, 2021), the lack of research on social well-being is surprising. Organisations' neglect of social well-being signals ignorance, affecting employees' job satisfaction and performance and retaining valued human resources (Gandy et al., 2018).

Social well-being
Social well-being forms an integral part of one's health and concerns the degree to which individuals function well in their social lives, including work relationships and the ability to function well alongside colleagues (Keyes, 1998). According to Keyes (1998), social well-being encompasses five theoretically substantiated dimensions. In a workplace context, these can be translated to mean the following: Open Access Social acceptance (the acceptance of the diversity of colleagues), actualisation (the belief in one's organisation, team and colleagues' potential), coherence (the belief that one's organisation and social relations at work are both meaningful and comprehensible), contribution (the belief that one's daily work tasks add value to one's team, department, and organisation) and integration (the belief that one experiences a sense of communal connectedness and belongingness). (Redelinghuys et al., 2019, p. 2) Previous research by Janik and Rothmann (2015) indicated that the acceptance of employees by their supervisors and co-workers had a significant influence on their work lives (Janik & Rothmann, 2015). On the one hand, where positive feelings and close social ties exist between colleagues, a sense of belonging and high levels of relatedness are promoted. Those employees are more likely to remain with the organisation and are willing to go the extra mile to support and help their colleagues because of the social well-being experienced . On the other hand, when employees think that their work conditions are unfair, compensation for work is inadequate or they lack training or developmental and career advancement opportunities, their growth is stifled and as a result they will regress and languish (Rothmann, 2014).

Job satisfaction
Job satisfaction refers to the degree to which individuals feel positive or negative towards the aspects of their jobs (Alegre, Mas-Machuca, & Berbegal-Mirabent, 2016;Cek & Eyupoglu, 2020). The work environment and relationships with coworkers affect job satisfaction and the incapacity to maintain social relations is often associated with low job satisfaction (Moller & Rothmann, 2019). When employees perceive work as meaningful and their work conditions (such as their terms of remuneration and equal access to benefits) and treatment by their superiors as fair, they are more likely to feel connected to others and experience higher levels of job satisfaction (Redelinghuys et al., 2019). Work relationships with supervisors and colleagues and identification with the objectives and goals of the organisation are considered main aspects that can influence job satisfaction (Alegre et al., 2016).
Social exchange theory (Blau, 1964, as cited in Cek & Eyupoglu, 2020) supports the notion that employees who are satisfied with their jobs may feel inclined to reciprocate by demonstrating positive attitudes and behaviours towards others, which enhances co-operation and performance. The opposite may also be true: whereas a supportive work environment stimulates generosity within employees that propels businesses into prosperity, an unsatisfactory and unsupportive work environment may cause employees to feel psychologically unsafe, to experience less job satisfaction, and to become withdrawn, so that they are less likely to be psychologically available to others or to extend themselves beyond what is legally required (Cek & Eyupoglu, 2020;Janik & Rothmann, 2015).

Organisational citizenship behaviour
Organisational citizenship behaviour is a term used to describe constructive employee behaviours and actions that are not mandatory, but discretionary, behaviours performed voluntarily to support and benefit the organisation and employees (Cek & Eyupoglu, 2020;Thiruvenkadam & Yabesh, 2017). According to Smith, Organ and Near (1983), organisational citizenship behaviour can be defined as a type of contextual performance. It refers to actions taken to help others in the organisation or demonstrate conscientious behaviour. Research has shown two types of organisational citizenship behaviour (Organ, 1988): (1) altruism or helping others and (2) generalised compliance or following the rules and procedures of the organisation. Organisational citizenship behaviour enhances effectiveness, efficiency, adaptability to changing circumstances and organisational competitiveness (Cek & Eyupoglu, 2020). Help-oriented behaviours also increase commitment and prompt intentions of employees to stay with the organisation (Thiruvenkadam & Yabesh, 2017).
When employing a multidimensional measure of well-being, Redelinghuys et al. (2019) found well-being to be significantly related to organisational citizenship behaviour. Mukherjee (2020) also demonstrated that subjectively experienced wellbeing was significantly related to citizenship behaviours when directed at the organisation, whereas workplace wellbeing was positively associated with citizenship behaviours directed towards teammates and the organisation as an entity.

