The effect of corporate social responsibility on organizational commitment of employees of rural and community banks in Ghana

Researchers have mostly focused on the effect of corporate social responsibility engagement on customer behaviour. In this paper, an attempt is made at assessing the effect of corporate social responsibility engagement on employee commitment to their organizations. A self-reported questionnaire was used to collect the data from 145 employees of 50 Rural and Community Banks (RCBs) across Ghana. The study found a strong positive relationship between engagement in corporate social responsibility and employee commitment. Engagement in corporate social responsibility explained 54.1% of the total variation on employee commitment. However, tfghhis relationship is insignificant when educational level and years of working with the bank is controlled for. Gender however does not confound this relationship. This must inform decision-making regarding the planning and implementation of CSR strategies in organizations. It is acknowledged that having particular concern for the welfare of employees will boost the employee’s commitment and by extension, their performance and ultimately the growth of the Organization. *Corresponding author: Henry Kofi Mensah, Department of Human Resources and Organisational Development, KNUST School of Business, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana E-mail: hkmensah@knust.edu.gh Reviewing editor: Tahir Nisar, University of Southampton, UK Additional information is available at the end of the article ABOUT THE AUTHORS Henry Kofi Mensah is a lecturer with the Department of Human Resources and Organisational Development at the School of Business of Kwame Nkrumah University of Science and Technology, Kumasi-Ghana. His research interest spans from social responsibility, employee development, business and community interactions and Sustainability. Ahmed Agyapong is a senior lecturer with the Department of Marketing and Corporate Strategy at the School of Business of Kwame Nkrumah University of Science and Technology, KumasiGhana. His research interests are focused on issues of entrepreneurship development in the informal sector, management strategy execution for performance development among others. Dorcas Nuertey is a lecturer with the Department of Supply Chain and Information Systems at the School of Business of Kwame Nkrumah University of Science and Technology, Kumasi-Ghana. Her research interests are focused on issues sustainable supply chain management, agricultural supply chains, supply chain integration with sustainable development among others. PUBLIC INTEREST STATEMENT Corporate Social Responsibility (CSR) has received a lot of attention in recent times due to its potential of leveraging strategic economic benefits for corporations. Using the stakeholder approach and with a special focus on employees as a critical stakeholders, this study explores how employeefocused CSR trigger an improved organizational commitment within the rural banking sector of Ghana. We find a strong positive relationship between engagement in corporate social responsibility and employee commitment. The relationship however is not strong when we control for levels of education and duration of working of the employee. This provides firms with valuable evidence that, paying particular attention to the welfare of employees can in the long run boost their commitment and performance, which will ultimately result in the growth of the organization. Received: 03 October 2016 Accepted: 08 January 2017 Published: 27 January 2017 Page 1 of 19 © 2017 The Author(s). This open access article is distributed under a Creative Commons Attribution (CC-BY) 4.0 license.


PUBLIC INTEREST STATEMENT
Corporate Social Responsibility (CSR) has received a lot of attention in recent times due to its potential of leveraging strategic economic benefits for corporations. Using the stakeholder approach and with a special focus on employees as a critical stakeholders, this study explores how employeefocused CSR trigger an improved organizational commitment within the rural banking sector of Ghana. We find a strong positive relationship between engagement in corporate social responsibility and employee commitment. The relationship however is not strong when we control for levels of education and duration of working of the employee. This provides firms with valuable evidence that, paying particular attention to the welfare of employees can in the long run boost their commitment and performance, which will ultimately result in the growth of the organization.

