The study aims to inspected the associations between different forms of corporate social responsibility (CSR) and banks' performance (BP).
Data were collected from 275 respondents with a structured questionnaire survey. The collected data were analyzed using partial least squares structural equation modeling (PLS-SEM) with the support of SmartPLS software version 3.0.
The statistical outcomes expose that all categories of CSR are positively associated with a bank’s financial performance, except CSR to the environment. Statistical findings reveal that CSR to employees was found to be the most significant, followed by CSR to community and CSR to customers in influencing BP.
The research contributes to the CSR literature as it represents a realistic indication of the impact of CSR practices on BP in Bangladesh.
1. Introduction
Corporate social responsibility (CSR) practices have become a significant field of study over the last few decades (Hart, 2010; Nofsinger and Varma, 2014; Jeriji and Louhichi, 2020), and organizations now develop policies to spend more on CSR practices. At the same time, researchers and practitioners are putting forth considerable efforts to explore the effect of CSR practices on organizational performance (Kim et al., 2010; Giannarakis et al., 2017). Morales-Raya et al. (2019) and Scott (2007) stated that corporate social contribution builds a better reputation for the organization in the society that helps attracting new graduates. Therefore, CSR is an integral and indispensable part of the long-term business, sustainable growth and success, and it plays a vital role in promoting values both home and abroad. The banking sector in Bangladesh has a glorious past of getting involved in diverse kinds of social activities, which are properly known as CSR like assistance to various groups, city beautification, patronizing cultural programs, etc. However, there is fabulous progress in this envelope since 2008; the Bangladesh Bank Guideline advised the banks to involve themselves in these events in a formal way (Saha, 2018; Mahbuba and Farzana, 2013). Though it is a popular research topic in developed countries, surprisingly in Bangladesh, very few research works have been done on the CSR practices of the banking industry. Roy et al. (2017) and Rahman et al. (2019) tried to explain CSR and the related matter and scrutinize the well-being performances of some state-owned commercial and private commercial banks in Bangladesh. They attempt to discover the degree to which the mentioned banks in Bangladesh accomplish their CSR activities and to search the level of disclosure for CSR in their annual reports. Kamrujjaman and Uddin (2015) and El-Diftar et al. (2017) found a strong correlation between CSR expenditure and sample banks’ deposit, loan and profitability. Another study by Hamid (2016) found that the contribution amount by the Non-Bank Financial Institutions (NBFIs) to CSR activities is very insignificant in proportion to their profit amount. Hossain and Hoque (2005) and Alipour et al. (2019) found a significant statistical difference in sales revenue and in employee size between companies having environmental disclosure and companies not having environmental disclosure in Bangladesh. Masud (2011) and Syed and Butt (2017) found that CSR practices are quite scant rather than profitable. Therefore, very few studies have examined CSR policies and practices in the banking sector of Bangladesh. This study advances the field by examining the impact of CSR from the perspective of four groups: the employees of the banks, and the banks’ customers, community and environment on banks’ performance (BP). Therefore, the present study aims to examine the relationship between CSR to different stakeholders and BP and provide recommendations for policymakers and regulators on CSR practices that might help to improve BP.
2. Literature review
2.1 CSR and corporate financial performance
The connectivity of CSR with financial success has been developed to analyze in the various theories. Others who have established a relationship between CSR and financial performance have discovered that organizations with a greater degree of social responsibility pay higher expenses, placing them at a competitive disadvantage relative to their less socially aware peers (Masum et al., 2021; McGuire et al., 1988). On the other hand, other experts feel that CSR and financial success are inexorably related. Numerous authors have seen an increase in employee and client satisfaction as a result of social responsibility (Davis, 1975; Nuskiya et al., 2021). Socially responsible operations may help a company’s reputation among important stakeholders such as bankers, investors and government authorities. Economic benefits may be realized by strengthening ties with these group constituents (Wijesinghe and Senaratne, 2011). Nonetheless, financial organizations and other investment firms have stated that social considerations influence their investment decisions significantly (Spicer, 1978). Investors may see organizations that are less socially conscious as riskier investments since they feel the management team has the skills essential to operate the business properly (Oliveira et al., 2019; Alexander and Buchholz, 1978; Spicer, 1978). For instance, a business’s financial performance may have an effect on its social policies and actions (Ullmann, 1985).
2.2 CSR to employees and banks’ performances
According to Lee and Miller (1999) and Biswas et al. (2018), executing a strategy requires the involvement of personnel, an organization’s human capital. Around 50% firms of global giant firms understand that personnel appreciation is an important driver of CSR (Australia, 2014). According to Pour et al. (2014), various levels of human resources have been defined as value job that results in the organization’s key success factors (Siueia et al., 2019; Post et al., 2011). Rego et al. (2010) conducted a study of the literature and concluded that a company’s CSR activities improve both its employees and bottom line. Additionally, CSR advantages include increasing productivity and minimizing absenteeism and turnover (Peterson, 2004; Glavas and Kelley, 2014). Another advantage of managing CSR is that if an employee is more committed to their job, the company’s performance is likely to increase (Farooq et al., 2014). There is a correlation between CSR efforts and employee commitment and an organization’s attraction to prospective employees (Turker, 2015; Greening and Turban, 2000; Rouf and Hossan, 2020). Organizations may increase employee morale and productivity by using successful internal marketing tactics. Employees who are satisfied with their employment and feel a feeling of loyalty to their employers are more likely to talk positively about their employers in their social networks (Du et al., 2010). According to Babiak and Wolfe (2009), the most important measure of an organization’s success is its dedication to its employees and its CSR. Another study conducted by Azim (2016) demonstrates that CSR for employees has a positive effect on organizational performance. Additionally, organizational performance has been associated with factors such as organizational climate (Welsch and LaVan, 1981) and the variability of organization-person fit (McDonald and Hung Lai, 2011; Reichers, 1986). According to Turker’s (2015) empirical study, CSR toward employees has a positive effect on a firm’s performance. According to Ali et al. (2010), employee CSR has been found to have a positive effect on organizational performance. The following hypothesis is being developed in this research.
