Employees’ perception of CSR and organizational commitment, the role of culture and job position
Submitted by Melissa Idemudia, s2933314 Department of Economics and Business
Master’s Thesis for MSc. International Business and Management University of Groningen
January 20th, 2020 Supervisor: Dr. Shin Co-assessor: Dr. Kuiken
Abstract
The relationship between employees’ perception of CSR and their organizational
commitment is studied. Multiple linear regression analyses have been performed, using primary data obtained via an online survey. A strong positive relation was found between employees’ perception of overall CSR and their organizational commitment. However, a weak negative relation was found between their perception of CSR towards employees and their commitment. Besides, job position was found to positively moderate the relation between employees’ perception of CSR towards employees and their organizational
commitment. Moreover, the results did not find support for the moderating role of culture.
Keywords: Corporate social responsibility, employees, organizational commitment
3
Table of Content
1. Introduction ………..4
2. Literature review ………..7
2.1 Defining the concepts ………..7
2.2 Theory ………...8
2.3 Hypotheses development ………..9
2.3.1 Overall CSR and CSR to employees ………..9
2.3.2 The moderating role of culture ………...12
2.3.3 The moderating role of job position ………17
2.4 Conceptual model ………20
3. Methodology ………..21
3.1 Target sample ………21
3.2 Data collection ………...21
3.3 Variables description ………22
4. Data and analysis ………26
4.1 Sample characteristics ……….26
4.2 Validity and reliability of measurements ……….28
4.3 Correlation ………...30
4.4 Multiple regression analysis ……….33
4.5 Statistical model ………...35
5. Results ………37
6. Discussion and Conclusion ………..42
6.1 Discussion ………42
6.2 Managerial implications………...45
6.3 Theoretical contributions………...….46
6.4 Limitations………47
6.5 Conclusion……….………48
7. References ………...49
8. Appendices ……….55
4
1 Introduction
Corporate Social Responsibility (CSR) is becoming more and more popular among companies around the globe. Companies are increasingly pressured to pursue social responsible behaviour from a variety of stakeholder groups (Al-bdour, Nasruddin & Lin, 2010; Brammer, Millington, & Rayton, 2007; Rupp, Skarlicki, & Shao, 2013). Therefore, firms have to consciously think about their responsibility towards their stakeholders.
Since the 1960s and 1970s, companies’ social responsibility is a frequently discussed topic in the managerial and social scientific literature. The increasing interest in this subject has occurred with a tremendous increase of formal CSR programs and reporting (Mueller, Hattrup, Spiess & Lin-Hi, 2012). KPMG (2017) who studied CSR reporting, argue that nowadays corporate responsibility reporting is standard practice for large and mid-cap companies around the globe. They found that approximately three quarters of the studied companies in their survey publish these reports. Also, every sector has a reporting rate of 60% or more, for the first time in history. These findings signal that an incredible
percentage of firms engages in CSR.
According to Peterson (2004), early research focused on the relation between companies’
performance on social issues and their financial success. However, Wood and Jones (1995) argue that it might be more effective to study which particular stakeholders are affected by a companies social performance, rather than focusing on financial success. Accordingly, based on Freeman’s (1984) stakeholder theory, studies concerning CSR often focus on the relation between the company and a particular stakeholder group (Carroll, 1999).
Frequently studied stakeholders are investors, shareholders, managers, consumers and job candidates (Brammer et al., 2007). This study will focus on employees. A millennial
employee engagement study (Cone Communications, 2016) found that 83% of the
employees would be increasingly loyal to a firm that supports them in contributing to social and environmental issues.Furthermore, 63% would not take a job if their potential
employer does not have strong CSR practices. These findings indicate the importance of CSR to employees. Moreover, it is argued that employees are of primary importance and thus a critical internal stakeholder of the firm (Rodrigo & Arenas, 2008). Accordingly, it is valuable to study employees’ perception of CSR.
5 Several studies have shown support for the relation between employees’ perception of CSR and their work attitudes (i.e. Al-bdour et al., 2010; Farooq, Payaud, Merunka &
Valette-Florence, 2014; Mueller et al. 2012). However, research studying this topic in a cross-cultural setting is scarce. The minimal research that has been conducted, implies that the perception of CSR is contingent upon cultural context (Farooq, Rupp & Farooq, 2017;
Mueller et al., 2012; Rupp et al., 2013). Therefore, this study takes into account employees’
cultural values. It examines whether the different cultural dimensions of Hofstede moderate the relation between employees’ perception of CSR and their organizational commitment.
Moreover, as the majority of studies have focused on Western contexts (Farooq et al., 2014), there is a lack of understanding of whether the relation between employees’
perception of CSR and their work attitudes is equivalent for Western and non-Western contexts. Accordingly, this study attempts to take into account Western and non-Western countries. Hence, this study attempts to satisfy the need for studying this topic in other regions of the world. And might provide a better understanding of the influence of cultural values on the relation between CSR and employees’ work attitudes.
Besides, this study takes into account the job position of the employee, in terms of how closely they are connected to the customer. The relevance of this job characteristic will be discussed. Hence, this study will investigate the following main research questions:
“Is the relation between employees’ perception of CSR and their organizational commitment moderated by their cultural values?”
“Is the relation between employees’ perception of CSR and their organizational commitment moderated by their job position?”
The first question is of managerial interest because it can enhance organizations’
understanding about the moderating role of culture. It is stated that culture is one of the finest predictors of work place outcomes (Taras, Steel, & Kirkman, 2011). Accordingly it is important for managers to understand how CSR can improve the work attitudes of their firms’ employees in different cultural contexts.