Introduction
Corporate social responsibility (CSR) is a primary avenue for businesses to respond to the social needs of people in the environment in which they operate. Various schools exist with respect to the motive behind a firm's CSR practices. The first and possibly the most popular one recognizes CSR practices as means of improving customer patronage (Pirsch, Gupta, & Grau, 2007). Another famous school considers CSR practices as an ultimate way to give back to society (Pirsch et al., 2007). Proponents of a third school are also of the view that engagement in CSR activities is associated with the goal of impressing customers and consequently improving patronage, and contributing to environmental and social progress (Santoso, 2014;Tuzcu, 2014). CSR activities are often known to be aimed at customers and community members. Of course, every firm engages in CSR to primarily or partly impress customers and potential customers in their immediate environment. Yet some firms recognize employees as the closest customers who must be impressed through CSR activities (Santoso, 2014;Tuzcu, 2014). Most studies have focused largely on CSR for external stakeholders and how that impact performance, with very little attention on employee (Aguilera, Rupp, Williams, & Ganapathi, 2007;Gond, El-Akremi, Igalens, & Swaen, 2010). With recent increasing concern about high employee turnover, employee absenteeism and employee low motivation towards work and organization, Ali, Rehman, Ali, Yousaf, and Zia (2010) and other studies have suggested that CSR can build strong employee bond with corporations and assist to achieve better employee and organizational commitment and ultimately performance. Servaes and Tamayo (2013) observed that CSR for employees are relevant to enhancing employee commitment, satisfaction and performance. It is agreeable that a firm's ability to savor CSR towards improved business performance is based on employee commitment and performance. Hence, it could be a good idea to engage in CSR for employees given that they are an important pillar in every business. Considered as the immediate internal stakeholders, employees are a crucial component of the target group for most CSR agenda. Crane, Matten, and Spence (2008) emphasized that CSR issues relating to employees relate to the internal functioning of corporations regarding workforce issues including working conditions, health and safety, equal opportunity, remuneration and benefits, and others.
According to Aguilera et al. (2007), these have some positive outcomes on employee satisfaction and for that matter their organizational commitment. Yet at least in a Ghanaian context and in terms of Rural and Community Banks (RCBs), it remains empirically unconfirmed that CSR for employees positively influence employee commitment. Studies on CSR in Ghana have characteristically concentrated on multinational enterprises and scoped around environment and community development for extractive sector and CSR reporting and customer loyalty among universal banks (Abugre & Nyuur, 2015;Boon & Ababio, 2009;Hinson, Boateng, & Madichie, 2010;Ofori, Nyuur, & S-Darko, 2014), RCBs in Ghana are known for their engagement in corporate social responsibility in rural areas in Ghana. A content analysis of reports of their Annual General Meetings (AGM) reveals that these banks appreciably engage in CSR for employees. Rationally speaking, these banks are engaging in CSR for employees to enhance employee commitment, satisfaction and performance. But as to whether or not the engagement in CSR has adequately reflected in an increased employee commitment to the banks, has not been verified. This study therefore assesses the effect of employee-related CSR on employee commitment with a focus on RCBs in Ghana.

Literature review
Corporate Social Responsibility is increasingly being adapted to enhance firm and customer value. Though CSR research has assumed global attention in recent times, lingering in many studies are issues regarding what constitutes CSR, what are the firm's target stakeholders and the facets of society that benefit from a firm's in CSR endeavour (Maignan & Ferrell, 2004).

Strategic corporate social responsibility
It is admitted that firms expend financial resource on their CSR engagement and that could be a potential drain on profit maximization. It must however be emphasized that a firm's CSR engagement is also embedded in its quest to maximize the largest benefits for both the firm and the society at large. This is extensively discussed in strategic CSR which aims at achieving large and distinctive social and economic benefits from a strategically focused set of initiatives (Gill, 2007). More essentially, Porter and Kramer (2006) explained that the core of CSR must be whether a cause presents an opportunity to create shared value that is of a meaningful benefit for society and at same time valuable to the business alike. Initiating the discussion on Strategic CSR, Porter and Kramer (2006) asserted that social and economic issues by themselves create markets and market opportunities, and that companies, by finding solutions to and solving social problems, will go a long way to make profits and advance their reputations.
They further explain that, in order for CSR to be strategic, it has to be seen to be contributing to firm value-chain practices and generally, to the improvement of the firm's competitiveness. They give the example that Strategic CSR activities must help corporations to secure purchased inputs, reduce operational costs, ensure a smooth inbound or outbound logistics and generally, contribute to the marketing and sales function of the value chain of the corporation. These activities are largely performed by the employees of the firm. If these employees are not given a focus in the firm's CSR endeavour, their commitment to working hard may be affected and that is where the firm's profit maximization goal will be affected. Firms also stand the potential benefit of attracting top talent and capital (Kramer, 2011;Schreiber, 2011) when they engage in CSR activities particularly those that tends to satisfy employees and trigger improved organizational commitment.

