Master Thesis
The effect of Corporate Governance on the
CSR strategy of multinational enterprises:
The moderating role of the strength of CSR
requirements of the home country
MSc Business Administration | International Management Student Marieke de Boer
Student number 11136227 Supervisor Dr. L. DiVito
Second reader Dr. M. Westermann-Behaylo
Statement of originality
This thesis is written by student Marieke de Boer who declares to take full responsibility for the contents of this document.
I declare that the information presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Business and Economics of the University of Amsterdam is solely responsible for the supervision of completion of the work, but not for the contents.
Table of content
Abstract 4
Introduction 5
Literature review 8 Corporate social responsibility 8
The concept of corporate social responsibility 8
Integration of CSR into international business 10
CSR strategy 11
Corporate governance 13
The concept of corporate governance 13
Corporate governance models 14
Institutional environment 17
Institutional theory 17
Institutional environment shaping CSR requirements 18
Existing research gap 20
Conceptual model 21
Methodology 26
Research strategy 26
Data collection procedure 26
Description of the sample 27
Measures 28
Data analysis 33
Results 35
Descriptive statistics 35
Binary logistic regression 36
Moderation 38
Discussion 41
Findings 41
Contributions to the literature 43
Managerial implications 44
Limitations and future research 45
Conclusion 47
Abstract
The goal of this study is to test the effect of corporate governance on the corporate social responsibility (CSR) strategy of multinational enterprises (MNEs). A quantitative study is conducted on the world’s top non-financial MNEs from developed and developing countries. The findings suggest that MNEs with a shareholder orientation are more likely implement a centrally coordinated global CSR strategy, while MNEs with a stakeholder adopt a decentralized local CSR strategy. However, the probability that a firm in a stakeholder-country adopts a local CSR strategy is smaller than the likelihood that a shareholder-oriented MNE adopts a global CSR strategy. The results highlight that the strength of the CSR requirements of the home country where the MNE is headquartered moderates the relationship between corporate governance and CSR strategy. This signifies that there are differences in the effect of corporate governance on CSR strategy between MNEs from developed and developing countries. This research has both academic and practical implications. An important gap within academic business literature concerning CSR strategy and corporate governance is filled. The understanding of the relationship between corporate governance and CSR strategy has been enriched by examining the moderating role of the strength of CSR requirements of the home country in this relationship. Besides this, the paper makes managers aware of the importance of CSR strategy and they get a better understanding of how they can manage CSR. Eventually, this will help to improve business performance.
Key words: Corporate social responsibility strategy, corporate governance, stakeholder orientation, shareholder orientation, corporate social responsibility requirement
Introduction
Increasing global problems such as human rights violations, poverty, inequality and climate change have attracted widespread attention in the last decade in international business (Kolk and Tulder, 2010). Multinational enterprises (MNEs) are encouraged by societies to provide a high level of corporate social responsibility to address such problems (Kolk and Tulder, 2010; Marano and Kostova, 2016). Corporate social responsibility (CSR) can be defined as ’the social responsibility of business encompasses the economic, legal, ethical and discretionary expectations that society has of organizations at a given point in time’’ (Carroll, 1979, p.500). Because MNEs are embedded in multiple institutional environments and institutions are mostly country specific, organizational practices such as CSR vary from country to country (Williams and Aguilera, 2008). As a result, MNEs are confronted with conflicting pressures from numerous stakeholders such as governments, non-governmental organizations, employees, customers, and suppliers to adopt a policy of social responsibility (Marano and Kostova, 2016). Managing CSR strategically has become crucial because CSR strategy helps MNEs to address the expectations of stakeholders who are embedded in different institutional environments (Husted and Allen, 2006; Muller, 2006; Porter and Kramer, 2006). The concept CSR strategy is, however, underdeveloped. Thus, it is essential to explore this concept further. Firms can develop centrally coordinated global CSR strategies or decentralized local CSR strategies (Muller, 2006). The corporate governance model that is dominant in the home country is expected to have a role in such a decision-making process since corporate governance can be defined as ‘’the study of power and influence over decision making within the company’’ (Aguilera and Jackson, 2010, p. 487). Firms can either adopt a stakeholder-oriented model whereby firms try to balance the interests of different stakeholders or a shareholder-oriented model whereby the primary goal of a firm is maximizing shareholder value (Aguilera and Cuervo-Cazurra, 2004).