Intention to leave
Intention to leave signals the attitudinal readiness to withdraw in favour of seeking alternative employment and is the best indicator of actual future turnover (Janse van Rensburg, Rothmann, & Diedericks, 2017). When employees experience insufficient well-being, their performance will plummet and they are likely to quit their jobs (Redelinghuys et al., 2019). Harmonious relationships with supervisors and co-workers serve as a buffer against intention to leave, whereas meaningless work roles and unsupportive colleagues will increase employees' intention to leave (Janik & Rothmann, 2015;Janse van Rensburg et al., 2017). Govender (2017) cites several sources (e.g. Shmailan, 2016;Yadav & Aspal, 2014) that empirically confirmed that job satisfaction is directly and statistically significantly related to individual productivity and employee retention. According to Masilela (2018), low job satisfaction reduces organisational citizenship behaviour, weakens social ties and leads to turnover. Studies (Crede, Chernyshenko, Stark, Dalal, & Bashshur, 2007;Murphy, Athanasou, & King, 2002;Zito et al., 2018) found associations between job satisfaction, organisational citizenship behaviour and intention to leave. Therefore, it is critical to investigate which factors elevate job satisfaction and organisational citizenship behaviour, as failure to do so translates into more intentions to leave (Keim et al., 2014). De Neve, Krekel and Ward (2018) showed that social relationships had a significant and sizeable effect on job satisfaction. From 12 domains of workplace quality, social relationships explained the largest part of the variation in job satisfaction. Cek and Eyupoglu (2020) concluded that talent retention, job satisfaction and willingness to demonstrate citizenship behaviour should be of primary concern to organisations because of the effects on productivity and competitiveness. Therefore, a more in-depth understanding of inter relationships amongst these variables is needed.

Social well-being and individual and organisational outcomes
Positive relations with supervisors and social well-being of employees promote a sense of psychological safety, social connectivity and higher levels of relatedness and belonging, which spill over into higher levels of job satisfaction, willingness to engage in organisational citizenship behaviour and remain in the job (Janik & Rothmann, 2015;Redelinghuys et al., 2019).
Based on the description of the research problem and the review of previous studies, the following hypotheses were formulated: Hypothesis 1 (H 1 ): Social well-being is positively associated with job satisfaction.

Hypothesis 2 (H 2 ):
Social well-being is positively associated with organisational citizenship behaviour.

Hypothesis 3 (H 3 ):
Social well-being is inversely related to intention to leave.

Hypothesis 4 (H 4 ):
Job satisfaction mediates the relationship between social well-being and organisational citizenship behaviour.

Hypothesis 5 (H 5 ):
Job satisfaction mediates the relationship between social well-being and intention to leave.
The uniqueness of this study is threefold: firstly, it was the first study to focus on the social well-being of a utility organisation in a developing country. Secondly, findings from this study can promote positive social change, as indepth knowledge of social well-being and how it relates to important work outcomes will help eradicate harmful effects such as high intentions to leave, which contribute to organisational inefficiency, profit losses and substandard service delivery.

Research design
A quantitative cross-sectional survey design was used in this study. This design was considered suitable because there was limited empirical evidence to support the nature of potential relationships between the variables of interest, which necessitated an initial exploration of such possibilities (Spector, 2019).

Participants and setting
The participants included employees representing a cut across various designations and levels (senior management, middle management, junior management and all skilled employees) who were permanently employed in the electricity industry of South Africa. A total of 403 eligible employees in the organisation, scattered throughout all 9 provinces of South Africa, responded by completing the questionnaire in full. The most common age distribution ranked between 31 and 40 years; this meant that almost one quarter (38.7%) of the workforce were at the halfway mark of their professional lives. All employees who participated in the study had a minimum qualification level of Grade 12; more than half (51.6%) of the sample group had obtained a postgraduate qualification as their highest level of education and occupied various ranks (low, middle and senior) in the organisation. Employees were seen to have had approximately a minimum of 11 years of service in the organisation.

