IMPACT OF CORPORATE SOCIAL RESPONSIBILITY ON FINANCIAL PERFORMANCE: MEDIATING ROLE OF QUALITY OF WORK LIFE

This study examines the influence of perceived corporate social responsibility (CSR) on corporate financial performance (CFP). Based on win-win paradigm, this study uncovers the mediating role of Quality of Work Life (QWL) on CSR-CFP nexus. A questionnaire based cross-sectional survey was conducted to accumulate data from 355 employees working in public and private sector banks of Pakistan. Structural Equation Modeling (SEM) was used to examine the hypotheses. The results depict that CSR is positively related with CFP, and employee QWL partially mediates this relationship. According to our noesis, the review of previous literature regarding the association of CSR and CFP reveals that QWL has not been empirically tested as a mediator between these two variables. This study findings proposes that banks engaged in CSR activities are actually establishing a healthy work environment for their employees which ultimately helps the banks to improve their financial performance. This study highlights the importance of CSR activities for enhancing the CFP of the banking sector in the developing countries.


INTRODUCTION
Now-a-days, the concept of corporate social responsibility (CSR from here to onward) has become necessity for the organizations (Bapat & Upadhyay, 2021;Islam et al., 2016) as it involves a business conduct to be socially responsible, law abiding and economically profitable (Carroll, 1983). The idea of CSR was initiated in early 1950s by Bowen (1953) from business perspective in his seminal work "social responsibilities of businessman". Followed Bowen's (1953) approach, Carroll (1983) further developed the concept of CSR. However, the concept gained popularity in 21 st century, where businesses have become sensitive about human rights and ethics (Kanji & Chopra, 2010;Murphy & Schlegelmilch, 2013). In present times, CSR is considered as scientific work (Lu et al., 2020), and society expects responsible behavior from companies owing to the harmful results of their working on natural environment and humanity (Turker, 2009). Stakeholders have become the main driving force behind the adoption of CSR practices by the companies (Juniarti, 2020).
Given that, businesses are now conscious to re-analyze their role in society (Jain et al., 2017) and are adopting CSR as a differentiation technique to attract employees (Bapat & Upadhyay, 2021) as well as clients (Arslanagic-Kalajdzic, & Zabkar, 2017). Owing to this swift adoption of CSR by many business organizations, a debate regarding its long-term impacts has been aroused (Park & Lee, 2020). Stakeholders theory proposes that CSR expenditures could result in ameliorated financial performance by building positive relations with Stakeholders (Wu, 2006). However, adoption of CSR also raises a central question of shifting the organizational goal from maximization of shareholders wealth to satisfaction of stakeholders through building employees and society relations (Singh & Mittal, 2019). Moreover, firm costs might be increased and agency problem would be emerged (Chang et al., 2018). That's why Agency theory regarded CSR as Principal-agent problem owing to overinvestment by influence on CFP. Jia (2020) collected data of 3401 public companies in China for the period (2009)(2010)(2011)(2012)(2013)(2014)(2015) and concluded that CSR enhances CFP. Simpson and Kohers (2002) carried out research in banking sector, Beck et al. (2018) carried in public owned organizations and Naseem et al. (2020) analyzed 1021 Asian companies over the period of 2006-2016. All these studies discovered a positive relationship between CSR-CFP. Even the research based on data taken from Shanghai Stock Exchange (SSE) for the period of 2009-2013, by using agency problem as moderator, concluded that High Return on Assets (ROA) and Return on Equity (ROE) is associated with CSR engagement (Chang et al., 2018). Generally, this positive relation is expected based on stakeholders theory as well as agency theory (Aras et al., 2010).
Firms invest in CSR activities for the sake of delighting all its stakeholders. On the other side of coin, some argue that investment in social responsibility practices causes agency problem and decrease in wealth of shareholders (Chang et al., 2018). Negative CSR-CFP relation was found by others (Peng & Yang, 2014). CSR-CFP relation was probed with a sample of 213 firms, having impact of innovation, indicated that previous studies were over enthusiastic about the positive benefits, rather CSR has a 'dark side' that creates negative impact (Bocquet et al., 2017). Some argued that increase in CSR expenses leads to deteriorated profits (Alexander & Buchholz, 1978).
Some studies suggest that profit is neither increased nor decreased due to CSR activities. Various studies have found insignificant association between CSR-CFP (Alexander & Buchholz, 1978;Soana, 2011). Some studies narrate that CSR-CFP nexus can be explained as U-shaped instead of linear one (Franco et al., 2019;Park & Lee, 2009), stating high CFP is associated with low investment on CSR, low CFP is with moderated CSR performance, and highest CFP is backed by high CSR activities. Despite the plenty of literature on the nexus, literature flunks to furnish decisive evidence (Galant & Cadez, 2017). However, more common form of relationship is positive (Galant & Cadez, 2017). Thus, H1: CSR positively influences CFP.