The stakeholder view of corporate social responsibility
The stakeholder view of CSR is perhaps suggested as an alternative to the classical economic view of shareholder theory espoused by Friedman (1970). To Friedman, the only responsibility of business is to maximize profit for the shareholder. On the contrary, Freeman (1984) suggested that the responsibilities of business are rather those that are upheld towards those who affect or are affected by a decision or an action of a corporation popularly referred to as stakeholders. To Donaldson and Preston (1995), a stakeholder is a person or group with legitimate interests in procedural and/or substantive aspects of corporate activity. Hence, persons, groups, neighbourhoods, organizations, institutions, societies and even the natural environment could generally qualify as actual or potential stakeholders since they could all be affected or can affect the business operations (Mitchell, Agle, & Wood, 1997).
It is worth noting that, in the view of the proponent of this view, the socially responsible business entity is the one with a commitment to contribute to economic development, improving the quality of life of the workforce, their families and the local community in general (WBCSD, 1999). Identified stakeholder groups include among others shareholders, customers, employees and local communities. These are what have been described in literature as primary stakeholders which include those whose continued participation in the corporation is crucial for its survival (Clarkson, 1995). For instance, while firms focus their CSR on customers to strategically leverage loyalty and enhance customer and firm value (Garriga & Melé, 2004), a focus on community-related CSR is largely philanthropic in motive and are geared towards paying back to society where the firm operates especially where there is a prevalence of risks to which members of the community are exposed to as a result of the activities of the business (Albdour & Altarawneh, 2012).
The focus of this paper is on employee as stakeholders in firm's activities. Employees' role in CSR span from pushing corporations to adopt socially responsible behaviour; designing and implement effective CSR programmes and policies among others (Gond et al., 2010). It is ironic to note that notwithstanding the major role employees play in planning and implementation of firm's CSR engagement, less attention is given to those CSR that related to them. For most studies, the introduction of employees as a critical stakeholder has been theoretical. While some studies consider employees as "an independent stakeholder variable" seeking to explain the emergence of CSR (Aguilera et al., 2007), others recognize them as "dependent stakeholder" influenced by CSR and as a "mediating stakeholder" of CSR seeking to influence on corporate performance (Maignan & Ferrell, 2004).
CSR for employee groups (referred to in this paper as employee related CSR) is normally premised on the firm's motive of rewarding its employees in a special way using appropriate CSR models. Employees may therefore be offered special incentives, motivational packages, job designs and a fair organizational system that includes assurance of organizational justice. Some studies (e.g. Santoso, 2014;Tuzcu, 2014) contended that CSR for employees are relevant to successfully implementing CSR for the other stakeholders. This is logical because employees play a leading role in implement CSR for community and customers, hence CSR would make better impact if it is first geared towards Employee Organizational Commitment (EC).

Employee organizational commitment and CSR
Three component of organizational commitment has been featured prominently in Organizational Commitment literature. Meyer and Allen (1991) identify them as affective, normative and continuance commitments. In it earliest and simplest explanation O'Reilly (1989) defined Organizational Commitment as a psychological bond, which an individual employee has with the organization. This may include an employee's loyalty to the organization, readiness to adopt values and goals of the organizational; fulfilment of his/her job responsibilities, among other (Slack, Orife, & Anderson, 2010). In most cases, employees who have high levels of job satisfaction have a high propensity to be commitment to their organization. Notwithstanding their satisfaction level, it is believed that a committed employee is more likely to uphold and pursue the goals and values of his/her organization. Meyer and Allen's (1991) three-component model of organizational commitment posits that employees experience commitment in terms of affective, normative and continuance commitments. Jernigan, Beggs, and Kohut (2002), explains that affective commitment refer to an employee's attachment to, identification with and involvement in the organization. It encompasses the emotional ties, which an individual develops in relation to the organization primarily as a result of positive work experiences as explained by Meyer and Allen (1991). Normative commitment has also been explained as the feeling of obligation to continue to work for the organization (Jernigan et al., 2002). According to Slack et al. (2010), this is premised on building a sense of duty and value, and the degree to which an individual feels obligated to stays in an organization. Lastly, as explained by Continuance Commitment accrue from a perceived cost (both economic and social) associated with leaving the organization Meyer, Stanley, Herscovitch, and Topolnytsky (2002). These components of commitment have been described largely to be the predictors of employee behaviour and intentions at the organizational level (Meyer et al., 2002) and in the view of the current study not exclusive of employees of RCBs in Ghana.
