“Antecedents of employee wellbeing in the banking sector: the moderating role of working environment”

This study examined the influence of work-life balance, work stress, employee en- gagement, and working environment on employee wellbeing in the banking sector of Pakistan. Due to complex human resource policies in Pakistan, employee wellbeing is neglected in several banking institutions; this study addresses a research gap in this way. Drawing upon job demands-resources theory, the study employed a quantita- tive methodology through a survey of 360 employees from private and public banks in Pakistan. The results from PLS-SEM (Partial Least Squares Structural Equation Modeling) demonstrate that employee engagement and work stress are significantly related to employee wellbeing, while working environment has a significant interactive effect between employee engagement and employee wellbeing. Theoretically, the study contributes to broadening the existing literature on human resource management. Practically, this study provides guidelines to human resource practitioners, managers, and policymakers on devising strategies for their employee wellbeing in going forward. 2010). Despite the literature available on employee wellbeing, there is a dearth of studies in developing countries, especially in Pakistan. It is argued that exploring the antecedents of employee wellbeing, notably, work-life balance, work stress, employee engagement, and working environment on employee wellbeing, would reveal explanations why organizations should formulate employee-friendly policies at the workplace, which has been ignored in earlier studies (Kossek et al., 2014). This study provides guidelines to managers about ways to improve employee wellbeing in the workplace.


INTRODUCTION
In the modern era, work-related stress has been greatly exacerbated by adversarial psychosocial working conditions. It is the organizations' responsibility to look after the mental health and well-being of employees (Johnson et al., 2020). It is imperative that the 'third goal' of Sustainable Development Goals (SDGs) of the United Nations is to focus on "good health and wellbeing," which has increased awareness and importance among policymakers and researchers globally . In this light, any country's financial system is a vital institute, and the banking system has been professed as a hub of the financial system (Kaur & Sandhu, 2010).
Despite the literature available on employee wellbeing, there is a dearth of studies in developing countries, especially in Pakistan. It is argued that exploring the antecedents of employee wellbeing, notably, worklife balance, work stress, employee engagement, and working environment on employee wellbeing, would reveal explanations why organizations should formulate employee-friendly policies at the workplace, which has been ignored in earlier studies (Kossek et al., 2014). This study provides guidelines to managers about ways to improve employee wellbeing in the workplace.
Although working environment and work stress also play a vital role, the prime concern is that every organization should ensure a sound work-life balance. While studies on this concern are quite rare (Greenhaus & Allen, 2011;Haar et al., 2014), it focuses on the intra-individual transfer of balance. Therefore, management research needs to strive for a higher impact in a competitive labor market (George, 2016). Evidence suggests that employees who perceive work-life balance and are effective in life roles tend to have a higher sense of employee wellbeing (Lyness & Judiesch, 2014). Though, employee engagement has been defined as a distinctive and novel construct comprising cognitive, emotional, and behavioral components allied with individual role performance (Saks, 2006).
In a nutshell, employees' competitiveness and sustainability in the service sector in the last decade have continued to trigger scholars' interest in this subject matter (Hussain et al., 2020). Therefore, the main purpose of this study is to address the contextual gap by conducting a study in banks on employee wellbeing, which is linked to employee performance. This study also broadens the empirical evidence in the field of human resource management. Again, it is also imperative to reiterate the antecedents of this research theme, such as work-life balance, work stress, employee engagement, and working environment about employee wellbeing. To implement this goal, this study encompasses the literature by applying job demands-resources theory (JD-R) in precise by opening up the scientific discussion about the comparisons that underlie JD-R.