Measuring instruments
A biographical questionnaire was developed by the researcher and was aimed at measuring demographics.
The Social Well-being Scale of Keyes (1998), as validated for a South African workplace context by Redelinghuys (2016), was used to assess social well-being in terms of five subdimensions: social integration (e.g. 'during the past month, how often did you feel that people in your organisation are basically good?'); social acceptance (e.g. 'during the past month at work, how often did you feel that you really belong to your organisation?'); social contribution (e.g. 'during the past month, how often did you feel that you had something important to contribute towards your organisation?'); social growth or actualisation (e.g. 'during the past month at work, how often did you feel that your organisation is becoming a better place for people like you?') and social comprehension or coherence (e.g., 'during the past month, how often did you feel that the way your organisation works, makes sense to you?'). The items were scored on a sixpoint scale, ranging from 1 (never) to 6 (every day).
Respondents had to answer questions regarding the frequency with which they had experienced specific symptoms of social well-being during the past month. This response option allowed for the categorisation of levels of well-being, similar to the three classes used to assess positive mental health (Keyes, 2002). Individuals who were neither flourishing nor languishing fell into the category of moderate well-being. In a South African study, the reliability coefficient of this scale was 0.89 (Rautenbach & Rothmann, 2017).
The Job Satisfaction Scale (JSS) (Saks, 2006) was used to measure job satisfaction. Five items assessed individuals' satisfaction with their jobs (e.g., 'most days I am enthusiastic about my work' and 'I consider my job rather unpleasant').
Response options ranged from 1 (totally disagree) to 5 (totally agree). The Cronbach's alpha coefficient for the JSS was found to be 0.84 (Diedericks, 2012).
Organisational citizenship behaviour was measured using an adapted version of the Organisational Citizenship Behaviour http://www.sajip.co.za Open Access Questionnaire (OCBQ; Konovsky & Organ, 1996). There were six items in the questionnaire that were rated on a sevenpoint scale, ranging from 1 (strongly disagree) to 7 (strongly agree). Factor analysis confirmed that the OCBQ consisted of two factors, namely altruism (e.g. 'I assist others with their duties') and generalised compliance (e.g. 'I defend the organisation when other employees criticise it'). Redelinghuys et al. (2019) reported that the internal consistency of the scale was acceptable.
The Turnover Intention Scale (TIS; Sjöberg & Sverke, 2000), used to measure intention to leave, comprised three items (e.g. 'I am actively looking for other jobs'), with response options ranging from 1 (strongly disagree) to 5 (strongly agree). A previous South African study recorded a reliability coefficient of 0.71 for this scale (Janse van Rensburg et al., 2017).

Research procedure
The researcher obtained permission from the highest level of authority in the organisation to conduct the study in the utility industry. Scientific and ethics clearance was secured from a reputable institution of higher education to conduct the study and all ethical standards as prescribed by relevant legislation were always observed during the implementation phase. The questionnaire was presented in the form of an electronic booklet explaining the purpose of the study, emphasising confidentiality and stipulating processes to follow in case further clarity was needed.
Candidates were also made aware that participation was voluntary and that they were at liberty to withdraw at any stage of the research procedure. An independent service provider collated the data collection process. Once consent forms had been completed, access to the online survey was provided. The survey was constructed in such a manner that all items on any given page had to be completed before the participant could proceed to the next question. This removed all risk of the occurrence of missing values. A time frame of 2 weeks was allocated for completion of the online survey. The researcher obtained access to results by means of an anonymised format to analyse data accordingly.

Statistical analysis
Descriptive statistics were computed to describe the data and the sample characteristics. Point estimates of scale reliability were computed using confirmatory factor analysis (CFA) (see Raykov, 2009;Wang & Wang, 2020). A cut-off value for scale reliability of 0.70 (Nunnally & Bernstein, 1994) was used.
The measurement and structural models in this study were tested by using latent variable modelling with Mplus 8.6 (Muthén & Muthén, 1998. All variables were considered as continuous and a robust maximum likelihood method (MLM) was used as an estimator. The following indices were utilised to assess model fit: the chi-square statistic (the test of absolute model fit), standardised root mean square residual (SRMR), root mean square error of approximation (RMSEA), Tucker-Lewis index (TLI) and comparative fit index (CFI) (West, Taylor, & Wu, 2012). Tucker-Lewis index and CFI values higher than 0.90 are acceptable although values above 0.95 are preferred; RMSEA and SRMR values lower than 0.08 indicate a close fit between model and data (Wang & Wang, 2020). The Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were used in addition to other fit indices to assess the fit of competing models. The AIC is meaningful when estimating different models, with the lowest AIC indicating the best-fitting model. The BIC indicates model parsimony (Kline, 2010). Simple mediation analysis was performed using Mplus 8.6 (Muthén & Muthén, 1998.