Since these models seem to contradict each other, it is expected that this will lead to the development of different CSR strategies. In addition, the strength of CSR requirements of the home country where the MNE is headquartered and the degree of internationalization are expected to moderate the relationship between CSR strategy and corporate governance. Therefore, this research addresses an important gap by proposing the following research question: What is the effect of corporate governance on the CSR strategy of multinational
enterprises and what are the roles of the degree of internationalization and the strength of CSR requirements of the home country herein?
A quantitative research methodology is chosen to establish relationships between the variables. As its theoretical basis, this paper relies upon CSR theory (Carroll, 1979; McWilliams and Siegel, 2001; Baumann-Pauly et al., 2013), and in particular CSR strategy (Muller, 2006). This research uses the theoretical contributions of Rossouw (2008) to consider the impact of corporate governance on CSR strategy. Institutional theory provides a theoretical umbrella under which CSR strategies and the predictive capacity of corporate governance can be understood.
The conclusion of this study demonstrates that MNEs with a shareholder orientation are more likely to adopt a global CSR strategy, while MNEs with a stakeholder orientation adopt a local CSR strategy. The likelihood that a stakeholder-oriented MNE adopts a local CSR strategy is, however, smaller than the probability that an MNE in a shareholder-oriented country adopts a global CSR strategy. The strength of the CSR requirements of the home country where the MNE is headquartered moderates the relationship between corporate governance and CSR strategy. This signifies that the effect of corporate governance on CSR strategy differs between MNEs from developed and developing countries.
This research contributes to the literature and management practices in several ways. First of all, this research builds knowledge in the literature of CSR strategy and corporate
governance. This paper seeks to add a new dimension to this debate by examining the moderating role of the strength of CSR requirements of the home country. Besides this, examining how MNEs with contrasting corporate governance models respond to stakeholder pressures will help managers to develop and implement appropriate CSR strategies.
The paper is structured as followed. First, the literature review provides an overview of the relevant existing knowledge on this subject. This section ends with the developed hypotheses, which are formed based on the literature review. Second, the appropriate research design is discussed. This is further followed by a description of the process of data analysis. Third, the findings of the research are presented. The fourth section discusses these findings, the contributions to the literature, and managerial practices. The limitations and suggestions for future research are also discussed in this section. The study ends with a conclusion.
Literature review
This chapter provides the theoretical background of this study. The first section explains the concept corporate social responsibility and explores the concept in the context of international business. In addition, the strategic use and types of CSR strategies are discussed. The second section evaluates the concept corporate governance and discusses the different corporate governance models. Lastly, the institutional environments in which multinational enterprises (MNEs) operate and their relation to CSR and corporate governance are described. This chapter ends with the formulation of the research question and hypotheses.
Corporate social responsibility
The concept of corporate social responsibility
In the past decade, corporate social responsibility has drawn increased attention (Kolk and Tulder, 2010). A lot of research has been conducted on this topic. However, numerous definitions have been proposed and no consistent definition is given (Carroll, 1979; Williams and Aguilera, 2008). For example, some researchers focus on the social issues such as product safety for which a company has a responsibility, while other scholars suggest that social responsiveness is more concerned with the philosophy of response than with issues that should be addressed (Caroll, 1979). Since researchers refer to different dimensions of CSR, Carroll’s comprehensive definition of CSR is used to explain this concept.
Carroll (1979) addresses the entire range of obligations firms have to society. The author categorizes social responsibilities into four groups. The first component of CSR,
economic responsibilities, refers to the actions that serve a firm’s ultimate goal of maximizing
profits. Carroll (1979) emphasizes that high profits contribute to society because it facilitates growth. This component is the most important element upon which the three other elements rest. The second element, legal responsibilities, is formed out of legal laws a firm has to apply
to. Third, ethical responsibilities are additional behaviors and activities that are not necessarily and explicitly included in legal restrictions. The fourth and last category is constituted by
discretionary responsibilities. These responsibilities include voluntary philanthropic activities.