Mediating Role of Quality of work Life
By drawing on the win-win paradigm, this study introduces QWL as a mediator to test indirect impact of CSR on CFP. Furthermore, Kim et al. (2017) also urged to probe the mediating role of QWL in the relationship between CSR and CFP. QWL is a broader concept which is comprised of different aspects on satisfaction like work life and overall life of employee i.e., financial, social and family (Sirgy et al., 2001). Rahman et al. (2017) viewed QWL as a philosophy which shapes employee satisfaction and their retention by affecting their work life. Accordingly, firms are trying to provide an improved work environment by satisfying their legal demands. This effects employees psychological, spiritual and physical well-being (Berg et al., 2003). QWL is a source to create social, psychological and economic relation between firm and its employees (Mirvis & Lawler III, 1984). QWL of employees could be increased by providing them desirable environment like fair remuneration, job security and supporting families etc., through CSR activities of firm (Bohdanowicz & Zientara, 2009). Ethical and philanthropic dimensions of CSR have the potential to meet employees QWL needs of actualization and self-esteem (Kim et al., 2017). Research has demonstrated that CSR activities effect the employees in different ways by improving their QWL and well-being (Kim et al., 2020). Moreover, now-a-days organization faces strong pressure from customers and government in shape of boycotts and sanctions respectively for negatively impacting society. So, if the company is not operating for the welfare of all stakeholders then its existence is endangered and causes employees job insecurity (Kim et al., 2017). Based on this argument, we can say CSR activities directed to all stakeholder can affect the QWL of employees (Kim et al., 2017). So, this study aims to test the influence of different CSR 's dimensions on employees QWL. When firms allocate resources to employees QWL, this increases their willingness to perform well for their firm (Asante Boadi et al., 2020). QWL is a source to increase organization performance and efficiency from organization's perspective and to create a comfortable working environment from employee's perspective (Rahman et al., 2017). Previous studies also found a significant and positive relation between QWL and organization's performance (CFP) (Chib, 2012). Here, our purpose of winwin situation seems to be achieved, when QWL is positively related with CFP (Lau & May, 1998). In nutshell, studies depict an undeniable relation among CSR, employee's QWL and CFP supported by Jakubczak and Gotowska (2015). Based on the previous literature, this study provides testable hypotheses and conceptual model which is explained in Figure I. H2: QWL mediates the direct relation between CSR and CFP.

Figure I: Conceptual Model METHODOLOGY Data collection procedure and sample
This study involves cross-sectional design and employees working in different banks of Pakistan were surveyed. Banking sector was opted to collect data because banks are gradually increasing their investment in CSR related activities (Pérez & Del Bosque, 2012). The study employed a questionnairebased survey method and a total of 1150 employees were approached on convenience basis. The intention was to take response from all individuals involved in the development or effectuation of CSR. Respondents' participation in data collection was volunteer; hence, no additional benefits to encourage participation were offered. The questionnaire was divided in two parts with a cover letter explaining the research purpose. In the first part of the questionnaire, respondents were asked about their personal characteristics (e.g., gender, qualification and age etc.). Whereas, in 2 nd part questions regarding their perception about CSR, QWL and CFP were asked. A total of 360 employees returned the questionnaire (response rate is 31.3%), which is acceptable in social sciences (Baruch, 1999). Five questionnaires were incomplete. Hence, they were not included and analysis was performed on 355 responses.
The sample comprises of 294 males (82.2%) and 61 females (17.2%), which denotes the gender-wise distribution of the respondents. 12.1% of the respondents are between the age limit of 20-25 years, 47.3% are between 26-30 years of age, 24.5% are between 31 and 35 years, and 16.1% are older than 35 years. 25.9%, 59.2% and 14.9% of the respondents hold Bachelor, Master and Above Master degrees respectively. Majority of respondents are of managerial ranks i.e., Branch manager, Operation manager, Relationship manager, General Manager and Assistant manager. 4.2% of the sample has less than 1 year experience, 23.4% between 1 to 3 years, 18% between 3 to 5 years, and 54.4% has more than 5 years of experience.