Results
Based on the results from Harman's single-factor test, the fit statistics of loading the model onto one factor were as follows: χ 2 = 1756.65, df = 464, CFI = 0.66, TLI = 0.63, SRMR = 0.11 and RMSEA = 0.11. The fit statistics showed that the model did not fit, which indicated that common method variance (CMV) was not a problem (Tehseen, Ramayah, & Sajilan, 2017). If there had been model fit to the one factor, CMV could have posed a problem for the study.

Testing the measurement model
Four competing measurement models were tested by using CFA with Mplus 8.6. Model 1 was constructed as theory proposes: Social well-being was constructed as a secondorder latent factor consisting of five first-order latent variables, namely social integration, social actualisation, social coherence, social acceptance and social integration. Each of these five subdimensions was measured by three items. Five directly observed indicators measured job satisfaction. Six directly observed indicators were used to measure organisational citizenship behaviour in terms of two subdimensions, namely altruism (three items) and generalised compliance (three items), with organisational citizenship behaviour being a second-order latent variable.
Results for the first model indicated that a perfect fit was not attainable (χ 2 = 740.209, p < 0.05). However, because of severe dependence of the chi-square test of model fit on sample size, this test is not problem-free and the literature recommends that it be considered in combination with additional fit indicators to obtain a more precise picture (Saris, Satorra, & Van der Veld, 2009). When doing so, it was clear that the RMSEA (0.05, p > 0.05), the SRMR (0.07) and the CFI (0.93) and TLI (0.93) values pointed to an acceptable approximate fit of the model to the observed data. From an inspection of the standardised factor loadings of the items, it was, nevertheless, evident that the third item of the JSS ('each day of work seems like it will never end') did not load significantly onto its underlying factor (p = 0.922). It is possible that the participants found the item too confusing. Consequently, this item was removed, and a revised model was tested.
All components of the revised model were specified as in Model 1, except for job satisfaction, Job satisfaction was only measured by four items because of the omission of the third item. Although the chi-square test value for Model 2 was still significant, all indicators of the revised model pointed to a more precise representation of the observed data, namely CFI = 0.94, TLI = 0.93, SRMR = 0.07 and RMSEA = 0.05, p > 0.51, with all the items also loading significantly onto their respective constructs (p = 0.000).
Model 3 was constructed like Model 2, except for organisational citizenship behaviour, which was now measured as a first-order latent variable measured by six directly observed variables. Comparison of the fit indices indicated that Model 2 fitted the data best relative to the competing models. The χ 2 value of the model was also significant (p = 0.00) -similar to those of the alternative models -and, thus, indicative of an imperfect fit. However, the overall fit when all indices were considered indicated that this model met the requirements for an approximate fit. It provided a superior fit to the data compared with that of the alternative competing models. Furthermore, both the AIC (35698.46) and BIC (36086.35) values confirmed Model 2 to be the superior model. The standardised regression coefficients of this model were all statistically significant (p < 0.01), and all items were loaded on their respective constructs as expected, with values ranging from -0.44 to 0.94. The standard errors for each of the standardised estimates were also small and suggested accuracy in estimating these values. Table 2 displays the descriptive statistics, reliabilities and correlations of the scales.
All correlations, except for the relationship between intention to leave and organisational citizenship behaviour, were statistically significant: p < 0.01. As expected, social wellbeing was positively associated with job satisfaction and organisational citizenship behaviour, but inversely related to turnover intent. Correlations of large effect were found between subscales social well-being and job satisfaction and between job satisfaction and intention to leave. Correlations of medium effect were found for the association between social well-being and intention to leave, social well-being and organisational citizenship behaviour and organisational citizenship behaviour and job satisfaction.
As a result of the significant correlations amongst several variables, including those of social well-being with job satisfaction, organisational citizenship behaviour and intention to leave, hierarchical regression analyses were performed to investigate the possibility that social well-being might have a significant directional association with the dependent variables (job satisfaction, organisational citizenship behaviour and intention to leave).