Bank performance is positively related to corporate social responsibility toward employees.
2.3 CSR to customer and banks’ performance
Consumers are indeed among the organization’s most influential stakeholders. Product safety, customer service and the management of consumer discontent all come under this area, as does a firm’s obligation to its customers and goods. Customers are important to every business’s success, which is why organizations prioritize cultivating and maintaining great connections with them. External CSR initiatives, according to a study, have a significant influence on a company’s overall performance (Brammer et al., 2007; Garanina and Aray, 2021). According to Fornell et al. (2006), a firm’s long-term profitability and market value are inextricably related to customer satisfaction. Luo and Bhattacharya (2006) addressed exactly three lines of inquiry regarding the environmental performance issues. As a starting point, both organizational and stakeholder theories assert that firms’ function indicates customers' complex identity a member of a family, as a member of society, along with a member of a nation (Handelman and Bello, 2004). The second advantage is that when a firm has a great CSR track record, its positive brand image is boosted (Gurhan-Canli and Batra, 2004). Thirdly, consumer empowerment has evolved into a kind of social engagement. Customer rights, providing high-quality goods and services and disseminating accurate and relevant information are all included in the frameworks for CSR (Davenport, 2000; Wood, 2010). According to the study of Brown and Dacin (1997) and Anderson and Bieniaszewska (2005), CSR activities have been found to have an effect on the satisfaction of actual customer, creating an organization’s value in the market. On the other hand, Omar et al. (2019) have shown that CSR to customers has an effect on organizational performance. Prutina (2016) also said that CSR programs focused on customers have a significant influence on an organization’s success. Wolf et al. (2014) explored a remarkable affirmative effect on an organization’s performance with CSR. Consequently, we can hypothesize in the following way.
Bank performance is positively related to corporate social responsibility toward customers.
2.4 CSR to environment and banks’ performance
Numerous external elements have a substantial influence on an institution’s functioning, both socially and financially. Acar and Temiz (2020) conducted empirical analysis on corporate environmental performance and environmental disclosure in an emerging market context: socio-political theories versus economic disclosure theories. International Journal of Emerging Markets evaluated ten research and development (R&D), assembly and production offices and demonstrated how each office’s external environment had a direct effect on the division’s organization, administration and financial outcomes. Williamson (1981), for example, advocates for an emphasis on transaction costs. In light of their results, it is recommended that a firm’s resource exchange system be based on the interactions between the enterprise and its environment. There is a public eagerness on protecting the environment, which is widely considered (Mazurkiewicz, 2004). Mishra and Suar (2010) addressed that there are three major factors that manage the variability of organizational performance; as such, innovation in product includes the manipulation ingredient utilization, process technologies (which include such efficient production processes) and management practices (such as auditing processes). According to a study, efficient environmental management may improve a firm’s market value, image and economic position (Klassen and McLaughlin, 1996; Kilincarslan et al., 2020). Both Davenport (2000) and Wood (2010) included environmental activities in their CSR plans to demonstrate their commitment to sustainability and environmental challenges. Elsayed and Paton (2005) and Porter (2008), both prominent researchers, pinpointed respectively that the betterment of atmosphere or saving both company and giant community – in short, the equality between corporate success and environment issues mentioned by Palmer (2012). It has been suggested that financial success is correlated with social or environmental success (e.g. in Elsayed and Paton, 2005; McWilliams and Siegel, 2001). Given the increasing public knowledge of environmental concerns over the last several years, persons who work for environmental organizations may feel an enhanced sense of duty. Additionally, according to Social Identity Theory (SIT) literature, comparing a business to its competitors may result in an increase in the firm’s production (Turker, 2009). CSR initiatives aiming at environmental protection have been demonstrated in empirical research to be positively connected with an organization’s performance (Rehman et al., 2022). When employees embrace CSR values, organizational performance increases, and this effect is amplified when the whole workforce adheres to CSR principles. Environmentally responsible CSR may also have an effect on how workers see CSR (Prutina, 2016). As a result, it is predicted that,
Bank performance is positively related to corporate social responsibility toward environment.