In an attempt to explain the relationship between CSR and commitment, Gond et al. (2010) premised this relationship on social exchange. In their view, the reactions of employees to CSR initiatives are largely governed by reciprocity and generally described as a pattern of mutually contingent exchange of gratifications. Within an organization, employees under the circumstance of CSR feel obligated to reciprocate the positive treatment given to them by the organization (Gond et al., 2010). On the other hand, Greenborg (2002) cited in Gond et al. (2010) intimated that a dissatisfied employee, who feel unfairly underpaid is more likely to show noncommittal to the achievement of the organization's goals. A considerable number of empirical studies (e.g. Madison, Ward, & Royalty, 2012;Santoso, 2014;Tuzcu, 2014) has also confirmed that EC is positively affected by CSR for employees. On the basis of this evidence, researchers are of the view that CSR for employees contribute to an enhancement of EC. This empirical evidence however does not embrace some sectors, specifically the Rural and Community Banking sector in Ghana, though CSR engagement in here is considerable. It is believed that this evidence is relevant to decision-making among the RCBs and would contribute to an expansion of the existing literature. It is further argued in this study that some background variables (e.g. educational level, gender, work experience or years of working with bank, etc.) can confound the relationship between CSR-employees and EC. By implication, gender, educational level and work experience need to be controlled for testing the CSR-EC relationship. In view of these arguments and lessons taken from the literature review, the following hypotheses are tested in this study: H1: CSR-employees make no significant effect on employee commitment in the selected RCBs H2: CSR-employees make no significant effect on employee commitment in the selected RCBs The relationship between CSR-employees and employee commitment is still significant even after the gender is controlled for H3: The is a significant relationship between CSR-employees and employee commitment when years of working with bank is controlled for H4: There is no relationship between CSR-employees and employee commitment is when educational level is controlled for Though other variables such as job description and employee level can also serve as control variables in the CSR-EC relationship, they are not captured in this study. Testing for their confounding effects on the CSR-EC relationship is left for future research.

Methods
The study applied a cross-sectional research technique to collect primary data on 50 selected RCBs in the 10 regions of Ghana. We selected and focused on RCBs known to engage in CSR in their communities. The study population was employees of the selected RCBs. Employees selected were those who had worked in the selected RCBs for at least 2 years. We selected employees who had worked in the banks for at least 2 years to ensure that participants had been exposed to a considerable period of CSR. Of course, assessment of the effect of CSR on employees requires employees' exposure to CSR for a considerable period. We used top management members because these employees had in-depth knowledge on CSR for employees and better satisfied the selection criterion. Three employees each were selected from 50 RCBs making a convenient sample of 150 employees. The employees were selected based on satisfaction of the criterion identified above.
In all, two variables were measured in this study. The independent variable (IV) is CSR practice for employees (i.e. CSR-employees) and the dependent variable (DV) is Employee Organizational Commitment. The control variables are gender, educational level and work experience. Employee Organizational Commitment was measured using items borrowed from the study of Adekola (2012). CSR activities for employees were measured using items borrowed from the study of Khan and Jan (2015). The control variables were measured by assigning values to their levels as follows: Gender-Male (1); Female (2); Educational level-Basic/secondary (1); Diploma (2); Degree (3); Master's degree (4); PhD or higher (5); Work experience-Up to 2 years (1); 2-4 years (2); 5-7 years (3); 8-10 years (4); and above 10 years (5).
CSR-employees and OC were measured using a Likert scale which allowed the participants to respond on a scale of 1-5 in indicating their extent of agreement or disagreement to each CSR activity or item. The scale of the Likert scale includes: 1 = strongly disagree; 2 = disagree; 3 = not sure; 4 = agree; and 5 = strongly agree. In coding however, not sure was corresponded to 0, since it represents neutrality and uncertainty.