Job demands-resources theory (JD-R)
Job demands-resources (JD-R) theory is related to the work-linked features that influence employees' job execution, physical and psychological wellbeing (Hobfoll, 2001). According to Schaufeli (2013), an individual's job and personal traits combine to reinforce high levels of engagement. The theory explains how job characteristics and job performance are related (Menguc et al., 2017). In recent research by Kim and Beehr (2018), there is a difference between job demands, challenges, and obstacles. Eldor (2017) argues that employees feel bored due to high job demands at the workplace. Employees face psychological and health issues due to job demands at the workplace (Bakker et al., 2007;Hakanen et al., 2006). More importantly, multiple factors, such as learning and personal development, reduce job demands (Bakker & Demerouti, 2017).
The World Health Organization has defined wellbeing as "a comprehensive condition of mental, physical and social wellbeing, not only lack of infirmity or disease". Durand (2015) stated in his study that the OCED has lately established a range of parameters of wellbeing as a fragment of 'better life' inventiveness. He also focused on the subjective wellbeing that is defined as 'a good mental state that comprises of numerous optimistic and negative evaluations that individuals make of their lives and the affective response of people and their experience.' In this era, work-life balance is a foremost concern of employees working in an organization, especially if they were given flexible working arrangements. It is recommended that there be a balance between working hours and employees' workload (Haddon, 2018). Petrou et al. (2012) emphasize that an individual who is provided with work-life balance opportunities has more satisfied and committed attitudes at the workplace. According to their research findings, they found a positive relationship between work-life balance and employee wellbeing. Enehaug et al. (2016) explored the correlation between worklife balance and employee wellbeing. According to the world population, the sixth highly populated country in the world is Pakistan, with 210 million people (Worldometer, 2020). The banking sector in Pakistan is deliberated as a leading industry with diverse branches across the country. As stated by the State Bank of Pakistan (SBP, 2020), many individuals are working under the umbrella of the banking industry after its reform in the early 1990s. In banks, work conditions have changed drastically in the last two decades (Khilji, 2006).
According to Sonnentag and Frese (2012), work stress is associated with a maximum number of possible outcomes. Although several studies are related to stressor-to-strain relationships, few other studies look into the inverse relationship, i.e., from wellbeing to job stressors, and have published empirical support by Ford et al. (2014). However, Friedenreich et al. (2016) found that few studies failed to determine the negative relationship between work stress and employee wellbeing.
The researchers have generally neglected the importance of work stress at the workplace (Eldor, 2017). According to McDonald and Westphal (2013), work stress created an equal impact on male-female employees because the work stress was more or less similar for each employee. Importantly, Beehr et al. (2001) argued that stress is always a response from emotional problems, so it is the organization's responsibility to provide emotional stability to its employees. At least employees should be given less pressure from the employer. This study has analyzed that work stress is significantly related to employee wellbeing (Hendrie & Pickles, 2010;Nesse, 2000).
Employee engagement research has gained encouragement (Albrecht et al., 2010). The cognitive, emotional, and behavioral energy boost the performance of employees and organizations by their engagement at the workplace concepts of job demand and job/personal resources. The idea of employee engagement is crucial to management, and there is limited research about this new construct (Saks, 2006).
In almost every organization employee who is a high achiever is always in demand. When an employee feels that they have independence in decision-making power and can give feedback on their work, it leads to engagement and satisfaction (Truss et al., 2013). When an employee can work independently and have authority and empowerment, it refers to employee engagement. Therefore, employees who have decision-making power and autonomy of work do not need to take permission for each decision and feel more satisfied and contribute to the organization's progress. According to Truss et al. (2013), employee engagement has always been a problem associated with wellbeing, and it has also been substantiated by earlier literature. It is a crucial relationship concerning organizational performance because it is obvious that employees having mental peace always benefit their organization through their performance. According to Tan et al. (2020), there is a strong relationship between employee engagement and wellbeing.
Kossek, T. Kalliath, and P.    2017) highlighted the workplace environment and its impact on employee health. They considered that if the employee is paid a salary for his work, there is no need to improve their working conditions. A few studies on the working environment indicate that employee wellbeing is strongly affected by the working conditions. They include a healthy relationship with peers, a manageable workload, leading practices, and autonomy for work (Nielsen et al., 2017).
The current study addresses the research gap and aims to: 1) examine the factors that impact employee wellbeing in the banking sector; 2) assess the moderating effect of working environment between work-life balance, work stress, employee engagement, and employee wellbeing in the banking sector.
To implement the aims stated above, the paper hypothesizes the following: H1: Work-life balance is positively associated with employee wellbeing.
H2: Work stress significantly affects employee wellbeing.
H3: Employee engagement is positively associated with employee wellbeing.
H4: Working environment moderates the relationship between work-life balance and employee wellbeing.
H5: Working environment moderates the relationship between work stress and employee wellbeing.
H6: Working environment moderates the relationship between employee engagement and employee wellbeing.
From the literature review, the research constructs (see Table 1) and the conceptual framework (see Figure 1) and their hypotheses are summarized.