2.5 CSR to community and banks’ performance
The community aspect is a big part of CSR, and it’s broken down into sub-elements like education, health, housing and security. When a firm’s size, geographical breadth and complexity grow, Trigilia (2001) and Nuskiya et al. (2021) found that it pays more attention to the patron of exercising societies, which may help an organization perform better. Furthermore, Husted (2003) explored that when the organization has completely fulfilled its corporate social functions to the various stakeholders and focuses on the interested ones who are directly involved in the organization, they are highly appreciated for the organizational profitability and sustainability. On the other hand, Berman et al. (1999) and Mishra and Suar (2010) asserted that earlier research shows a negative association between CSR and company performance. Tsoutsoura (2004), Gilley et al. (2018) and Brine et al. (2007), as depicted in their numerous studies, have suggested that the effect of a company’s links to the community on its performance is difficult to assess. As a result, Porter (2008) claimed that the link between community ties and CSR and performance is less critical. Omar et al. (2019) showed that CSR to the community has an effect on employee loyalty to the business. According to research, a company’s CSR policies, especially its external ones, have a significant influence on its performance (Brammer et al., 2007). Turker (2009) examined the impact of CSR efforts directed towards external stakeholders on employees, building on the results of firm performance. As a result, a hypothesis is stated as follows.
Bank performance is positively related to corporate social responsibility toward the community.
3. Methodology
3.1 Identifying population
The objective of the study is to find the relationship between CSR practices and the different stakeholders and BP in the banking industry in Bangladesh. This section briefly covers the research areas, in particular, banking industry in Bangladesh, which is categorized into two major types, namely scheduled banks and non-scheduled banks. There are 59 scheduled banks conducting activities in Bangladesh under the supervision of the central bank, namely Bangladesh Bank Order 1972 as well as the Bank Company Act 1991 (amended up to 2013). Scheduled bank is further broken down into four categories as follows:
State-owned commercial banks (SOCBs): There are six SOCBs, which are fully or majority owned by the government of Bangladesh.
Specialized bank (SDBs): Three SDBs are now operating, which were established to fulfill specific purposes like agriculture or industrial development.
Private commercial banks (PCBs): There are 43 PCBs that are majority owned by private entities. PCBs are also categorized into two groups.
Conventional private commercial bank: There are 33 interest-based PCBs now operating in the banking industry.
Islamic Shariah-based private commercial bank: There are 10 Islamic Shariah-based PCBs in Bangladesh, and they are operating banking activities according to Islamic Shariah-based principles, i.e. profit-loss sharing mode.
Foreign commercial bank (FCBs): There are nine FCBs doing business in Bangladesh as the branches of the banks, which are incorporated abroad.
Based on this context, the population of this study is referred to as the private commercial banks of Bangladesh. The researcher excluded SOCBs and SDBs and FCBs because of less involvement in CSR, and the study focused more on third-generation banks other than the second-generation. The total number of private commercial banks is 41, of which 33 are conventional banks and 8 are Shariah-based banks. The executives (branch managers) of banks are considered as respondents because they are knowledgeable and have sufficient knowledge about CSR practices and the impact of CSR on bank performance. Since there are different categories of banks, it is quite difficult to study them individually. All the branches of banks are situated all over the country. For the researchers, it is almost impossible to collect data from all the branches due to time, budget and accessibility. The researchers attempt to cluster them based on geographical area, and they only focus on the branches that are situated in Dhaka city. In Dhaka, all the PCB branches are available.
3.2 Sampling frame
A good number of samples is necessary as it allows the findings to be generalized to the population (Sekaran and Bougie, 2016). In this study, the sampling frame consisted of a list of executives in all private commercial banks in Bangladesh. The availability of such a list would help the researcher to sample the population through the use of a table of random numbers that ensures random sampling and prevents biases. There are two major types of sampling design, namely probability sampling and non-probability sampling. However, the present study follows a cluster sampling method for selecting the respondents and the cluster sampling method is based on geographical location. Only the branches (800) of Dhaka city have been taken into consideration for the study.
3.3 Sample size and data collection
A total of 800 branches of the taken banks are available in Dhaka city. According to Sekaran and Bougie (2016), if the total population is 800, the minimum sample size should be 260. However, to avoid uncertainty and the risk of non-response cases, the present study distributed 400 sets of questionnaire papers among the respondents so that at least 260 sets could be returned in useable form. And 50% of each bank’s branches were covered to collect responses. All those 400 sets of questionnaires were physically distributed among the respondents, and 275 sets were collected back in useable form.
3.4 Research framework
It is important to understand that CSR can be used as an effective strategy for improving organizational performance. A substantial number of research studies have been conducted in the various dimensions of CSR in the literature (Bolton et al., 2011; Kim et al., 2010). Through an extensive literature review and the problem statement of this study, a research framework has been developed, which is shown in Figure 1 to achieve the objectives of this study. The framework includes CSR to employees (CSREM), CSR to customers (CSRCUS), CSR to environment (CSREN) and CSR to community (CSRCOM), which are used as the determinants of BP.