Besides, information on the moderating role of job position can enhance managers’
knowledge on whether certain employees, experience a more positive influence of CSR on their work attitudes. Managers can use this knowledge to determine on which employees
6 they should focus particularly, when for example communicating their CSR practices.
Overall, this study attempts to provide managers with a better understanding of the relation between employees’ perception of CSR and their work attitudes. This information is valuable for managers because work attitudes are associated with several positive
outcomes such as reduced absenteeism, increased worker productivity and lower turnover rates (Meyer, Paunonen, Gellatly, Goffin, & Jackson, 1989).
7
2 Literature review
Since employees’ perception of CSR is a key part of this study’s research, findings from the literature that relate to how employees perceive CSR have been reviewed. More specifically, how employees’ perception of CSR influences their organizational commitment is the topic of interest.
Previous research studying the impact of CSR on employees can be categorized into two categories (Farooq et al., 2014; Turker, 2009). One category focuses on prospective employees, these studies argue that CSR enhances employers’ reputation and therefore enables them to become more attractive for employees. Whereas the other focuses on how CSR affects current employees. These studies indicate that CSR positively affects employees’
work attitudes (i.a. Brammer et al. 2007; Peterson, 2004; Turker, 2009). Due to the increasing importance of CSR for stakeholders and specifically employees, it can be
expected that CSR will become even more essential as an aspect contributing to employees’
loyalty, morale, productivity, and involvement in the organization (McWilliams & Siegel, 2001).
2.1 Defining the concepts
While CSR is frequently studied, there is still no well-developed consensus about how to define CSR (Farooq et al. 2014). Peterson (2004) showed that favourable perceptions of corporate citizenship were positively related to employees’ organizational commitment.
They used the Corporate Citizenship Scale as a measure to assess how employees perceive the social performance of their companies. This scale is based on Carroll’s (1979)
framework of CSR. Accordingly, they defined corporate citizenship as the extent to which companies meet the economical, legal, ethical and discretionary responsibilities demanded by their stakeholders. In contrast, Carroll (1999) distinguishes the economic component of CSR from the legal, ethical and discretionary components. They argue that the economic component encompasses what the company does for itself, while the other dimensions concern what the company does for others. Accordingly, Turker (2009) bases her definition upon this, and defines CSR as companies’ behaviour directed to positively influence
stakeholders, which goes beyond its economic interest. Additionally, Turker (2009) makes a distinction between various stakeholders, when studying the relation between CSR and
8 organizational commitment. She includes CSR to social and non-social stakeholders, CSR to employees, CSR to consumers and CSR to the government. Her results revealed that CSR to all of these stakeholders, except for CSR to the government, were significant predictors of organizational commitment. It is important to note that CSR to the government did not influence employees’ organizational commitment. This dimension of CSR concerns the legal requirements for companies. So her results indicate that employees might be sceptical about whether legal requirements are a dimension of CSR. Since, it is argued that
employees’ perception of CSR exclude economic interests and legal requirements, this study will adopt McWilliams and Siegel’s (2001) definition of CSR, in which CSR is defined as
‘‘actions that appear to further some social good, beyond the interests of the firm and that which is required by law’’ (p. 117).
As an indicator of employees’ work attitudes, organizational commitment is frequently used. Therefore, this research will also study this concept. Porter, Steers, Mowday & Boulian’s (1974) definition of organizational commitment will be adopted, in which organizational commitment is defined in terms of strength of an individual’s identification with and involvement in a particular organization. Besides, organizational commitment can be divided into three types, namely affective commitment, normative commitment and continuance commitment. Research studying the relation between CSR and organizational commitment frequently focused on affective commitment (Brammer et al., 2007; Farooq et al., 2014; Mueller et al. 2012; Turker, 2009). Earlier studies have
indicated that affective commitment has the strongest relation with various important work outcomes and behaviours (Allen & Meyer, 1990; Brammer et al., 2007). Therefore, this study will also use affective commitment as an indicator for organizational commitment.
Affective commitment can be defined as “the employee’s emotional attachment to, identification with, and involvement in the organization” (Meyer & Allen, 1991, p. 67).
2.2 Theory
Earlier research often used social identity theory (SIT) when studying the relation between CSR and employees’ work attitudes (Brammer et al. 2007; Mueller et al., 2012; Peterson, 2004; Turker, 2009). Therefore, this study will also be based on SIT. SIT suggests that a person’ sense of who they are is based on their membership in social categories (Tajfel,
9 Turner, Austin, & Worchel, 1979). How an individual perceives himself, also named self- concept, is affected by their membership in social organizations. Accordingly, the
organization at which the individual is employed affects the individual’s self-concept.
Individuals try to establish or improve their positive self-concept by comparing the
characteristics of themselves and the group to which they belong with other individuals and groups (Tajfel et al., 1979; Tajfel & turner, 1986). Comparisons in which the status of the group to which they belong is higher than the status of other groups, results in an improved self-concept. While comparisons in which the status of the group to which they belong is lower than the status of other groups, results in a reduced self-esteem of an individual.
Furthermore, it is suggested that employees’ perception of an organizations’s identity affects the strength of identification of the employee with the organization (Brammer et al., 2007). According to SIT it is hypothesized that when individuals associate themselves with positive reputable organizations this leads to an enhanced self-concept (Maignan & Ferrell, 2001; Tajfel & Turner, 1985). So, when employees’ perceive their organization as being socially responsible, the association with this organization may improve their self-concept (Brammer et al., 2007). Therefore, SIT proposes that when employees start to become proud of their membership in a socially responsible organization, their work attitudes can be influenced positively (Brammer et al., 2007; Maignan & Ferrel, 2001; Peterson, 2004). In short, it can be aruged that employees’ improved social identity, resulting from the
membership in an organization with a positive reputation, will influence the employees’
work attitudes.