If a firm does not engage in these activities, it does not mean that it is unethical per se. This category is most difficult to define because of its subjective character. The four categories are related to some of the proposed definitions of CSR. However, Carroll’s pyramid (1979) reflects CSR in a more exhaustive manner, which makes this concept an important contribution to the CSR literature. Eventually, the following definition of CSR is created: ‘’the social responsibility of business encompasses the economic, legal, ethical and discretionary expectations that society has of organizations at a given point in time’’ (Carroll, 1979, p.500).
After the development of this definition, the concept of corporate social responsibility has evolved. According to McWilliams and Siegel (2001, p.117), CSR can be defined as ‘’situations where a company goes beyond compliance and engages in actions that appear to further some social good, beyond the interests of the firm and that which is required by law’’. This definition mainly reflects the ethical and discretionary responsibilities of Carroll’s pyramid (1979) and is often used by scholars to explain the concept of CSR. An even more recent study lays emphasis on the multi-faceted and complex relationships between business and society (Baumann-Pauly et al., 2013), which is consistent with the recognition that firms operate in environments with multiple conflicting stakeholder expectations and institutional forces (Marano and Kostova, 2016).
In summary, objectives of organizations go beyond simple profit maximization or compliance with laws and regulations. Today’s successful firms should aim to positively contribute to a social cause (Carroll, 1979; McWilliams and Siegel, 2001; Baumann-Pauly et al., 2013). Therefore, this research also emphasizes additional behaviors and activities that are
not necessarily and explicitly mentioned in legal and economic responsibilities. The next section explains why CSR became especially important for international business.
Integration of CSR into international business
According to Muchlinski (2001), MNEs are encouraged by societies to play a part in solving social problems such as human rights violations. An MNE can be defined as ‘’a firm that owns and controls activities in two or more different countries’’ (Buckley and Casson, 2009, p. 1564). States are no longer the only institution to solve social problems, but MNEs have been expected to adopt a policy of social responsibility to address challenges in the world (Muchlinski, 2001). The past years have seen renewed attention for the social and environmental influences of international business (Kolk and Tulder, 2010). Kolk and Tulder (2010) also argue that this is a result of pressing global problems such as human rights violations, climate change, poverty, and inequality. Kolk and Tulder (2010) agree with Muchlisnki (2001) by arguing that especially MNEs have been expected to solve such social problems. This is because MNEs are confronted with a wide range of issues, stakeholders and institutional contexts in home and numerous host countries. Marano and Kostova (2016) even argue that stakeholders and institutional forces have a significant impact on the CSR practices of a firm.
Since business norms and standards, regulatory frameworks, and stakeholder demand for CSR vary substantially across countries, regions, and lines of business (Williams and Aguilera, 2008; Marano and Kostova, 2016), organizations will not be able to conform to all stakeholder expectations (Marano and Kostova, 2016). Therefore, MNEs must carefully consider how they manage conflicts of interest (Polonksy and Jevons, 2009; Marano and Kostova, 2016). Husted and Allen (2006) even note that when MNEs do not manage CSR strategically, this may lead to economic consequences. Many scholars seem to focus on the
tension between business and society rather than the interdependence between them, which will be explained in the next section.
CSR strategy
According to Friedman (1970), social related initiatives lead to higher profit margins. Organizations should see CSR as a tool to manage stakeholder expectations. Porter and Kramer (2006) also argue that CSR can be more than just a constraint. Strategic management of CSR practices can give organizations significant benefits such as a sustainable competitive advantage or performance improvement (Porter and Kramer, 2006). Burke and Logsdon (1996, p. 496) suggest that corporate social responsibility is strategic when it ‘’yields substantial business-related benefits to the firm, in particularly by supporting core business activities and thus contributing to the firm’s effectiveness in accomplishing its mission’’. This definition reflects mainly the economic layer of Carroll’s pyramid (1979). Since CSR encompasses more than maximizing profits (Carroll, 1979; McWilliams and Siegel, 2001; Baumann-Pauly et al., 2013), it is important to look into this concept further.