The framework diagram shows two vertical sections labeled “Independent Variables” on the left and “Dependent Variable” on the right. In the “Independent Variables” section, four text boxes are arranged in a vertical series and labeled from top to bottom as follows: “C S R to Employee”, “C S R to Customer”, “C S R to Environment”, and “C S R to Community”. Each text box has a right-pointing arrow leading to a single text box in the “Dependent Variable” section labeled “Banks’ performance”.Research framework. Source: Constructed by the authors based on literature review
The framework diagram shows two vertical sections labeled “Independent Variables” on the left and “Dependent Variable” on the right. In the “Independent Variables” section, four text boxes are arranged in a vertical series and labeled from top to bottom as follows: “C S R to Employee”, “C S R to Customer”, “C S R to Environment”, and “C S R to Community”. Each text box has a right-pointing arrow leading to a single text box in the “Dependent Variable” section labeled “Banks’ performance”.Research framework. Source: Constructed by the authors based on literature review
Therefore, there are four Independent Variables (IVs) (CSREM, CSRCUS, CSREM and CSRCOM) in this study and one Dependent Variable (DV) (BP).
3.5 Data analysis procedure
Generally, statistical methods are used to analyze collected data. In this study, data were analyzed using partial least squares structural equation modeling (PLS-SEM) technique using SmartPLS 3.0 software. Researchers are using the PLS-SEM method in order to measure the estimation that relates to the relationship in the field of path models involving latent constructs PLS-SEM analysis follow two steps, namely measurement model and structural model. In the measurement model, reliability and validity of the data are measured, and the structural model provides the findings for hypothesis testing.
4. Analysis and findings of data
4.1 Descriptive statistic
There is an absolute idea, which is explored from Table 1, that all factors serve maximum presence of 5 but the minimum response, CSR to customers has a minimum response of 3.67; CSRCOM, CSREM, CSREN and productivity of the bank have lower reply, which is only 1. Mean response has been drawn in the mean column, whereas the CSR moves on the customers outcome reflected 4.411, CSR towards community is 4.010, CSR towards employees is 4.15, CSR towards environment is 3.914 and BP is 4.025. Thus, all the means are above four except CSREN, and its mean value is very close to 4 (3.914); the outputs support the statements which are mentioned in the questionnaire. In the column of standard deviation (SD), the values represent the deviation of teach variables from its mean value. The values of SD also processed the mean and absolutely distributed of data or not. In the same way, there is a strong association between both the SD value and points. In Table 1, all factor values of SD are lower than 1. Depends upon the acquired outcome, almost all variables are comparatively lower that again mentioned a lot of connectivity remains in the both points of data (Aron et al., 2013; Chua, 2013).
4.2 PLS-SEM analysis results
The values of the test of reliability, validity and path coefficient accompanied by the determination of coefficient has been found from the PLS-SEM. The Figure 2 has been shown below which indicates the anticipated representation of picture constructed by Smart PLS software which version is 3.0.
The five latent variables are each represented by a circular node with the following labels: “E M P”, “C U S”, “C O M M U”, “E N V”, and “B P”. “E M P” is positioned at the upper left. From “E M P”, six individual leftward arrows connect to six rectangles positioned to the left side of “E M P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 0.588 points to “E M 1”. A second arrow with a path coefficient of 0.686 points to “E M 2”. A third arrow with a path coefficient of 0.591 points to “E M 3”. A fourth arrow with a path coefficient of 0.802 points to “E M 4”. A fifth arrow with a path coefficient of 0.738 points to “E M 5”. A sixth arrow with a path coefficient of 0.819 points to “E M 6”. To the right of “E M P”, “C U S” is positioned at the upper center. From “C U S”, three upward arrows connect to three rectangles positioned above it. The rectangles are arranged in a horizontal series and labeled from left to right as follows: A first arrow with a path coefficient of 0.725 points to “C 1”. A second arrow with a path coefficient of 0.682 points to “C 2”. A third arrow with a path coefficient of 0.718 points to “C 3”. A rightward diagonal arrow labeled 0.293 connects “C U S” to “B P”, which is positioned at the right center. Another rightward arrow labeled 0.429 also connects “E M P” to “B P”. “B P” is represented by a circular node with a value of 0.831 inside it. From “B P”, nine rightward arrows connect to nine rectangles positioned to the right side of “B P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 0.714 points to “B P 1”. A second arrow with a path coefficient of 0.663 points to “B P 2”. A third arrow with a path coefficient of 0.778 points to “B P 3”. A fourth arrow with a path coefficient of 0.868 points to “B P 4”. A fifth arrow with a path coefficient of 0.623 points to “B P 5”. A sixth arrow with a path coefficient of 0.588 points to “B P 6”. A seventh arrow with a path coefficient of 0.684 points to “B P 7”. An eighth arrow with a path coefficient of 0.666 points to “B P 8”. A ninth arrow with a path coefficient of 0.767 points to “B P 9”. Below “E M P”, “C O M M U” is positioned at the bottom left. From “C O M M U”, five individual leftward arrows connect to five rectangles positioned to the left side of “C O M M U”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 0.810 points to “C O M 1”. A second arrow with a path coefficient of 0.744 points to “C O M 2”. A third arrow with a path coefficient of 0.654 points to “C O M 3”. A fourth arrow with a path coefficient of 0.857 points to “C O M 4”. A fifth arrow with a path coefficient of 0.818 points to “C O M 5”. A rightward diagonal arrow labeled 0.404 connects “C O M M U” to “B P”. At the bottom center, “E N V” is positioned to the lower right of “C O M M U”. From “E N V”, six downward arrows connect to six rectangles positioned below it in a horizontal series labeled from left to right as follows: A first arrow with a path coefficient of 0.683 points to “E N V 1”. A second arrow with a path coefficient of 0.784 points to “E N V 2”. A third arrow with a path coefficient of 0.607 points to “E N V 3”. A fourth arrow with a path coefficient of 0.891 points to “E N V 4”. A fifth arrow with a path coefficient of 0.772 points to “E N V 5”. A sixth arrow with a path coefficient of 0.886 points to “E N V 6”. An upward arrow labeled 0.140 connects “E N V” to “B P”.PLS measurement model. Source: Output of PLS-SEM