2.3 Hypotheses development
2.3.1 Overall CSR and CSR to employees
Research has revealed that employees’ perception of CSR influences their work attitudes (i.a. Brammer et al., 2007; Peterson, 2004; Turker, 2009). This study will distinguish between two types of CSR, namely CSR to employees and overall CSR. CSR to employees concerns the CSR practices directly related to the employees. While overall CSR is related to CSR practices directed at all different stakeholder groups, so for instance directed to
consumers, environment, community etc. Moreover, earlier research often uses the terms internal and external CSR. Internal CSR concerns CSR practices directed specifically at
10 employees such as creating good working conditions, while external CSR encompasses CSR practices directed at customers, suppliers, the local community, the environment, etc. So directed at external stakeholders (Al-bdour et al., 2010). The distinction is made because it is expected that both affect employees organizational commitment, but through different ways. This is supported by Farooq et al. (2017), which argue that external and internal CSR affect organizational identification via differing paths. As such, it seems reasonable to make this distinction.
Firstly, the relation between employees’ perception of overall CSR and their organizational commitment will be discussed. Peterson (2004), uses SIT and shows that employees’
favourable perceptions of CSR are positively related to their organizational commitment.
Furthermore, Brammer et al. (2007) also use SIT. They distinguish between internal and external CSR. Furthermore, they considers an external aspect of CSR, namely employee perceptions of CSR in the community, and find that it is also positively related to employees’
organizational commitment. Turker (2009) bases her analysis on SIT as well. She makes a distinction between various stakeholders when studying the relation between CSR and organizational commitment. She includes CSR to social and non-social stakeholders, CSR to employees, CSR to consumers and CSR to the government. Her results revealed that CSR to all of these stakeholders, except for CSR to the government, were significant predictors of organizational commitment. Besides, Farooq et al. (2014) examine the indirect effect of employees’ perception of CSR on their organizational commitment and propose that organizational trust and organizational identification mediate the relationship. They use social exchange theory as well as SIT as a foundation for their predictions. Their findings support their anticipated mediation effects. Furthermore, CSR to the community and CSR to consumers were found that have a relation with employees’ organizational commitment.
While CSR to the environment had no effect.
Thus, it is often found that CSR to external stakeholders relate to organizational commitment. The explanation for these results is that overall CSR is expected to affect organizational commitment because it can positively affect companies’ reputation (Branco
& Rodrigues, 2006). As SIT argues, when employees associate themselves with positive reputable organizations, their self-concept will be enhanced. As such, an employee can feel
11 proud to be a member of a socially responsible organization (Dutton, Dukerich, & Harquail, 1994). Therefore, it is expected that employees’ perception of overall CSR positively
influences their organizational commitment.
H1: There exists a positive relationship between employees’ perception of overall CSR and their organizational commitment.
Secondly, the relation between employees’ perception of CSR to employees and their organizational commitment will be discussed. As mentioned before Brammer et al. (2007) distinguish between external and internal CSR. They consider two internal aspects, namely procedural justice in the organization and the provision of employee training. They find that these CSR aspects are positively related to employees’ organizational commitment.
Furthermore, Turker (2009) also examines the relation between employees’ perception of CSR to employees and their organizational commitment. From the different that they studied, CSR to employees was the most significant predictor of organizational
commitment. Besides, Al-bdour et al. (2010) focus on 5 aspects of internal CSR, human rights, health and safety, workplace diversity, training and education and work life balance.
They argue that each aspect is positively related to organizational commitment. Farooq et al. (2014) found that CSR to employees is the most important predictor of employees’
organizational commitment, which is in accordance with previous findings from Turker (2009).
So it is frequently found that employees perception of CSR to employees relate to their organizational commitment. Companies’ CSR practices towards employees are directly linked to the psychological and physical working climate of employees (Al-bdour et al., 2010). When a company engages in CSR towards its employees, it signals to care about a good working environment. Consequently, it can improve a company’s reputation as a good employer (Turker, 2009). Therefore, employees’ perception of their respect and status within the company might be strengthened (Farooq et al., 2014). Resulting in improved self-image of the employees. As a result, employees will generate greater emotional
attachment and feelings of obligation toward the company (Hofman & Newman, 2014). As
12 such, it is expected that CSR to employees is positively related to workers’ organizational commitment. Thus, the following hypotheses will be proposed:
H2: There exists a positive relationship between employees’ perception of CSR to employees and their organizational commitment.
2.3.2 The moderating role of culture
As Gjølberg (2009) argues, CSR is characterized by being a global concept. Nowadays, CSR is supported by key global institutions (i.a. the OECD, the World Bank, the UN). These global characteristics of CSR might result in the belief that national dynamics play a small role or even no role at all. However, it is argued that despite the global nature of CSR, it is practiced in different manners across various cultural contexts (Gjølberg, 2009). What might be usual and legal in one country can be considered illegal or unethical in other countries. For
instance, countries differ in their attitude towards issues regarding women’s role in
business, the environment, bribery, the importance of work relative to family. This signifies the impact culture has on the business environment (Blodget, Lu, Rose & Vitell, 2001).