Measures
To ascertain validity and reliability of scales, all measures were taken from previous studies (Ouakouak et al., 2019). All constructs were measured on Five-Point Likert scale ranging from [1=Strongly Disagree] to [5 =Strongly Agree]. CSR Many approaches have been used to measure CSR but researchers lack consensus on a single method (Galant & Cadez, 2017). In this study, with the intent to measure CSR as perceived by employees, we espoused a 17-item scale (Fatma et al., 2014). This 17-item scale is based on 5 dimensions namely Customer, Employee, Shareholder, Environment and Society. Four-item scale was used to measure "customer" dimension of CSR with a Cronbach's Alpha (i.e. α) value of 0.818. "Employees" dimension was quantified by three-item scale with α value of 0.806. Three-item scale was used to measure "Shareholder" dimension of CSR with a Cronbach's α value of 0.664. The dimension of "Environment" was measured by three-item scale, with α value of 0.662. Four-item scale was used to measure "Society" dimension of CSR, having α value of 0.786. The overall Cronbach's Alpha value for CSR was noted as 0.905. Quality Work Life (QWL) It was reckoned by adopting a 5-item scale developed by (Hart, 1994). The Cronbach's Alpha value for QWL was 0.894. Corporate Financial Performance (CFP) CFP can be ciphered by 80 types of measures (Griffin & Mahon, 1997). For present study, CFP perceived by employees was measured by 4-Item perceptual measure of financial performance from precedent studies (Kim & Han, 2015). Ab initio, this measure was germinated by (Marsick & Watkins, 2003). Cronbach's α value for CFP was 0.798.

Data Analysis Strategy
Mean, Standard Deviation and correlation were computed for the variables in this study to comprehend the characteristics of sample. Cronbach's Alpha, AVE (average variance extracted) and CR (composite reliability) were employed to know reliability of used scales. Reliability and validity of variables were also evaluated by Employing Confirmatory Factor Analysis (CFA) and the heterotrait-monotrait ratio of correlations (HTMT) analysis in IBM AMOS 24.0. Moreover, Structural Equation Model was utilized to know the direct relation between Independent and Dependent variables. To quantify the mediation effect of QWL between CSR-CFP nexus, Baron and Kenny (1986) approach was used. Bootstrapping has also been employed to confirm the extended direct and indirect effect of mediation in this study.

Common Method Bias
The study tackled common method error, by using two techniques. Firstly, all measures used to collect data have been taken from prior studies, hence eliminating the risk of misinterpretation of the scale items (Ouakouak et al., 2019). Moreover, Harman's one-factor test was used on all 3 factors by converting them into one factor as suggested by Podsakoff and Organ (1986). This test was conducted with the purpose to know either one factor elaborates about specific amount of variance. Variance of 39.72% was unveiled by the results of test that is less than 50% threshold level. It means that Common Method bias is not a cause of discomfort here.