The five latent variables are each represented by a circular node with the following labels: “E M P”, “C U S”, “C O M M U”, “E N V”, and “B P”. “E M P” is positioned at the upper left. From “E M P”, six individual leftward arrows connect to six rectangles positioned to the left side of “E M P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 0.588 points to “E M 1”. A second arrow with a path coefficient of 0.686 points to “E M 2”. A third arrow with a path coefficient of 0.591 points to “E M 3”. A fourth arrow with a path coefficient of 0.802 points to “E M 4”. A fifth arrow with a path coefficient of 0.738 points to “E M 5”. A sixth arrow with a path coefficient of 0.819 points to “E M 6”. To the right of “E M P”, “C U S” is positioned at the upper center. From “C U S”, three upward arrows connect to three rectangles positioned above it. The rectangles are arranged in a horizontal series and labeled from left to right as follows: A first arrow with a path coefficient of 0.725 points to “C 1”. A second arrow with a path coefficient of 0.682 points to “C 2”. A third arrow with a path coefficient of 0.718 points to “C 3”. A rightward diagonal arrow labeled 0.293 connects “C U S” to “B P”, which is positioned at the right center. Another rightward arrow labeled 0.429 also connects “E M P” to “B P”. “B P” is represented by a circular node with a value of 0.831 inside it. From “B P”, nine rightward arrows connect to nine rectangles positioned to the right side of “B P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 0.714 points to “B P 1”. A second arrow with a path coefficient of 0.663 points to “B P 2”. A third arrow with a path coefficient of 0.778 points to “B P 3”. A fourth arrow with a path coefficient of 0.868 points to “B P 4”. A fifth arrow with a path coefficient of 0.623 points to “B P 5”. A sixth arrow with a path coefficient of 0.588 points to “B P 6”. A seventh arrow with a path coefficient of 0.684 points to “B P 7”. An eighth arrow with a path coefficient of 0.666 points to “B P 8”. A ninth arrow with a path coefficient of 0.767 points to “B P 9”. Below “E M P”, “C O M M U” is positioned at the bottom left. From “C O M M U”, five individual leftward arrows connect to five rectangles positioned to the left side of “C O M M U”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 0.810 points to “C O M 1”. A second arrow with a path coefficient of 0.744 points to “C O M 2”. A third arrow with a path coefficient of 0.654 points to “C O M 3”. A fourth arrow with a path coefficient of 0.857 points to “C O M 4”. A fifth arrow with a path coefficient of 0.818 points to “C O M 5”. A rightward diagonal arrow labeled 0.404 connects “C O M M U” to “B P”. At the bottom center, “E N V” is positioned to the lower right of “C O M M U”. From “E N V”, six downward arrows connect to six rectangles positioned below it in a horizontal series labeled from left to right as follows: A first arrow with a path coefficient of 0.683 points to “E N V 1”. A second arrow with a path coefficient of 0.784 points to “E N V 2”. A third arrow with a path coefficient of 0.607 points to “E N V 3”. A fourth arrow with a path coefficient of 0.891 points to “E N V 4”. A fifth arrow with a path coefficient of 0.772 points to “E N V 5”. A sixth arrow with a path coefficient of 0.886 points to “E N V 6”. An upward arrow labeled 0.140 connects “E N V” to “B P”.PLS measurement model. Source: Output of PLS-SEM
4.3 Measurement model (outer model)
The Cronbach’s alpha is required to be greater than or similar to 0.70 for a satisfactory scale and 0.60 for a scale for empirical purposes for the reliability of data (Hair et al., 2012). The existing study’s all variables have Cronbach’s alpha values under the standard score such as CSR to customer (CUS) with (0.715), CSR to community (COM) (0.897), CSR to employee (EMP) (0.784), CSR to environment (ENV) (0.716) and finally banker’s financial performance (FP) (0.789). Cronbach’s alpha is constructed like a composite reliability for inside the reliability of consistency assessment. The acceptable value of composite reliability for PLS-SEM analysis is 0.7, and the contemporary study’s variables encounter that criterion. The existing study’s variables have a composite reliability value of more than 0.70; hence, the model postulated the internal consistency of lameness mentioned by Hair et al. (2012). As displayed in Table 2, the composite reliability values for all variables are higher than 0.7, where CUS (0.715), COM (0.897), EMP (0.784), ENV (0.716) and FP (0.898). According to Hair et al. (2012), the item loadings that are equal to or higher than 0.5 are considered satisfactory. Table 2 shows that all items’ loadings are higher than 0.5. The average variance extracted (AVE) reveals the discrepancy apprehended by the pointer’s comparison to dimension blunder. The examination research of the research, the AVE values fluctuated from 0.5 to 0.7, representing an upright step of construct validity of the dealings cast-off (Barclay and Smith, 1995). Thus, the item loadings 0.5 and AVE values endorse the convergent validity of the data.