Out of 150 questionnaires sent out, 145 were retrieved and deemed duly completed, representing a 97% response rate. Descriptive statistics (including skewness and kurtosis) was used to check data for outliers and then dependent variable was tested for data normality (see results in Table 3 and Figure 1). A confirmation of data normality made it possible to use the parametric statistical tools, precisely ordinary least square regression analysis.
The first research hypotheses were tested using ordinary least square regression analysis. The three remaining hypotheses were tested using multiple linear regression analysis. Yet in testing the first hypotheses, we used Factor Analysis, precisely Principal Component Analysis (PCA) to assess the relationship between items of CSR for employees and OC. The use of PCA compliments results reached using ordinary least squares regression analysis. Thus, PCA was employed as a robustness test of the first and primary hypothesis.

Results
In this section, data are analysed to address the research hypotheses. Firstly, PCA is used to explore the first hypothesis. Theoretically, the validity of results of the PCA is based on various statistical indicators nested in the PCA. The first of these indicators is the correlation coefficients formed by each pair of the indicator variables. The rule of thumb is that a high number of these coefficients must be greater than .3 (Ringnér, 2008;Suhr, 1999). In other words, no pair or just a few pairs of indicator variables should have a correlation coefficient less than .3. In Appendix Table A1, this requirement is met. It is therefore more likely that the PCA is sufficiently valid and would therefore give rise to valid principal components. However, the validity of the PCA depends on the results of other diagnostic tests such as those shown in Table 1.
In Table 1, the value corresponding to the Kaiser-Meyer-Olkin Measure of Sampling Adequacy (MSA) is required to be greater than .50 (Suhr, 1999), whereas higher values are better (Ringnér, 2008;Suhr, 1999). Moreover, the Bartlett's Test of Sphericity is required to be significant at 5% significance level (Ringnér, 2008). From Table 1, both requirements are met, further buttressing the validity of the PCA and its resulting principal components. In Appendix Table A2, the MSA value of all variables is represented in the leading diagonals (i.e. values in bold). It can be observed that each of these values is greater than .50, therefore the MSA requirement is satisfied for each indicator variable as well. It is worth saying that the MSA value in Table 1 is for all indicator variables, whereas those in Appendix Table A2 are for the individual indicator variables.
Based on statistical evidences produced in Appendix Table A1, Table 1 and Appendix Table A2, the PCA is sufficiently valid and would therefore produce valid principal components. In Appendix Table  A3, communalities of all indicator variables are shown. Generally, the size of the communalities or extraction values is the basis for removing indicator variables from an iteration of the PCA. An indicator variable is removed when its communality is less than .5 (Suhr, 1999). In the first iteration of this analysis, none of the variables has a communality value less than .50. This result suggests that all indicator variables of CSR-employees and EC are retained in the PCA.
Appendix Table A4 shows the number of component formed, the variance explained by each component and total variable explained. In Appendix Table A4, Total under Initial Eigenvalues is an important statistic used to identify components formed. A component extracted must produce a Total Eigenvalue of at least 1 (Suhr, 1999). On the basis of this criterion, four components were formed by both the indicator variables of CSR-employees and EC. The first component explains a variance of 57.3% of the total variance, whilst the second explains a variance of 8.2% of the total variance. The third component explains 6.4% of the total variance. The fourth component account for 4.6% of the total variance. The four components thus explain 76.5% of the total variance.
In Appendix Table A5, the first component includes a majority of the indicator variables of CSRemployees and EC (i.e. CSR 3 , CSR 4 … EC 7 ). This means that the first component represents highly correlated variables of both CSR-employees and EC. This is the first ultimate statistical evidence of the relationship between CSR-employees and EC. The second component is made up of only one indicator variable of EC (i.e. EC 8 ), whereas the third component is constituted by the first two indicator variables of CSR-employees (i.e. CRS 1 , and CSR 2 ). The fourth component is made up of only one indicator variable of CSR-employees (i.e. CSR 11 ). The positive relationship between CSR-employees and EC is buttressed by evidences in Table 2.