Descriptive statistics and correlations
Information regarding Mean (X), S.D and correlation (r) for independent, dependent and control variables is presented in Table I. The values show that, CSR has positive and significant correlation with QWL (r = 0.742, p < 0.01), and CFP (r = 0.577, p < 0.01). In addition, QWL is positively and significantly correlates with CFP (r = 0.589, p <0.01). However, control variables do not have any significant correlation with observed variables of the study.  (Hu & Bentler, 1999). CFA results are presented in (Table II) with high significance (p < 0.001) and all indicators have loadings greater than 0.60 (Bagozzi and Yi, 1988) except SHAR1, ENV1 and FP4 (we keep these items due to significant loadings) (Ouakouak et al., 2019). Cronbach's tests were conducted to check Reliability. All latent variables have Cronbach's alpha scores of more than minimum required threshold value of 0.70 (Hair et al. 2018; see Table II) used by . The convergent validity of the measures was examined through the values of CR and AVE which were well above the criteria (CR > 0.70 and AVE > 0.50), and by comparing the values of CR and AVE where the values of CR were greater than AVE (Ouakouak et al., 2019). To measure discriminant validity, Heterotrait -Monotrait ratios of correlation (HTMT) was conducted where Output values of HTMT analysis should be less than 0.90 and the situation will be ideal if the values are less than 0.85 (Henseler et al., 2015;Voorhees et al., 2016)) Table III shows that this model fulfills the conditions. So, it was proven that this model does not have any validity problem.  Table IV). As assumed in H1, positive and significant association is between CSR-CFP. (β = 0.684, p =0.000). These results strengthen the claim made in H1. Mediation Effect. The mediated impact of QWL was tested on the CSR-CFP nexus by following the Howell's (2007) approach which is based on the work of (Baron & Kenny, 1986). First, the impact of CSR on CFP was evaluated by ignoring the impact of QWL. This direct model provided the followings model fit indices (CFI = 0.949, IFI = 0.950; TLI; 0.942; CMIN/Df = 1.818, RMSEA = 0.048). SEM shows that Model fit was good having a significant relationship with a beta value of 0.684. Afterwards, CSR effect on QWL was examined (β = 0.858, p = 0.000). This relationship is significant and good Model fit indices were produced (CFI = 0.951; TLI = 0.944; IFI = 0.951; CMIN/df = 1.892, RMSEA = 0.050). Next to this, the nexus between QWL and CFP was evaluated, showing a significant relationship (β = 0.680, p = 0.000). All fit indices were acceptable (CFI = 0.988; TLI = 0.983; IFI = 0.988; CMIN/Df = 1.763; RMSEA = 0.046). Finally, after involving mediation of QWL in between the CSR-CFP nexus, now the path between CSR to QWL has (β = 0.859, p = 0.000), from QWL to CFP has (β = 0.353, p = 0.004) and from CSR to CFP has (β = 0.381, p = 0.003). Good Model fit indices were established (CFI = 0.946; TLI = 0.940; IFI = 0.947; CMIN/df = 1.806, RMSEA = 0.048). Here, it was concluded that beta value between CSR to CFP has reduced as a result of mediation but all path still showing significant P value. According to Baron and Kenny (1986) approach, if the beta value is reduced as a result of mediation and all variables shows significant impact, than it shows that partial mediation exits here. So in its result, H2 is supported partially. Table V shows mediated effects. The bootstrap method was also used in AMOS 24.0 to explore the extended mediation impact of QWL. By using confidence level of 95%, the sample for bootstrap was opted as 2000. To assess this mediation impact, direct and indirect effects with their p values were calculated. These findings are presented in Table VI.

Figure II: SEM-Mediation effect
Note: P-value significance (*** P < 0.01, ** P = 0.01, * P< 0.05) Results present that QWL acts to partially mediate the CSR-CFP nexus and confirmed the H2. To show both direct and indirect effects, a structural equation model was built for hypothesized mediation nexus (see figure II). Partial mediation exists if both paths are significant, and full mediation when indirect path is significant and direct path is insignificant. The Standardize Indirect effect of CSR on CFP is 0.303 with significant indirect relation (P = 0.028, so here P < 0.05), which demonstrate that CSR has indirect impact on CFP mediated by QWL and directly from CSR to CFP has (β = 0.381, p = 0.014), which means direct relation is still significant so showing partial mediation exists here.

DISCUSSION AND CONCLUSION
Previous studies have predominately centered on CSR-CFP direct and indirect relation but not employees centered. By focusing on employees' perception, an integrated study was designed that examined CSR as independent, employee's QWL as mediator, and CFP as outcome.
As hypothesized, empirical results signaled that CSR positively influences CFP in the banking sector of Pakistan. As such, CSR activities would seem to heighten the relation between banks and its staff which ultimately helps to enhance CFP. This model is consistent with prior research that perceived CSR activities can result in high financial performance by building employee's QWL (Lau & May, 1998). The scientific results depict that CSR efforts could result in generating QWL for employees. Based on encompassing literature review, this model may be first one to empirically explore the indirect impact of CSR on CFP mediated by Employee's QWL in the context of banks. A recent study found that perceived QWL is about to allocate resources inside and outside the firm for the satisfaction of employees (through CSR). Asante Boadi et al. (2020) further strengthening the above suggested relationship that CSR results in positive QWL. Further empirical results confirmed that CSR can lead to QWL generation for employees (Asante Boadi et al., 2020) .
A recent study's findings suggest that QWL can result in enhanced firm's performance (Leitão et al., 2019). This finding appends credence to this proposition that CSR has indirect positive impact on CFP. Previous researchers have asked for more exploration of CSR-CFP nexus to find causal relation of CSR activities. This notion is validated by the results of this model that banks which are socially responsible have strong ties with their employees. This Positive ties in shape of QWL partially mediate the impact of CSR on CFP. Findings further strengthen this understanding about the important role played by employee's QWL in explicating the effect of CSR activities on Bank's performance.