Measurement model output
| Construct | Item | Loadings | AVE | Composite reliability | Cronbach’s alpha |
|---|---|---|---|---|---|
| CSR to customers | CUS1 | 0.627 | 0.601 | 0.788 | 0.715 |
| CUS2 | 0.615 | ||||
| CUS3 | 0.666 | ||||
| CSR to community | COM1 | 0.753 | 0.622 | 0.899 | 0.897 |
| COM2 | 0.764 | ||||
| COM3 | 0.740 | ||||
| COM4 | 0.762 | ||||
| COM5 | 0.674 | ||||
| CSR to employees | EMP1 | 0.630 | 0.544 | 0.821 | 0.784 |
| EMP2 | 0.668 | ||||
| EMP3 | 0.694 | ||||
| EMP4 | 0.775 | ||||
| EMP5 | 0.756 | ||||
| EMP6 | 0.650 | ||||
| CSR to environment | ENV1 | 0.757 | 0.587 | 0.778 | 0.716 |
| ENV2 | 0.780 | ||||
| ENV3 | 0.746 | ||||
| ENV4 | 0.676 | ||||
| ENV5 | 0.647 | ||||
| ENV6 | 0.694 | ||||
| Banks’ performance | FP1 | 0.691 | 0.598 | 0.898 | 0.789 |
| FP2 | 0.839 | ||||
| FP3 | 0.870 | ||||
| FP4 | 0.839 | ||||
| FP5 | 0.682 | ||||
| FP6 | 0.588 | ||||
| FP7 | 0.685 | ||||
| FP8 | 0.666 | ||||
| FP9 | 0.767 |
The sigma of the AVE of individual hypothesis was equated with the correlation between that construct and the other constructs for measuring discriminant validity. Table 3 demonstrates that the sigma of the AVE outdoes the significant familiarity in both stipulated and actual constructs (Fornell and Larcker, 1981), providing support for discriminant validity of the constructs in this study.
Correlations of constructs and discriminant validity assessment
| BP | COM | CUS | EMP | ENV | |
|---|---|---|---|---|---|
| BP | 0.773 | ||||
| COM | 0.588 | 0.788 | |||
| CUS | 0.488 | −0.297 | 0.775 | ||
| EMP | 0.541 | 0.411 | −0.409 | 0.737 | |
| ENV | 0.325 | −0.448 | 0.318 | −0.556 | 0.766 |
R2, which means determination of the coefficient value, indicates the amount of exogenous variable could be influences. The value indicates the extent to the exogenous variables can influence an endogenous variable. The current research has a BP, for which the R2 value is 0.831. The R2 value of 0.831 indicates that the performance of banks (dependent variable) is influenced by the independent variables by 83.10%. As an old rule of thumb, a R2 value of 0.75, 0.50, or 0.25 for endogenous latent variables can be respectively described as substantial, moderate or weak (Hair et al., 2011; Moosbrugger et al., 2009). The present study has a R2 value of 0.831, which indicates a high degree of effect. Thus, all (4) independent variables, namely CSR to customers, employees, environment and community, have substantial effects on the BP.
4.4 Structural model assessment for hypothesis testing
First step is the measurement model output, which sanctions the reliability and validity of data; the following step is to assess the output of the structural model. The hypothesis testing can be done in the structural model of the PLS analysis. Here, the path coefficients, t-statistics, p-values and errors are considered. A hypothesis is said to be accepted if it is significant at 5% level (t value > 1.96 or p < 0.05) (Henseler and Fassott, 2009). The output of the PLS structural model is shown in Table 4, and these findings are used for hypothesis testing. Figure 3 also shows the PLS structural model output.