With reference to Appendix Tables A4 and A5, it could be observed that a greater part of the total variation explained is accounted for by the first component. This evidence can be verified from Appendix Table A5 and is indicated by the fact that all indicator variables are strongly positively related to Component 1 (see component loadings formed by component 1 and indicator variables in Appendix Table A5). Moreover, against components 2, 3 and 4, each indicator variable is relatively weakly related (see Appendix Table A5). The strongest evidence regarding the positive relationship between CSR-employees and EC is shown in Moreover, components 2 and 3 are positively correlated (R = .533), likewise components 2 and 4 (R = .362). Components 3 and 4 are also positively correlated (R = .457). It is therefore evident that components formed by indicator variables of both CSR-employees and EC are positively correlated. The first null hypothesis is therefore not confirmed. The alternative hypothesis is therefore provisionally accepted.
The second null hypothesis is tested using OLS regression analysis as follows. Yet before looking at the results of the OLS regression analysis, there is a need to ensure that the data came from a normally distributed population. This verification is done using statistics in Table 3. Table 3 shows two tests of normality of data. The first test, Kolmogorov-Smirnov, is only applicable for large samples of at least 2,000 cases. Considering the fact that the sample size of this study is less than 2,000, we would focus on the second test, Shapiro-Wilk. This test is based on the 5% significance level. It can be seen that the p-value of the Shapiro-Wilk test is less than the chosen level of significance (p < .05), suggesting that data did not come from a normally distributed population. But the deviation from normality is not serious, considering the Box Plot of Figure 1. Moreover some researchers (e.g. Sawilowsky, 2005) observed that such moderate deviation from normality is acceptable. We therefore decided to proceed with the test of the second hypothesis using OLS regression analysis.
In Table 4, there is a strong positive correlation between CSR-employees and EC (R = .736). Figure  2 shows the line of best-fit associated with the relationship between EC and CSR-employees and EC. It can be observed that this line is a perfect straight line, with a variation of 54.1% explained by CSRemployees confirmed. In Table 4, this total variation explained by the predictor is confirmed; thus CSR-employees explain 54.1% of the variation. The error term of the regression model therefore explains 45.9% of the total variation. It is therefore deemed that our model moderately fitted. In Table 5, the ANOVA test is significant at 5% significance level (F = 168.66, p = .000). Significance of ANOVA test is the evidence to the prediction of EC from CSR-employees. In Table 6, CSR-employees significantly predicts EC at 5% significance level (t = 12.99, p = .000, β = .367). In addition, a unit change in CSR-employees changes the conditional mean of EC by .367 within a confidence interval of .311-.423. The relationship between CSR-employees and EC is expressed as follows: EC = .367 * CSR-employees + 7.347  The positive coefficient (that is .367) produced in the mathematical model supports the positive correlation between the two variables as seen in Table 4. It is therefore evident that enhancing the level of engagement in CSR activities for employees increases employee commitment. The second null hypothesis is therefore not confirmed and the alternative hypothesis is provisionally accepted. Though CSR (for employees) significantly predicts Employee Commitment in the selected RCBs, there is the need to control for background variables that are likely to influence this relationship between EC and CSR-employees. As stated earlier, controlling for these variables is aimed at eliminating alternative effects on EC. Table 7 shows a model summary of the prediction of EC from CSR-employees. From Table 7, CSR and the three background variables (Gender, Educational Level, and Years of Working in the Bank) account for a 56.1% of the total variation in EC, with a residual variation of 43.9% accounted. In Table 4, CSR alone accounts for 54.1% of the total variation. This means that the three background variables explain just .02% of the total variation. Moreover in Table 8, the Analysis of Variance is significant at 5% significance level [F (4, 131) = 41.86, p = .000], suggesting that EC can be expressed as a linear combination of CSR and the three background variables.  In Table 9, CSR-employees significantly predicts EC at 5% significance level (t = 12.62, p = .000, β = .349). Gender also significantly predicts EC at 5% significance level (t = −2.52, p = .000, β = −1.89). Educational level and "Years of working with bank" however fail to predict EC at the same level of significance (p > .05). Though gender significantly predicts EC, the general effect of the background variables on the relationship between CSR-employees and EC is scanty and almost negligible. For instance, with respect to Table 4, the background variables, especially Gender, account for just 5% of the effect of CSR-employees on EC. Yet this small influence of the background variables cannot be totally ruled out, though it is largely contributed by Gender.