Furthermore, several studies suggest that the perception of CSR is contingent upon cultural context (Farooq et al., 2017; Mueller et al., 2012; Rupp et al., 2013). Cultural values
influence employees’ organizational identification, which can be defined as the extent to which an individual feels a oneness with the organization (Ashforth, Harrison and Corley, 2008). So organizational identification depends on individuals’ own values and on how they perceive the practices of their company. This can be related to CSR, when an individual perceives the company’s CSR practices as favourable, it can positively affect the individual’s self-concept and because of that its organizational commitment (Peterson, 2004). National culture shapes individuals’ beliefs. Therefore culture which encompasses different
religions, traditions, heritages, and customs can cause distinct learning circumstances and histories. Consequently, these differing circumstances lead to diversity in beliefs and moral standards (Blodget et al., 2001). Accordingly, the CSR practices of an organization may be perceived desirable by one but may be perceived less so by someone else (Blodget et al., 2001). Because culture influences individuals beliefs and moral standards, it affects individuals’ perception of their organizations’ CSR practices. And in turn, the relationship between employees’ perception of CSR and their organizational commitment. This is
13 supported by Taras et al. (2011), they state that culture is one of the finest predictors of workplace outcomes. And argue that cultural values strongly influence ethics as well as organizational commitment. Therefore, it is important to take into account employees’
cultural values when studying the relation between employees’ perception of CSR and organizational commitment.
This study will adopt Hofstede’s (2011) definition of culture, in which culture is defined as “the collective programming of the mind that distinghuishes the members of one group or category of people from others”. Moreover, Hofstede’s initial four cultural
dimensions will be used, namely power distance, individualism versus collectivism, masculinity versus femininity and uncertainty avoidance.
Moreover, Mueller et al. (2012) studied the moderating effects of various cultural value dimensions on the relation between employees’ perception of CSR and their organizational commitment. They used the data from the Global Leadership and Organizational Behaviour Effectiveness (GLOBE) project. Their results display that several cultural value dimensions indeed moderated the relationship. Furthermore, Farooq et al. (2014) study the relation between CSR and employees’ organizational commitment through the mediation of organizational identification and trust. They conducted their study in Pakistan, which is characterized by a collectivistic culture. A remarkable finding is that environmental CSR seems insignificantly related to organizational commitment. Farooq et al. (2014) suggest that this could be due to cultural differences. Since Pakistan is characterized by scoring low on the individualism index of Hofstede. When comparing this to a study in the United States, which scores very high on the Hofstede’s individualism index, for example Stites and
Michael (2011) found that CSR concerning the environment is indeed positively related to organizational commitment. The difference could be because of the difference in cultural value between Pakistan and the United States. Consequently, Farooq et al. (2014) argue that the relation between CSR and employees is moderated by individual and national cultural values. Besides, they argue that future research should consider individual cultural values, to enable an understanding of boundary conditions.
14 Individualism vs. Collectivism
Mueller et al. (2012) adopt a definition of CSR which is more related to overall CSR. Their findings indicate that the relation between employees’ perception of CSR and their
organizational commitment is stronger in cultures higher on institutional collectivism and ingroup collectivism. Moreover, the results by Farooq et al. (2017) suggest that differences in cultural orientation, in terms of individualism vs. collectivism, moderate the relation between CSR an organizational identification. Specifically, their findings indicate that the impact of external aspects of CSR on organizational identification, through the mediation of prestige, was stronger for employees that scored high on collectivism.
In cultures characterized by a high level of individualism, connections between individuals are loose. Individuals are expected to take care of themselves and their direct family (Hofstede, 2011). While in cultures characterized by a high level of collectivism, individuals are integrated into groups. These groups not only include direct family but also extended family which take care of each other in exchange for questioning loyalty
(Hofstede, 2011). Blodget et al., (2001) study the effect of culture on sales agents’ ethical sensitivity towards various stakeholders (i.a. customers, competitors, and colleagues). They found that individualism negatively relates to sales agents’ ethical sensitivity towards stakeholders. They argue that individualistic employees are more likely to view themselves as the most important stakeholder. Whereas collectivistic employees are more likely to be concerned with the interests of other stakeholder groups. As overall CSR encompasses various stakeholders, it is proposed that collectivistic employees’ organizational
commitment will be more affected by the perception of overall CSR than individualistic employees.
H3a: The relation between employees’ perception of overall CSR and their organizational commitment is stronger for employees characterized by a high degree of collectivism.
Farooq et al. (2017) also study internal CSR. Their findings indicate that the relation between internal CSR and organizational identification is stronger for individualistic employees. Internal CSR is directly affecting employees. As individualistic employees are more likely to view themselves as the most important stakeholder (Blodget et al., 2001), it
15 can be hypothesized that the relation between their perception of CSR to employees and organizational commitment will be stronger than for their collectivistic counterparts.
H3b: The relation between employees’ perception of CSR to employees and their
organizational commitment is stronger for employees characterized by a high degree of individualism.
Masculinity vs. Femininity
These poles relate to the distribution of emotional roles between men and women. The masculine pole is characterized by assertive and competitive values. While the feminine pole is characterized by modest and caring values (Hofstede, 2011). Moreover, masculine individuals put more weight on money and material success than on immaterial wellbeing.
While feminine individuals are more focused on social relationships, and are more
concerned with immaterial wellbeing such as the quality of life (Hofstede, 1984). Since CSR concerns the social performance of a firm and not the financial performance, it can be argued that feminine individuals are more concerned about a firm’s CSR practices than masculine individuals. Accordingly, Blodget et al. (2001) propose that feminine individuals are more sensitive to the ethical interests of stakeholders and their results support this proposition. Thus it is anticipated that feminine employees’ organizational commitment will be stronger affected by overall CSR which encompasses different stakeholders.
H4a: The relation between employees’ perception of overall CSR and their organizational commitment is stronger for feminine employees than for masculine employees.