Structural model path coefficients with p-values (path analysis)
| Hypothesized relationship | Path coefficients (β) | Standard errors | T-statistics | p-values | Result |
|---|---|---|---|---|---|
| H1: CSR to customers → Banks’ performance | 0.293 | 0.096 | 2.972 | 0.001 | Supported |
| H2: CSR to employees → Banks’ performance | 0.429 | 0.069 | 6.183 | 0.000 | Supported |
| H3: CSR to community → Banks’ performance | 0.404 | 0.065 | 6.195 | 0.000 | Supported |
| H4: CSR to environment → Banks’ performance | 0.140 | 0.123 | 1.142 | 0.254 | Not supported |
The five latent variables are each represented by a circular node with the following labels: “E M P”, “C U S”, “C O M M U”, “E N V”, and “B P”. “E M P” is positioned at the upper left. From “E M P”, six individual leftward arrows connect to six rectangles positioned to the left side of “E M P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 13.574 points to “E M 1”. A second arrow with a path coefficient of 15.867 points to “E M 2”. A third arrow with a path coefficient of 9.580 points to “E M 3”. A fourth arrow with a path coefficient of 21.998 points to “E M 4”. A fifth arrow with a path coefficient of 16.468 points to “E M 5”. A sixth arrow with a path coefficient of 11.082 points to “E M 6”. To the right of “E M P”, “C U S” is positioned at the upper center. From “C U S”, three upward arrows connect to three rectangles positioned above it. The rectangles are arranged in a horizontal series and labeled from left to right as follows: A first arrow with a path coefficient of 11.024 points to “C 1”. A second arrow with a path coefficient of 11.038 points to “C 2”. A third arrow with a path coefficient of 10.127 points to “C 3”. A rightward diagonal arrow labeled 2.972 connects “C U S” to “B P”, which is positioned at the right center. Another rightward arrow labeled 6.183 connects “E M P” to “B P”. “B P” is represented by a circular node. From “B P”, nine rightward arrows connect to nine rectangles positioned to the right side of “B P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 14.133 points to “B P 1”. A second arrow with a path coefficient of 3.419 points to “B P 2”. A third arrow with a path coefficient of 19.501 points to “B P 3”. A fourth arrow with a path coefficient of 43.108 points to “B P 4”. A fifth arrow with a path coefficient of 11.804 points to “B P 5”. A sixth arrow with a path coefficient of 11.391 points to “B P 6”. A seventh arrow with a path coefficient of 11.573 points to “B P 7”. An eighth arrow with a path coefficient of 14.079 points to “B P 8”. A ninth arrow with a path coefficient of 17.615 points to “B P 9”. Below “E M P”, “C O M M U” is positioned at the bottom left. From “C O M M U”, five individual leftward arrows connect to five rectangles positioned to the left side of “C O M M U”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 11.135 points to “C O M 1”. A second arrow with a path coefficient of 12.703 points to “C O M 2”. A third arrow with a path coefficient of 11.799 points to “C O M 3”. A fourth arrow with a path coefficient of 11.777 points to “C O M 4”. A fifth arrow with a path coefficient of 21.061 points to “C O M 5”. A rightward diagonal arrow labeled 6.195 connects “C O M M U” to “B P”. At the bottom center, “E N V” is positioned to the lower right of “C O M M U”. From “E N V”, six downward arrows connect to six rectangles positioned below it in a horizontal series labeled from left to right as follows: A first arrow with a path coefficient of 21.188 points to “E N V 1”. A second arrow with a path coefficient of 10.916 points to “E N V 2”. A third arrow with a path coefficient of 10.848 points to “E N V 3”. A fourth arrow with a path coefficient of 21.213 points to “E N V 4”. A fifth arrow with a path coefficient of 11.229 points to “E N V 5”. A sixth arrow with a path coefficient of 21.215 points to “E N V 6”. An upward arrow labeled 1.142 connects “E N V” to “B P”.The structural model. Source: Output of PLS-SEM
The five latent variables are each represented by a circular node with the following labels: “E M P”, “C U S”, “C O M M U”, “E N V”, and “B P”. “E M P” is positioned at the upper left. From “E M P”, six individual leftward arrows connect to six rectangles positioned to the left side of “E M P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 13.574 points to “E M 1”. A second arrow with a path coefficient of 15.867 points to “E M 2”. A third arrow with a path coefficient of 9.580 points to “E M 3”. A fourth arrow with a path coefficient of 21.998 points to “E M 4”. A fifth arrow with a path coefficient of 16.468 points to “E M 5”. A sixth arrow with a path coefficient of 11.082 points to “E M 6”. To the right of “E M P”, “C U S” is positioned at the upper center. From “C U S”, three upward arrows connect to three rectangles positioned above it. The rectangles are arranged in a horizontal series and labeled from left to right as follows: A first arrow with a path coefficient of 11.024 points to “C 1”. A second arrow with a path coefficient of 11.038 points to “C 2”. A third arrow with a path coefficient of 10.127 points to “C 3”. A rightward diagonal arrow labeled 2.972 connects “C U S” to “B P”, which is positioned at the right center. Another rightward arrow labeled 6.183 connects “E M P” to “B P”. “B P” is represented by a circular node. From “B P”, nine rightward arrows connect to nine rectangles positioned to the right side of “B P”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 14.133 points to “B P 1”. A second arrow with a path coefficient of 3.419 points to “B P 2”. A third arrow with a path coefficient of 19.501 points to “B P 3”. A fourth arrow with a path coefficient of 43.108 points to “B P 4”. A fifth arrow with a path coefficient of 11.804 points to “B P 5”. A sixth arrow with a path coefficient of 11.391 points to “B P 6”. A seventh arrow with a path coefficient of 11.573 points to “B P 7”. An eighth arrow with a path coefficient of 14.079 points to “B P 8”. A ninth arrow with a path coefficient of 17.615 points to “B P 9”. Below “E M P”, “C O M M U” is positioned at the bottom left. From “C O M M U”, five individual leftward arrows connect to five rectangles positioned to the left side of “C O M M U”. The rectangles are arranged in a vertical series and labeled from top to bottom as follows: A first arrow with a path coefficient of 11.135 points to “C O M 1”. A second arrow with a path coefficient of 12.703 points to “C O M 2”. A third arrow with a path coefficient of 11.799 points to “C O M 3”. A fourth arrow with a path coefficient of 11.777 points to “C O M 4”. A fifth arrow with a path coefficient of 21.061 points to “C O M 5”. A rightward diagonal arrow labeled 6.195 connects “C O M M U” to “B P”. At the bottom center, “E N V” is positioned to the lower right of “C O M M U”. From “E N V”, six downward arrows connect to six rectangles positioned below it in a horizontal series labeled from left to right as follows: A first arrow with a path coefficient of 21.188 points to “E N V 1”. A second arrow with a path coefficient of 10.916 points to “E N V 2”. A third arrow with a path coefficient of 10.848 points to “E N V 3”. A fourth arrow with a path coefficient of 21.213 points to “E N V 4”. A fifth arrow with a path coefficient of 11.229 points to “E N V 5”. A sixth arrow with a path coefficient of 21.215 points to “E N V 6”. An upward arrow labeled 1.142 connects “E N V” to “B P”.The structural model. Source: Output of PLS-SEM
Hypothesis 1: CSR practices towards customers are positively related to BP. This hypothesis is supported as Table 4 represents that the path coefficient value is 0.293 and the corresponding t-statistic is 2.972 (p < 0.01), which indicates 1% significance level. Thus, it is supported that CSR practices towards customers are positively related to the bank’s performance.