Discussion of findings
In our Factor Analysis, all items of employee commitment and CSR-employees are retained, with 76.5% of the total variation explained by the variables retained. Since the total variation explained is greater than 50% (variation > 50%), there is ample evidence of a strong relationship among the retained variables (Suhr, 1999). In addition, the components formed by items of EC and CSRemployees are significantly correlated (see Table 2). The Factor Analysis therefore produces the first evidence in support of the hypothesis that engagement in CSR for employees contributes to enhancing employees' organizational commitment. Yet the regression analysis more precisely confirms this hypothesis by virtue of revealing that CSR-employees is strongly correlated to EC (r = .736, p < .05) and contributes 54.1% of the total variation on EC. In addition, CSR-employees significantly predicts EC at 5% significance level (t = 12.99, p = .000, β = .367).   This confirmed positive relationship between CSR-employees and EC is empirically and theoretically consistent. From an adaptation of the stakeholder and integrative theories of Maignan and Ferrell (2004), EC is an organizational reward for engaging in CSR for internal stakeholders of a business, which are employees. A dimension of this reward is CSR's positive impact on organizational commitment of employees. Empirically, this confirmed relationship is consistent with several previous studies (e.g. Madison et al., 2012;Santoso, 2014;Tuzcu, 2014) conducted in various sectors. Therefore, the positive effect of CSR for employees on organizational commitment both from theoretical and empirical perspectives is consistent with the selected RCBs.
The insignificant effect of gender on the CSR-EC relationship suggests that CSR impacts the commitment of males and females at the same extent. This finding means gender does not significantly affect the relationship between CSR and EC. It controverts the argument of some researchers (e.g. Khan & Jan, 2015;Madison et al., 2012) that impact of organizational variables, activities or procedures on employees would differ across employee groups due to differences in these groups in terms of expectations, organizational roles, motivation, and other related variables. Thus, this argument is not supported by the gender group variable. Education and Tenure however support it. This means that the relationship between CSR and EC is as a result of the influence of education and years of working in the bank. Similarly, education and years of working in the bank influences both perceptions about CSR and employees' commitment, whereas gender does not. As a result of lack of research work on these control variables in this context, these findings are not backed by any study, but they provide a basis for future research.

Conclusion and recommendation
Items of employee organizational commitment and CSR-employees are all retained in the Factor Analysis, suggesting that items of the two constructs are significantly related. Moreover, factors formed by the two constructs are significantly positively correlated. The ordinary least squares regression analyses confirm a strong positive relationship between EC and CSR-employees. Results of both the FA and OLS regression analysis therefore confirm the hypothesis that engagement in CSR for employees enhances employees' organizational commitment. Yet, the relationship between CSRemployees and EC is influenced by the background variable of educational level and years of working with the bank. Thus, the CSR-employees and EC have no relationship when these variables are controlled for. However, the relationship between CSR and EC is still significant when gender is controlled for. It is therefore concluded that CSR-employees makes a significant effect on EC even when gender is controlled for. This relationship is not significant when education and work experience are controlled for. This means that this relationship is based on the effects of education and years of working with the bank. The findings reported in this study have reinforced the argument that CSR generally has a significant influence on employee's commitment towards his organization. However, the study also indicates that this relationship is premised on some background variables including gender, educational level and years of working. This should inform decision-making regarding the planning and implementation of CSR strategies in organizations. RCBs in particular should integrate CSR strategies with their human resource policies and must acknowledge that having particular concern for the welfare of employees goes a long way to boost the employee's commitment and by extension, their performance and ultimately the growth of the organization. To boost organizational performance therefore, managements would have to enhance and maximize their engagement in CSR for employees. According to Santoso (2014), this may demand that managements offer employees better conditions of service, a fair organizational system and a family-oriented organizational environment.

Significance of the study and limitations
CSR for employees is seldom mentioned in the literature, let alone its effect on organizational commitment. It is therefore hoped that this study will contribute to academic debate on CSR for employees and its effect on EC. This paper also tests and supports the importance of CSR for employees. The main limitation of this study is the fact that the sample size applied was relatively small. Theoretically, PCA produces the best results if the sample size is at least 500 (Ringnér, 2008). The fact that the sample size was less than 500 implied that findings yielded in the PCA do not represent the best possible outcome. Future researchers are therefore encouraged to consider using a larger sample.  202 .174 .150 .073 .131 .192 .090 .198 .197