Measurement of constructs
It was noted that the items used for measuring the constructs were adapted from the existing literature. Therefore, it is important to mention that a seven-point Likert scale was used to measure all items ranging from 1 = Strongly disagree to 7 = Strongly agree, and all the items were in affirmative statements. The measurement scale for work-life balance was taken from Tasnim

Data collection and analytic technique
This study adopted the survey method approach, and its philosophical positioning is the positivism paradigm. and sales officers from public and private sector banks from the metropolis city Karachi, situated in the Sindh province of Pakistan. This study adopted a convenience sampling technique to collect data from employees working in public and private banks. The data were collected from one province (Sindh) of Pakistan as it is an industrial hub of Pakistan. According to Etikan et al. (2016), the technique was adopted due to respondents' convenience, geographical nearness, eagerness to participate, participants' ease of access to the investigator, and affordability in terms of the cost related to the accomplishment unit of analysis.
The period to collect the data was from September to December 2019. The confidentiality of information was assured after sending a letter of participation to respondents (Khuwaja et al., 2020). The average time to fill the questionnaire was 5 minutes. Out of 500 samples, only 360 respondents signifying 72%, were valid for data analysis. The demographic summary of the respondents with frequency and percentage is illustrated in Table 1. Table 1 indicates that the sample of this study consisted of 55.55% male employees, while 44.44% are female employees working in banks. 55% of the respondents were aged between 30 to 40 years. At the educational level, 56.11% of respondents were holding a Bachelor's degree. 63.61% of employees worked in private sector banks, while 36.38% were working in public sector banks. Most of the employees had 6 to 10 years of experience in banks, with 54.44%. The bank manager population consisted of 16.11%, while 24.16% were senior branch officers-services.

Model assessment
Conferring to the pioneer specialists Hair et al. (2017), it is significant to assess the constructs' reliability using Composite Reliability besides Cronbach's Alpha coefficients. Table 2 indicates that all the values surpassed the threshold was presented, which also surpassed the minimum threshold of 0.5 as mentioned in Table 2. Regarding the indicator loadings of the covert constructs, all items were loaded implicitly to their conforming construct. The measured indicators have maximum loadings of nearly 0.9 and minimum loadings of nearly 0.5. Bagozzi, Yi, and Nassen (1998) stated that the best measurement of a latent variable under study should have a loading above the threshold of 0.6. These indicator variables have a maximum load of 0.884 and a minimum load of 0.551, and this specifies the measure of what they should measure. Therefore, Table 3 summarizes all the research constructs with their measurement items and corresponding coefficients (loadings).

Coefficient of determination (R-squared)
Concerning the research constructs' predictive power, the coefficient of determination (R 2 ) of the regression model was measured. The predictor (independent) variable explains the coefficient that specifies the proportion of the difference in the dependent variable. The R 2 of WLB (0.001) exhibited a 1% variation in the construct. The construct's EWB describes WLB. The adjusted R 2 shows the variance in the endogenous construct explicated by the exogenous construct. Table 3 shows that the estimated R 2 of EE (0.012) indicated that 1.2% of the EE variation is explicated by the construct EWB as an independent variable. At the same time, model R 2 of the dependent variable EWB (0.16) specifies 16% of the total variation of the construct. In Table 3, EWB is explicated by individual constructs' collective effect: EE, WLB, WS, and WE. The items whose factor loading threshold value was less than 0.60 were deleted (EE3, EE6, WLB2,  WS2, WE1, WE5, WE6, EWB1, EWB2, EWB3,  and EWB7). Simultaneously, to establish discriminant validity, the criterion was used to measure the existence of discriminant validity among the covert variables (Henseler et al., 2015). The findings from Fornell-Lacker criterion specified that constructs fulfill both stringent and basic assumptions and establish discriminant validity. It is significant to note the values in bold (diagonal) of the Fornell-Lacker criterion. Table 4 specifies AVE's of the measured constructs and must be greater than 0.5. AVE should be of higher value (coefficient) at both column and row position over other constructs to establish discriminant validity (Fornell & Larcker, 1981). The capacity of survey data collection instrument questions to capture data for its defined tenacity and distinguish themselves from other questions with construct and in between construct is measured through the test of discriminant validity by employing the Fornell-Larcker criterion (Hair et al., 2010). The Partial Least Squares Structural Equation Modeling (PLS-SEM) has been employed to calculate the value of the Fornell-Larcker criterion (Fornell & Larcker, 1981). The value of the Fornell-Larcker criterion of each construct should be higher than its contiguous value, and a higher value shows that each construct is different from the other and extends to only its relevant phenomena (Hamid et al., 2017). Therefore, the Fornell-Larcker criterion results of the current study show that every construct reports a higher value on its construct and a lower value on the contiguous construct. Thus, it can be concluded that the Fornell-Larcker criterion has been achieved.