Hypothesis 2: CSR practices towards employees are positively related to the bank’s performance. The results of the existing research verify this hypothesis. The variable has a path coefficient of 0.429, which is significant at 1% (t, 6.183; p, <0.01) level. Thus, it is accepted that CSR practices towards employees are positively related to a bank’s financial performance.
Hypothesis 3: CSR practices towards the community are positively related to the bank’s performance. This hypothesis is supported as Table 4 describes that the path coefficient is 0.404 and the matching t-statistic is 6.195 (p < 0.01), which indicates 1% significance level. Thus, it is supported that CSR practices towards the community are positively related to BP.
Hypothesis 4: CSR practices towards the environment are positively related to the bank’s performance. From Table 4, the hypothesis examining values show that CSR practices towards the environment is positively correlated with BP, but it is not significant at the 5% level. The path coefficient for this variable is 0.140, and the t-statistics is 1.42 (p > 0.05), which is insignificant. Thus, the findings reveal that CSR practices towards the environment are not significantly related to the bank’s performance indicates that hypothesis 4 is not supported.
5. Discussion and conclusion
The study inspected the associations between different forms of CSR and BP. The statistical outcomes expose that all categories of CSR are positively associated with a bank’s financial performance, except CSR to the environment. CSR to employee and community out of the mentioned four CSR categories was highly appreciated, followed by CSR to customer in influencing bank’s financial performance, but CSR to environment did not satisfactorily influence bank’s financial performance. The first hypothesis test results reveal that CSR to customers is a substantial factor for any bank’s financial performance. This finding is relevant to that of Brammer et al. (2007), Omar et al. (2019), Fornell et al. (2006) and Prutina (2016), who found that there was a strong assertive and momentous association between CSR to customer and bank’s performance. Therefore, banks should concentrate on CSR activity for actual and potential consumers, which might be recognized means of promotional approach as well as crucial mover of sustainable profitability as well as wealth building. The second hypothesis test outcomes explore the CSR to employees is influential for the bank’s financial performance. This result is highly associated with that of Rego et al. (2011) and Turker (2009), who explored the affirmative links banks' financial performance and CSR to the human resources of the organization. Thus, banks should consider CSR to employee satisfaction for significant performance. The third hypothesis findings disclose that CSR to community plays a vital role in a bank’s performance. The outcome of the study is significantly relevant to Berman et al. (1999), and Mishra and Suar (2010) identified that there was inconclusive and significant correlation of CSR to the community with the bank’s performance. Hence, the banking industry should focus on appreciating the CSR to the community for long-term profitability and wealth building. The final hypothesis testing result is not supported with the literature in the context of Bangladesh. So, CSR to the environment is not significant for a bank’s performance in this study.
On the whole, the relationship between CSR and BP is openly recognized. Consequently, the attachment between BP and CSR, which were empirically scrutinized in several research studies and explored an abundance of textual mistakes as well as negotiations regarding affirmative association in both CSR and performance of the bank. In our study, all the variables have positive and significant links between CSR and bank’s performance except CSR to the environment. The findings of this study indicate that banks must take into account the CSR activities for increasing their performance. Because the hypotheses test findings reveal that CSR to customers, strategic partners and society contribute a lot in enhancing BP. Therefore, banks should do more CSR activities towards these groups for their sustainability. At the same time, the policy makers should also devise rules or policies so that banks and financial institutions focus more on CSR activities. This study is essential for the banking industry as it reveals insightful findings by exploring how CSR practices influence banks’ profitability. Therefore, through the findings of the research, which are now proven that, CSR activities do not incur costs; rather, they bring forth substantial benefits in the long run. The present study’s findings also enrich the pages of existing literature on CSR and organizational performance by providing some empirical findings from a developing economy like Bangladesh. The findings of this study might help future researchers advance this field of study to a greater extent. Though the present study generates some vital, insightful findings, the outcomes are tolerating some shortcomings. Research that was done recently, which took samples only from the executives of the banks working in the capital city. So, future studies may be conducted taking samples from all over the country. Future research may also be conducted considering all the financial institutions, both banks and non-banking financial institutions.