Testing the structural model
The structural model was tested based on the preferred measurement model (Model 2), and the structural regressions were added to the model in line with what was to be expected based on the literature. The results also showed an acceptable fit to the data (TLI = 0.93; CFI = 0.94; RMSEA = 0.05; SRMR = 0.07). Table 3 displays the standardised regression coefficients for the structural model and illustrates that all the regression relationships were significant at a level of p-values < 0.01.
As is evident from Table 3, the portion of the model focusing on the association between social well-being and job  satisfaction showed a statistically significant standard path coefficient (β = 0.56; p < 0.01) and displayed the expected sign. Hypothesis 1 was supported. For the model portion focusing on organisational citizenship behaviour, the path coefficient of social well-being (β = 0.43; p < 0.01) was statistically significant and displayed the anticipated sign. Hypothesis 2 was supported. Furthermore, social well-being in relation to intention to leave also had a statistically significant standard path coefficient and displayed the expected sign (β = -0.33; p < 0.01). Hypothesis 3 was supported.
In terms of effect sizes (Cohen, 1988), the model displayed in Figure 1 accounted for the following percentages of variance explained (R 2 ): job satisfaction = 31.4% (medium effect), organisational citizenship behaviour = 18.7% (small effect) and intention to leave = 11.0% (small effect).
To investigate the possibility that social well-being might also associate with organisational citizenship behaviour through job satisfaction as an underlying facilitative factor (H4), a simple mediation analysis was performed using Mplus 8.6. By using the Hayes (2018) procedure, the indirect effects of social well-being on organisational citizenship behaviour and intention to leave were evaluated. Bootstrapping was used to construct confidence intervals based on an empirically derived sampling distribution of the indirect effect. With 5000 bootstrap samples, 95% confidence intervals (CIs) based on biascorrected estimates were constructed. The indirect association of social well-being with organisational citizenship behaviour through job satisfaction was found to be significant: (β = 0.11, p < 0.01 [0.05, 0.21]). Therefore, Hypothesis 4, indicating that job satisfaction mediated the association between social well-being and organisational citizenship behaviour, was accepted. Furthermore, the indirect association of social well-being with intention to leave through job satisfaction was found to be significant (β = -0.42, p < 0.01 [-0.63, -0.26]). Therefore, Hypothesis 5, indicating that job satisfaction mediated the association between social well-being and intention to leave, was accepted.