Structural model
Proceeding from the model fit assessment, the structural model (path analysis) of the hypothetical analysis is required. Noticeably, it is relevant to achieve this stage of the analysis since it identifies and establishes the causal effect or relationships of the underlying research aim's constructs. The results reveal the direct and indirect effects of factors that trigger employee wellbeing in the banking sector of Pakistan. Regarding the direct effect, the result revealed that the constructs: Employee Engagement (EE) and Work Stress (WS) have a significant effect on Employee Wellbeing (EWB) with (β = 0.105, t = 1.972) and (β = 0.341, t = 8.202), respectively. However, Work-Life Balance (WLB) has no significant direct effect on EWB with (β = -0.038, t = 0.569), as seen in Table 5.
Whereas concerning indirect (moderation) effect, the result was quite interesting, in that, with all three moderated hypotheses, only one was significant. Thus, the construct Working Environment (WE) as a moderating variable plays a significant moderation role between the interaction of EE and EWB with (β = 0.203, t = 4.032). It suggests that the working environment of employees determines the motivation level towards their zeal in the execution of their job, which would consequently boost the level of morale in their quest to achieve organizational goals and objectives (see Table 5).

DISCUSSION
This study considers it important to examine the work-life balance, work stress, employee engagement, and working environment with employee wellbeing in the banking sector, more importantly, all employees affected by the negative consequences of the wellbeing program. H1 result demonstrates that employee engagement has a direct and positive impact on employee wellbeing. The result is in line with the recent research by Rahman, Björk, and Ravald (2020); they discovered that employee engagement has a positive relationship with employee wellbeing. The second H2 found that work-life balance has no relationship with employee wellbeing in this study. Thus, the finding contradicts a study by Fotiadis et al. (2019); they found that work-life balance is associated with wellbeing.
It is also interesting to note that H3 found that work stress is positively associated with employee wellbeing in the banking sector. This study is in line with a study by Ross (1995); work stress is a crucial issue being faced by employees and managers, which leads to the impact on organizational performance and their wellbeing at the workplace. According to H4 and H5, working environment has no moderating effect on work-life balance and work stress with employee wellbeing. Conversely, this study found that according to H6, working environment moderates the relationship between employee engagement and employee wellbeing. This study is in line with the result of Saleem et al. (2020); they found that working environment positively impacts employee engagement.

Research implications
This study has provided two practical implications. Firstly, top management of organizations formulates employee-oriented policies that improve the wellbeing of employees at the workplace. Employees face several hurdles, such as work stress, high workload, work and family issues due to work. The workplace dynamics affect the employees' health and affect them psychologically due to a pressured working environment and high targets from managers. Secondly, this study tries to provide comprehensive advice for managers on how they should develop policies regarding flexible working hours for their employees. Employees are assets of organizations, so they must take care of their employees to achieve competitive advantage and reduce turnover. It will affect the overall performance of businesses.

CONCLUSION
This study contributes to the body of knowledge on employee wellbeing at the workplace by examining employees' engagement, work-life balance, work stress, and working environment in the banking sector of Pakistan. More importantly, the findings of this study reveal that Employee Engagement (EE) and Work Stress (WS) have a significant effect on Employee Wellbeing (EWB). This finding suggested that engagement with peers and managers improves their wellbeing at the workplace. However, work stress affects employees' performance and wellbeing due to long working hours when dealing with many customers in banks. Unexpectedly, Work-Life Balance (WLB) did not find sufficient evidence with employee wellbeing. Subsequently, an important finding to emerge in this study was Working Environment (WE) as a moderating variable, which plays a significant moderation effect between the interaction of EE and EWB, thus deviating from traditional thoughts and providing a fresh perspective on the subject matter. The findings of this study broaden the understanding of employee wellbeing in the redesign and adjust policies and strategies in the banking institutions. Notably, this study also provides guidelines to human resource practitioners, managers, and policymakers on devising strategies for their employee wellbeing programs to boost performance at the workplace while boosting their morale.
This study has few limitations, which paves the way for future research to understand the relationships examined in this study. Firstly, this study was conducted in a developing country such as Pakistan (South-Asian region); hence, future research could be conducted in developed regions such as Europe, Australia, the USA, the UK, and Japan. Secondly, this study focused only on one service sector; therefore, future research may test the model in other equally important sectors/ industries such as textile, oil and gas, cement, pharmaceutical, and other sectors such as tourism, education, hotel, and hospitality. Thirdly, the model of this study is revealed around a few predictors of employee wellbeing. Future research could also explore this subject matter by considering 'work stress' as a mediator variable to get interesting results about employee wellbeing. Therefore, this study encourages future researchers to replicate the model on two-way aspects: top managers and sales teams in the banks.