Theoretical Implications
The previous studies shown that CSR activities significantly influence the financial performance of banks, this study extends the existing literature through investigating the different dimensions of CSR their influence on CFP via employee's perceived QWL. In particular, a mediation model was examined and exposed that CSR enhances CFP via partial mediation of employee's QWL. By exploring this mechanism, this study examined the mediation mechanism (a new path by which CSR has a influence on CFP). To our best knowledge, this model was not empirically tested or no such study model examined by researchers yet. Secondly, this study has unique findings based on win-win paradigm to assess the influence of different dimension of CSR on employee's QWL and CFP. CSR literature has been enriched by adding evidence from the developing country regarding the effectiveness of different dimensions of CSR as main driver of the relationship between CSR and employee attitudes. Employee well-being from the lens of QWL has been emphasized which can enhance the CFP. Finally, as suggested by extant research that CSR is commonly linked with positive financial performance, it's demonstrated that the acts of satisfied employees have a direct influence on performance enhancement. Specifically, when employees highly appreciate their banks' social responsibility acts and work positively for their firm, it helps them to boost their financial performance.

Practical Implications
This study has entailments for management too. First, the final results of this model are much pertaining, crucial, and purposeful for management and policy makers. It is presented that CFP of banks can be heightened by employing CSR activities for stakeholders. These activities may comprise rendering responsible services which are environment friendly and have concerns for employees, shareholders, society, and customers of banks. Further, banks can involve their employees in making CSR related policies. This will help the banks to inculcate the views of the employees regarding the dimensions of CSR they consider pertinent in generating positive financial performance and understand the activities that are highly valued by them. Research suggests that companies must form their strategies after taking into consideration the priorities from stakeholders (Naseem et al., 2020).These initiatory activities will induce employees to have positive perspectives about the organization, and work in the best interest of banks. Ultimately, this will help to enhance financial performance. Moreover, it would induce them to work for the organization in a committed way to uplift CFP. Thus, this research model has many practical implications. The findings connote that HR manager can define a system of CSR communication for employees to have better understanding about organization's responsible behavior towards stakeholders. This study also enhances the knowledgebase by acknowledging the fact that employees QWL could be helpful to increase CFP. The management can arrange trainings for employees to let them know about the positive acts done by organization in the best interest of stakeholder and how these could be proved valuable for the bank employees. In nutshell, this current study highlights the importance of employee's attitude towards CSR in generating positive financial benefits for banks.

Limitations and future research directions
This research has various confinements that might be used as valuable source for future research. First, sampling method employed in this study is based on convenience sampling technique. No doubt, this method has been employed by researchers commonly, but it might be possible that it does not represent the full population. So, it is recommended that future researchers may consider other sampling techniques. Second, cross-sectional study design has been used to collect data at one time. Thus, there is a possibility that this model is ambiguous about definite inferences of cause-and-effect relation of CSR-CFP. Longitudinal research might be helpful to grasp more effective data over different point of times which can produce more accurate results. Third, data has been collected from banking sector which may cause generalizability problem for transferring the results of this study to other sectors. Moreover, this study is based on single country (i.e., Pakistan). As, CSR activities have been followed by many countries differently, so future research may be conducted in other countries to know cultural impact. Forth, different dimensions of CSR have been suggested by various researchers, which impact employees' behavior differently (Kim et al., 2017). Future research can address the impact of other dimensions of CSR. Finally, this study used QWL as a potential mediator between CSR-CFP, future studies are suggested other mediating and moderating variables to further understand these concepts.