Discussion
Instilling social well-being through improved levels of job satisfaction and organisational citizenship behaviour and lower intention to leave should form part of business priorities to promote growth beyond economic understanding; companies rarely realise the importance of social well-being in relation to these facets (Munzel, Meyer-Waarden, & Galan, 2018).
This study investigated how social well-being associated with important individual and organisational work outcomes, namely job satisfaction, organisational citizenship behaviour and intention to leave. The findings suggested that statistically significant associations existed between social well-being and job satisfaction (Hypothesis 1), organisational citizenship behaviour (Hypothesis 2) and intention to leave (Hypothesis 3). Furthermore, social well-being was found to be indirectly related to organisational citizenship behaviour via job satisfaction (Hypothesis 4).  The positive association between social well-being and job satisfaction was in line with recent studies by Alegre et al. (2016) and Cek and Eyupoglu (2020) that were conducted in settings other than the utility industry. As the study showed that social well-being promoted job satisfaction in a utility organisation as well, the findings added to the existing body of knowledge. In line with Schützenberger (2016), who emphasised the importance of job satisfaction for economically advanced and developing countries alike, the study illustrated the importance of the association between social well-being and job satisfaction in an organisation in a developing country context, which extended illustrations of the observed association between social well-being and job satisfaction beyond industry, profession and country-specific contexts.
Confirmation of the first hypothesis (H1) implied that interventions intended to promote job satisfaction ought to strengthen social support networks, for increased social well-being would yield positive results for job satisfaction (e.g. increased participative behaviour or employees having more creative problem-solving ideas); this can be performed, for instance, by offering employees productive spaces to build interpersonal relationships (Dutton & Ragins, 2007). From a financial perspective, encouraging job satisfaction through employee relations as embedded in social wellbeing is imperative because organisations make money through retention of happy and satisfied employees, which has a direct impact on organisational growth or revenue generation.
The second hypothesis focused on the link between social well-being and organisational citizenship behaviour. The finding in this regard also confirmed a positive association and provided further support for previous research that demonstrated that organisational citizenship behaviour was heightened through social well-being (Kumar, Jauhari, & Singh, 2016;Masilela, 2018;Rastogi & Garg, 2011). The findings indicated that improved social ties and positive social exchanges led to multiplication of organisational citizenship behaviour; the more satisfied employees were with their work, the more likely it was that they would be inclined to demonstrate organisational citizenship behaviour -and this behaviour would be directed towards co-workers and towards the organisation (Kumar et al., 2016;Rastogi & Garg, 2011).
The present findings suggested that organisations ought to explore methods of increasing social well-being to promote organisational citizenship behaviour as prompted by social ties (Kumar et al., 2016). Ways to do this can include incorporating employee suggestions more; this will free superiors to spend more time on strategic initiatives that will improve organisational directives and allow workers to feel appreciated (Attaran, Attaran, & Kirkland, 2019). Organisations should also look into motivating organisational citizenship behaviour through non-monetary interventions (e.g. encouraging further studies, on-the-job training or acting in higher positions, etc.) because research has shown that employees are not only fuelled by money (Allen, Peltokorpi, & Rubenstein, 2016). Organisations can increase organisational citizenship behaviour by acknowledging employees' religious customs as one means of doing so. This will also signal that organisations recognise that employees have a deeper life nourished by meaningful work over and above salary reasons, which will foster a sense of togetherness, which aligns organisational values (Hudson, 2014).
The third hypothesis focused on the relationship between social well-being and intention to leave, revealed that where high social well-being existed, a decrease in intention to leave was observed. This provided further evidence that the work environment did indeed play an important role in promoting employee social well-being and in retaining employees. This finding supported previous research showing a link between social unwellness and voluntary turnover (Janse van Rensburg et al., 2017;Redelinghuys et al., 2019). In addition, it strengthened previous research claiming that reasons for turnover extended beyond a lack of monetary incentives that traditional turnover models concentrated on why employees left, but did not consider the aspect of social well-being and such perspectives were too narrow to provide a holistic view of why employees left (Allen et al., 2016).
The fourth hypothesis of this study was to test whether the relation of social well-being with organisational citizenship behaviour was mediated by job satisfaction. This study showed that job satisfaction and organisational citizenship behaviour were associated with social well-being. Furthermore, the effect of social well-being on organisational citizenship behaviour seems to be enhanced by the presence of a high level of job satisfaction. The findings further confirm the conclusions of Isen and Baron (1991) that a high level of job satisfaction evokes positive moods, which, in turn, increase organisational citizenship behaviour. Job satisfaction also pointed towards a catalyst for organisational citizenship behaviour.
The final hypothesis of the study focused on job satisfaction as a mediator of the relationship between social well-being and intention to leave. Our findings showed that when social well-being is strengthened, intentions to leave the organisation are likely to decrease. The simultaneous presence of higher levels of job satisfaction would serve as an underlying mechanism through which the inverse effect of social well-being on turnover intention would be amplified. This finding supported a previous study by Gandy et al. (2018) that underlined the importance of both social relations and job satisfaction for retaining valued employees and extended the existing knowledge base. As far as we are aware, it was the first time that the facilitative role of job satisfaction in the association between social well-being and turnover intention was tested within a developing country context.
The findings of this study confirm previous research advocating that a more integrated approach to promote wellbeing, including its social aspects, ought to be adopted that would require employee well-being to be at the centre of company agendas (De Simone, 2014;Litchfield et al., 2016). It is, thus, imperative that the utility industry take note of the associations found between social well-being, job satisfaction, organisational citizenship behaviour and intention to leave because implementing ways to increase employee social well-being to retain human capital will be in the best interests of the industry. This could be beneficial to South African industries that already have a social advantage in the form of Ubuntu and leveraging existing social structures could catapult organisations into unlimited horizons when it comes to retaining employees.
Regarding the effects of social well-being on job satisfaction, organisational citizenship behaviour and intention to leave, the results showed that the effect size for job satisfaction was almost double the size for organisational citizenship behaviour and three times the size for intention to leave. Therefore, it seemed that social well-being had a bigger effect on employees' job satisfaction than on organisational citizenship behaviour and intention to leave. This means that when the organisation supports social well-being, benefits reaped will be evident in terms of improved employee wellbeing and in more organisational citizenship behaviour and decreased intention to leave; the most striking benefits will, however, be earned in terms of improved job satisfaction outcomes. The utility industry can employ indicators of social well-being (such as the measure employed by this study) to establish a baseline for establishing existing employee social well-being levels and set targets for improving social well-being that human resource practitioners can monitor to devise strategic plans. Social well-being can also be fostered by encouraging initiatives to be published in a monthly newsletter or wellness index that can monitor social well-being amongst workers and enrich the value of human capital (Sinobuntu), which forms an important part of the utility industry.
The findings of this study can be understood in more specific terms by considering the specific components of social wellbeing in more detail. Social integration (feeling part of a community): this has a positive effect on an employee's psychological state because it fosters a sense of belonging and well-being, which is nurtured through the employee's relationship with his or her work society or community. Employees feel that they form part of an organisation when they have something in common with other workers in the same environment; such employees stay longer and are less inclined to exit the organisation (Keyes, 1998). According to Geue (2017), 77% of workers deem workplace friendship to be a priority. Methods such as one-on-one engagement, incorporating remote workers into teams, celebrating milestones and encouraging employee input can increase belongingness amongst disheartened utility industry workers, and this dynamic compels managers to look into methods to enrich organisational relations.
Social acceptance (accepting others): employees who display high social acceptance work together and rely heavily on one another for support; it eases the burden of uncertainty, leading to less turnover (Geue, 2017). Individuals absorbed into an organisation become familiar with organisational culture through social acceptance -a process encouraged by social inclusion through fair recruitment or promotional opportunities. The workplace consists of people from various backgrounds and organisations should uncover what promotes social inclusion amid diversity; this will improve organisational cohesion, interaction and teamwork (Joubert, 2017).
Social coherence (understanding the social world): individuals who embrace opportunities and challenges hold companies in high regard and are less likely to resign from an organisation (Struwig et al., 2013). Coronavirus disease-19 (COVID-19) has forced organisations to increase efforts to strengthen citizenship. This can be done through employee engagement sessions (online sessions), 10-min interaction sessions before electronic meetings and planning employeeof-the-month virtual award ceremonies to recognise good work, which will increase employees' social well-being (Center for Collegiate Mental Health, 2017). In addition, organisations can implement exploratory analysis where questions are incorporated to provide the company with an indication of the level of social cohesion experienced amongst workers. This will enable an organisation to address identified pitfalls (Struwig et al., 2013).
Social actualisation (an individual's growth within society): employees perceive employers as caring about them once their suggestions have been incorporated. This allows employees to grow, which improves company efficiency. Excessive turnover rates expose organisational defects; money is often used as a band-aid, leaving the real reasons for turnover unresolved. Excessive intention to leave rates are not always negative, as they provide employees with a platform for coming up with innovations to curb intention to leave; this participation contributes to a growth in employees' morale and decreased intentions to leave (Songcaka, 2015).
Social contribution (an individual's sense of contributing to society): people provide organisations with insight to formulate better organisational strategies. Considering employees' contribution results in employees working harder when efforts are valued and becoming more dedicated to the company. Employees' intrinsic rewards from feeling valued are greater than financial compensation and are more likely to make them stay in the organisation (Mukherjee, 2020).