Moreover, CSR to employees can also include aspects related to workplace diversity, work- life balance and human rights (Al-bdour et al., 2010). As Hofstede (2011) argues feminine individuals tend to favour balance between family and work, while masculine individuals tend to perceive work over family. Also feminine individuals favour minimum social and emotional role differentiation between men and women. Therefore, it can be anticipated that feminine employees care more about CSR to employees than masculine employees.
H4b: The relation between employees’ perception of CSR to employees and their
organizational commitment is stronger for feminine employees than for masculine employees.
16 Power distance
Power distance concerns the degree to which the less powerful members of organizations and institutions anticipate and accept that power is divided unequally (Hofstede, 2011).
Furthermore, it is argued that an organization’s level of inequality is supported by the subordinates as much as by the superiors (Hofstede, 2011). The arguments regarding the moderation effect of power distance on the relation between CSR and organizational commitment, will not be discussed separately for overall CSR and CSR to employees. The rational for this is that the same argument can be applied to both types of CSR. Blodget et al.
(2001), hypothesize that power distance has a positive impact on employees’ ethical sensitivity. However, their results implied the opposite. They argue that higher power distance individuals may be guided by regulations and policies. And might perceive their behaviour as tolerable, considering there is no particular policy or rule indicating
differently. While low power distance individuals might put more weight on personal values. Therefore they are more likely to constrain their behaviour even in the absence of particular organizational rules regarding appropriate/inappropriate behaviour (Blodget et al. 2001). In agreement, Mueller et al.’s (2012) results reveal that the relation between the perception of CSR and organizational commitment was strong for employees with cultures lower in power distance. Davis (1960) states that there is a strong relationship between CSR and the interest in the use and the misuse of power. Mueller et al. (2012), argues that low power distance individuals tend to be suspicious of unequally distributed power.
Therefore, lower power distance individuals are expected to more sensitive towards CSR.
H5a: The relation between employees’ perception of overall CSR and their organizational commitment is stronger for low power distance employees than for high power distance employees.
H5b: The relation between employees’ perception of CSR to employees and their
organizational commitment is stronger for low power distance employees than for high power distance employees.
Uncertainty avoidance
Uncertainty avoidance concerns the level of uncertainty that is tolerated by society
(Hofstede, 1985). It displays to what degree members in a culture are programmed to feel
17 comfortable or uncomfortable in an unstructured situation, i.e. situations which are
unusual, new and unknown (Hofstede, 2011). It is found that high uncertainty avoidance individuals feel the need for written rules and regulations, are less willing to take risks, and do not tolerate deviant behaviour. Mueller et al.’s (2012) results indicate that uncertainty avoidance has no effect on the relation between employees’ perception of CSR and their organizational commitment. In contrast, Blodget et al. (2001) found uncertainty avoidance to be positively related to employees’ ethical sensitivity. Additionally, Kim and Kim (2010) found that uncertainty avoidance has a positive effect on public relation practitioners’
perception of CSR. Since low uncertainty individuals are characterized by being comfortable with uncertainty and tolerate various behaviours and beliefs, Yoo and Donthu (2002) argue that these individuals tend to care less about explicit and precise ethical regulations.
Consequently, it can be anticipated that low uncertainty avoidance individuals are less sensitive towards CSR. Furthermore, low uncertainty avoidance individuals have a greater willingness to take risks (Hofstede, 1983). Rallapalli, Vitell, Wiebe and Barnes (1994) argue that individuals who are more likely to take risks tend to violate rules to reach certain goals.
Accordingly, these individuals are more likely to engage in behaviour that is morally questionable. Therefore, it can be anticipated that the organizational commitment of high uncertainty avoidance employees will be more affected by CSR, compared to low
uncertainty avoidance employees.
H6a: The relation between employees’ perception of overall CSR and their organizational commitment is stronger for high uncertainty avoidance employees than for low uncertainty avoidance employees.
H6b: The relation between employees’ perception of CSR to employees and their
organizational commitment is stronger for high uncertainty avoidance employees than for low uncertainty avoidance employees.
2.3.3 The moderating role of job position
In this study, job position is described as the distance between the employee and the customers. This specific job characteristic is not frequently studied. However, several studies include various job characteristics when studying employee responses to CSR. For instance, Brammer et al. (2007), who examine the relation between employees’ perception
18 of CSR and their organizational commitment, include weekly working hours (part-time versus full-time). Moreover, earlier research has stated that non-managers are more likely than managers to react to- and evaluate a company’s CSR practices (Kim, Lee, Lee, & Kim, 2010; Rupp, Ganapathi, Aguilera & Williams, 2006). Besides, Turker (2009) includes the size and sector of employees’ companies. Since these earlier studies have incorporated various job characteristics when studying the relationship between employee perception of- and their responses to CSR, it is expected that job position might also influence the relation between employees’ perception of CSR and their organizational commitment.
Korschun, Bhattacharya and Swain (2014) studied the relation between CSR, customer orientation and job performance of frontline employees (i.e. customer service representatives, cashiers, waitstaff). Their results reveal differences in how CSR can affect job performance among frontline employees. As social identity theory proposes, individuals can be strongly affected by their membership in organizations (Tajfel et al., 1979). For frontline employees, who represent the company when dealing with customers, the identification with the company is particularly important (Korschun et al. 2014).
Organizational identification can be defined as the extent to which an individual feels a oneness with the organization (Ashforth et al., 2008). Besides, it is suggested that CSR influences companies’ reputation. If an employee considers CSR to be important for its self- view, this leads to identifying more with the company. As such, his work attitudes might be influenced. It is hypothesized that the relation between CSR and employees’ work attitudes might be stronger for employees which are closely connected to the customers, such as customer service representatives. Because they have to represent the organization, so for them it might be of particular relevance that they belief that their organization is socially responsible and that they feel happy to be a member of the organization (Dutton et al., 1994; Korschun et al., 2014). Therefore, the following hypotheses are proposed:
H7a: Job position moderates the relation between employees’ perception of overall CSR and their organizational commitment. More specifically, the more the employee has to deal with the customer the stronger will be the relation between employees’ perception of overall CSR and their organizational commitment.
19 H7b: Job position moderates the relation between employees’ perception of CSR to employees and their organizational commitment. More specifically, the more the employee has to deal with the customer the stronger will be the relation between employees’ perception of CSR to employees and their organizational commitment.
20
2.4 Conceptual model
The hypotheses result in the following conceptual model. As discussed above, employees’
perception of both types of CSR will positively relate to their organizational commitment (hypotheses 1 and 2). Furthermore, it is anticipated that employees’ individual cultural values moderate this relation between employees’ perception of CSR and their
organizational commitment (hypotheses 3 to 6). Finally, it is hypothesized that job position will positively affect the relations between employees’ perception of CSR and their
organizational commitment (hypothesis 7).
21
3 Methodology
This study has adopted a quantitative research method as this is most appropriate for answering the research questions. Since this research is interested in explaining
relationships between different phenomena, quantitative research has been conducted (Yilmaz, 2013). To avoid bias and other noise the researcher was disconnected from the data collection process which ensured that the researcher could not influence respondents (Firestone, 1987). Therefore, an anonymous online survey was conducted. The purpose of this study was to produce findings which are generalizable to some extent. Therefore, a quantitative method that aimed to reach a diverse target audience was utilized.
Additionally, formal and structured instruments from the literature were employed to gather information about respondents.
3.1 Target sample
The participants in this study were employees from diverse organizations. Employees who work part time or full time, at least 12 hours a week were included. Also, each employee had to be spending more time earning an income compared to studying. These
requirements were set to assure that respondents were serious about their job and work for a considerable time per week, increasing the likelihood of being able to answer all of the questions that were asked in the questionnaire.
3.2 Data collection
To answer the research question, this study has collected primary data and performed a quantitative data analysis. A questionnaire was distributed to respondents via social media and through direct contact. Qualtrics was used to create the questionnaire. As such, a link of the questionnaire could be easily distributed and filled out by the respondents on their phone or computer. The questionnaire mostly contained Likert-type scale items, which ranged from 1 (strongly disagree) to 5 (strongly agree). The variables that are represented in the survey are discussed in section 3.3.
The questionnaire was written in English. It was assumed that all the respondents that participated in the study sufficiently mastered the English language, so that they could
22 understand the questions that were asked from them.
The questionnaire started with a description of the requirements to be met by respondents, an introduction explaining the purpose of the study and rules about confidentiality. In this way, respondents knew what the survey was about and what their participation entailed, before answering the substantive questions. The data were gathered during a period of 2,5 weeks, from the 28th of November until the 15th of December 2019.
3.3 Variable description
Dependent variable
To measure organizational commitment, the widely used instrument from Allen and Meyer (1990)was implemented, measuring three types of organizational commitment, namely affective commitment, continuance commitment and normative commitment. The three different types of commitment each contain 8 items.
For affective commitment, respondents were asked to indicate to what extent they agreed with statements such as ‘I would be very happy to spend the rest of my career with this organization’ and ‘I feel ‘emotionally attached’ to this organization’. Regarding continuance commitment, respondents were asked to indicate to what extent they agreed with
statements such as ‘It would be very hard for me to leave my organization right now, even if I wanted to’ and ‘I feel that I have too few options to consider leaving this organization’.
Regarding normative commitment, respondents were asked to indicate to what extent they agreed with statements such as ‘I believe that a person must always be loyal to his or her organization’ and ‘If I got another offer for a better job elsewhere I would not feel it was right to leave my organization’.
Moreover, Allen and Meyer’s (1990) commitment items included reversed items (e.g. ‘I do not believe that a person must always be loyal to his or her organization’ and ‘I do not feel a strong sense of belonging to my organization’). These reversed items have been adjusted (‘unreversed’) to make it easier for the respondents, as it is argued that
respondents can get confused through reverse wording can cause data to become contaminated. Moreover, respondents inattention of reverse wording could also lead to contaminated data (van Sonderen, Sanderman, & Coyne, 2013).
23 Independent variables
The perception of overall CSR and the perception of CSR to employees are the main
independent variables. As Peterson (2004) argues, employees’ perceptions will determine their work attitudes, regardless of whether these perceptions are accurate or not. Therefore the items concerning CSR are based on employees’ perception of their company’s social performance.
To measure CSR to employees, a perceptual measure developed by Turker (2009) was used.
This scale consists of six items which include statements like ‘Our company encourages its employees to participate to the voluntarily activities’ and ‘Our company policies encourage the employees to develop their skills and careers’.
Furthermore, to measure employees’ perception of a firm’s overall CSR performance, Maignan and Ferrell’s (2000) instrument for overall responsible behaviours was used. The overall responsible behaviours of an organization can be divided into four dimensions, namely economic citizenship, legal citizenship, ethical citizenship and discretionary citizenship. As discussed above, earlier research indicated that employees are sceptical about the economic dimension as well as the legal dimension being part of CSR. Therefore this study only uses the ethical and discretionary dimension for overall CSR. The ethical citizenship scale consists of five items such as ‘A confidential procedure is in place for employees to report any misconduct at work’. And, the discretionary citizenship scale includes five items such as ‘Our business supports employees who acquire additional education’.