Limitations and recommendations for future research
Although the employment of a cross-sectional study was useful for exploring initial relations between constructs in a novel context such as the utility industry (which is hampered by financial constraints) in a cost-effective manner (Spector, 2019), the design made it precarious to draw conclusions regarding regression analysis based on the impact of social well-being. Future studies could follow up through longitudinal research to verify causal relations and gain a wider understanding that could refine recommendations for interventions.
The self-reported data might have been contaminated by CMV. This possibility is acknowledged but needs to be seen in perspective. Spector (2019), for example, questions the usefulness of CMV testing for several reasons, including the problem that none of the existing range of statistical techniques can eliminate the possibility of CMV beyond any doubt and also acknowledges that the use of self-reported questionnaires is still considered ideal to tap into employees' personal thoughts, feelings and subjective assessment of their own experiences despite the risk of encountering CMV in the process (Spector, 2019). Finally, this study only focused on study participants from a utility organisation and findings cannot be generalised to other contexts without expanding the scope.

Conclusion
Given the current state of the world of work, ensuring worker well-being and retaining human capital in the utility industry should be the top priority. Not only do workers provide an essential service; they are also at the forefront in driving South Africa's economy. The findings from this study underlined the importance of social well-being and its role in enhancing job satisfaction and organisational citizenship behaviour and in reducing employee turnover. This means that in contrast to throwing money at problems as a quick-fix attempt, businesses have to reinvent retention strategies by taking a closer look at how enhancing the social well-being of employees can be used to attract and keep the best employees.