Moderator variables Job position
The distance between the employee and the customer is in this study depicted by the term
‘job position’, two items were used to examine this concept. Namely, ‘Do you work in a sales department or department with direct relation to customers (including call center)?’ and
‘Does your job entail direct contact with customers on a daily basis?’. These items
distinguish employees who often have contact with customers from employees who do not
24 Individual cultural values
Based on Hofstede’s cultural dimensions, the power distance orientation, uncertainty avoidance orientation, collectivism orientation and masculinity orientation were measured using 20 items from Donthu and Yoo’s (1998) cultural values scale. This scale is widely used in the marketing and international business literature (Sheikh, Newman & Al Azzeh, 2013).
Furthermore, it has been proven to successfully capture Hofstede’s (2001) cultural value dimensions at the individual level (Yoo & Donthu, 2002; Yoo et al., 2011; Sheikh, Newman and Al Azzeh 2013). The scale shows adequate reliability, validity, and across-sample and across-national generalizability (Yoo et al., 2011). Items used included ‘People in higher positions should make most decisions without consulting people in lower positions.’, ‘It is important to have instructions spelled out in detail so that I always know what I’m expected to do.’, ‘Individuals should sacrifice self-interest for the group.’ and ‘It is more important for men to have a professional career than it is for women.’.
Control variables
A number of demographic variables, such as gender, age and education, are controlled for as earlier research has indicated that they may influence employee commitment (Brammer, 2007; Iqbal, 2010; Farooq et al., 2014; Peterson, 2004). Furthermore, leadership is found to influence affective commitment (Bono & Judge, 2003; Brammer et al., 2007). Also, the results of Igbal (2010) indicate that supervisors and managers are more committed to their organizations than the lower level employees. Dick (2011) shows that rank is positively related with organizational commitment. Therefore a variable named rank, in which the respondent has to answer in what rank he is currently in, is included.
Besides, earlier work argues that job satisfaction positively relates to organizational commitment (Brammer et al. 2007; Meyer et al., 2002; Mueller et al. 2012). As Meyer (2002) describes, job satisfaction has an affective aspect and therefore is likely to be correlated to affective commitment. To measure job satisfaction the instrument by Susskind, Borchgrevink, Kacmar and Brymer (2000) was used, as these questions are framed in a concise and clear manner. It consists of three items such as ‘Generally speaking, I am very satisfied with the job’ and ‘I am generally satisfied with the kind of work that I do in this job’.
25 Moreover, it is argued that there exists a link between the different types of commitment and turnover intention (Allen & Meyer, 1991). Since employees who are strongly
committed to their organization are those who are least probable to leave their
organization. Therefore, turnover intention is also included in the analysis. To measure turnover intention Arkoubi, Bishop and Scott’s (2011) instrument was used. This is based on the turnover intention instrument of Bluedorn (1982). It consists of four items such as ‘I am going to quit working here real soon’ and ‘I have started to consider my job options outside of the company for which I work now’.
26
4 Data and analysis
4.1 Sample characteristics (n=102)
Characteristic Frequency Percentage
Gender Male 47 46.1%
Female 54 52.9%
Unknown 1 01.0%
Age 18 - 24 22 21.6%
25 - 34 31 30.4%
35 - 44 13 12.7%
45 - 54 20 19.6%
Above 54 16 15.7%
Nationality Dutch 79 77.5%
European
(non-Dutch) 7 6.9%
Asian 9 8.8%
African 4 3.9%
American 2 2.0%
Unknown 1 1.0%
Education Lower than bachelor’s degree 16 15.7%
Bachelor’s degree 44 43.1%
Master’s degree or higher 42 41.2%
Direct contact No 58 56.9%
with customers Yes 44 43.1%
The sample consists of 102 respondents. The response rate is equal to 86%. All these respondents are employees. There is an almost even distribution between respondents who identified as male compared with respondents who identified as female (52.9%). One respondent preferred not to answer the question with regard to gender as well as with regard to nationality. However this respondent did fill in all of the remaining questions, therefore this response was kept.
The majority of the sample is aged between 25-34 (30.4%). Followed by respondents aged between 18 - 24 (21.6%). An almost similar amount of people age between 45 - 54 (19.6%). 16 respondents age above 54 (15.7%). Whereas the smallest group ages between 35 - 44 (12.7%). Overall it can be said that the sample is representative for the several age groups.
27 In addition, it can be observed that the majority of the sample (84.3%), has a
Bachelor’s degree or higher. This indicates that average educational level is remarkably high.
Most of the respondents are Dutch (77.5%). Then there are 7 respondents from Europe who are not Dutch (6.9%), such as German, Spanish, Bulgarian etc. Moreover the sample includes 9 Asian respondents (8.8%), 8 Chinese respondents and 1 respondent is Vietnamese. Then there are 4 African respondents (3.9%), 3 South African respondents and 1 Cameroon. Additionally, the sample includes 2 American respondents both are Canadian (2.0%). Since the sample mostly consists of Dutch respondents it might be biased.
Also, the share of respondents which are in direct contact with customers on a daily basis is fairly high (43.1%) compared to the general population.
Moreover, the respondents are employed in diverse industries (i.a. Trade and services, information technology and justice, security and public administration.). Furthermore, there is also quit some variety in company size.
The characteristics inform how the sample is composed. This composition is of interest for the generalizability of the results. As the sample includes almost an equal amount of males and females, this seems to be representative for the general population. Also all age groups are represented with not very diverse amounts, therefore this is also representative of the general population. However, there is an overrepresentation of Dutch respondents.
Additionally, the sample mainly consists of highly educated employees. As such the sample is biased towards Dutch and highly educated employees.
28
4.2 Validity and reliability of measurements
To determine the validity of the instruments used in this study, a confirmatory factor analysis (CFA) is performed. This is a statistical technique which is used to verify the factor structure of a combination of observed variables (Suhr, 2006). And statistically tests the hypothesis that there exists a relation between the observed variables and their underlying latent constructs. This hypothesis is based on theory and/or empirical research. The aim of performing a CFA is to check whether the scale is valid, meaning that the items measure the intended variable.
Moreover, to ensure the reliability of the variables a reliability analysis must be conducted.
Cronbach’s alpha is a method to check the degree to which the different underlying items are well enough related to be combined into a single variable (Field, 2009). It measures the internal consistency between the items, where a higher value means a better internal consistency. Before conducting a reliability analysis, all items should be coded in the same direction. Therefore one item of turnover intention was recoded.
Affective commitment
First, the process of performing the CFA for affective commitment is elucidated here, to demonstrate how it was executed. Since this process is similar for the subsequent variables, the results of the successive CFA’s will be discussed concisely.
In order to perform a factor analysis, two conditions have to be met. The data should be continuous and the sample size should be sufficiently large (Field, 2009). Likert scaling presumes the existence of a latent continuous variable, whose value characterizes the opinions of respondents (Clason & Dormody, 1994). Each Likert-type discretely
approximates the continuous underlying variable. Therefore it is assumed that Likert-type data is approximately continuous of nature, satisfying the first condition. Furthermore, by computing the Kaiser-Meyer-Olkin (KMO) coefficient, the adequacy of the sample size is observed (Kaiser, 1970). Any value above 0.6 was regarded as acceptable for a factor analysis (Hinton, 2014). For the 8 items related to affective commitment the sample adequacy is meritorious (KMO = 0.885).
It was also observed that there was no issue of multicollinearity, as the correlation table
29 indicates that no coefficient is higher than 0.9 (Tabachnick, Fidell, & Ullman, 2007).
Furthermore, none of the items display correlation coefficients consistently below 0.3, indicating no need to delete any items. The factor analysis showed that the first factor explains 57.60% of the total variance of affective commitment. The eigenvalue was well above 1, satisfying the Kaiser criterion (Kaiser, 1960).
The component matrix shows that all items have a factor loading of at least 0.4 (see appendix A, table 1), which is considered to be sufficient (Field, 2009). Furthermore the reliability analysis showed that there exists internal consistency between the 8 items (α = 0.886).
Overall CSR
Again, the sample size is sufficiently large (KMO = 0.804). The factor analysis for the items of the ethical dimension and the discretionary dimension showed that the total variance explained by one factor is 41.08%, which is relatively low. Moreover, the component matrix shows that each item has a sufficient factor loading of at least 0.4 (see appendix A, table 2).
The reliability analysis showed that the internal consistency among the items is sufficient (α
= 0.837).
CSR to employees
To assure whether all the items belonging to CSR to employees load well on one factor, a principal component analysis was conducted. The condition for sample size was met (KMO
= 0.839). The results showed that the first factor explains 51.4% of the total variance.
Therefore, measuring CSR to employees with these 6 items is valid. Additionally, The eigenvalue of the first factor equalled 3.1. Besides, the component matrix shows that each item sufficiently loads on the first factor (see appendix A, table 3). Furthermore, the reliability analysis shows sufficient internal consistency between the six items of CSR to employees (α = 0.806).
Individual cultural values
When conducting the factor analysis it was observed that the sample size was middling (KMO = 0.713), which justified conducting a factor analysis. Furthermore, the factor analysis was conducted using varimax rotation since the factors were expected to be
30 independent of each other (Field, 2009). The number of factors is fixed at four, as the items are intended to be measuring 4 dimensions of individual cultural values (Yoo et al., 2011).
Furthermore, if an item shows cross loading it could indicate that it measures two different constructs and therefore should be deleted. The rotated component matrix (see appendix A, table 4) displays that for the second item of the collectivism dimensions (‘col2’) a cross loading exists. The difference between the two loadings is smaller than 0.2, accordingly the researcher decided to delete this item. The other items showing cross loadings have a higher loading on the appropriate factor with at least a difference of 0.2 between the cross loadings, as such there is no reason to delete any other items.
Furthermore, a reliability analysis was done for the four variables separately. The internal consistency between the items was sufficient for collectivism, power distance and
uncertainty (α = 0.823, α = 0.699 and α = 0.741, respectively). However, the Cronbach’s alpha for masculinity was equal to 0.655. It was observed that deletion of item four, resulted in a higher alpha (α = 0.670). Therefore, it was decided to exclude item four.
Deleting another item would decrease the value of alpha. Since the coefficient lies between 0.6 and 0.7, the variable is still moderately reliable.
So with regard to cultural values, two items were deleted. When reconducting the analysis without these items, it was observed that four factors explain 56.05% of the variation in individual cultural values.
Job satisfaction and turnover intention
For the items of job satisfaction, the condition for sample size was miserable (KMO = 0.585).
The results displayed that the first factor explained 69.9% of the total variance.
Accordingly, measuring job satisfaction with 3 items is valid. Furthermore, the eigenvalue of the first factor equalled 2.1 and the eigenvalue of the second factor corresponds to 0.7. As such there was no evidence to suspect that the items load on multiple factors. Besides, the component matrix indicates that all items load sufficiently on the factor (see appendix A, table 5). As the sample size coefficient is miserable, interpreting the results of the factor analysis should be done carefully. In addition, The reliability analysis showed sufficient internal consistency between the items (α = 0.783).