MNEs can develop global or local CSR strategies (Muller, 2006). A global strategy is probably more efficient because it is centrally coordinated. However, this strategy may lack ownership and legitimacy at a local level. A local strategy is more locally responsive but is also criticized for lack of consistency across subsidiaries. In addition, this strategy makes it difficult to coordinate all divergent approaches. When a firm adopts a global strategy, CSR practices of the home-country are the basis of firm-wide CSR strategies (Muller, 2006). In contrast, when an MNE adopts a local strategy, host-country stakeholders drive the firm’s strategy (Muller, 2006). Thus far, it remains unclear which type of CSR strategy world’s largest MNEs adopt and how they exactly manage the expectations of stakeholders from different countries.
According to Attig et al. (2016), the importance of addressing stakeholder expectations increases with the degree of internationalization. Internationalization can be defined as ‘’the process by which the firm increases its involvement in international markets’’ (Dutot, Bergeron and Raymond, 2014, p. 674). Firms operate across geographical boundaries to create a competitive advantage, which in turn will lead to increasing profits, growth opportunities, as well as access to new knowledge and resources (Attig et al., 2016).
In general, it depends on the degree of internationalization whether an MNE can be categorized as a global company or not. According to Rugman and Verbeke (2004), MNEs are global when 20% of sales or more is achieved in the North American Free Trade Agreement (NAFTA), the European Union (EU) and Asia, but no more than 50% of sales in one region of this triad. The authors demonstrate that some MNEs are host region oriented. However, most of the world’s 500 largest MNEs have the vast majority of their sales within their home region. These MNEs benefit from higher degrees of home regional integration rather than facing a fully globalized competitive environment.
Rugman and Verbeke (2004) challenged the assumption that MNEs all operate on a global scale, which was associated with the belief that firms now operate in a globally integrated world. A more recent study of Asmussen and Goerzen (2013) also concludes that MNEs had most of their sales in the home region. Because these MNEs are mainly confronted with home region stakeholder expectations, it is expected that they will adopt mainly CSR requirements of the home country. However, this has not been researched yet. Therefore, more research and knowledge is needed regarding the role of internationalization in CSR strategy. Note that Muller (2006) argues that an MNE develops a global CSR strategy when they mainly adopt CSR practices of the home-country, while Rugman and Verbeke (2004) suggest that firms are not operating on a global scale when they have the vast majority of sales within their home region. This seems to be contradicted, but at the same time important to realize.
So far we have discussed the concepts CSR and CSR strategy. We can now move to the evaluation of the concept corporate governance, which is expected to be an important concept when it comes to CSR strategy.
Corporate governance
The concept of corporate governance
There is no universal and clear definition of the concept corporate governance. According to Aguilera and Jackson (2010, p. 487), corporate governance can broadly be defined as ‘’the study of power and influence over decision making within the company’’. Aoki (2000, p.11) defines corporate governance more specifically as ‘’the structure of rights and responsibilities among the parties with a stake in de firm’’. The debate about which parties have a stake in the organization is one that has been going on for a while and has led to different definitions of corporate governance. Some definitions are focused on protecting shareholders from the self-interested actions of executives, while other scholars take the complex and wider relationships among different stakeholders into account in their definitions (Aguilera and Jackson, 2010). These different perspectives on the concept corporate governance will be explained in the next section. First, the systems through which corporate governance can work will be described to get a better understanding of the concept.
Haxhi and Aguilera (2012) note that there are two regulatory systems through which corporate governance can work. First, through hard laws such as statutory rules. Second, through soft laws like codes of good governance. According to Aguilera and Cuervo-Cazurra (2004, p. 417), codes of good governance can be defined as ‘’a set of best practice recommendations regarding the behavior and structure of the board of directors of a firm’’. For example, codes of good governance can provide recommendations on the role and composition of the board of directors, dismissal of directors and top managers and relationships with
shareholders and top management. Codes of good governance complement the binding legal system and try to improve the quality of corporate governance (Aguilera and Cuervo-Cazurra, 2004). The content of these codes differs across countries and compliance with these codes is voluntary. However, research indicates that firms comply to a large extent with the recommendations because they see the need for a good governance system (Aguilera and Cuervo-Cazurra, 2004).
The need for an effective governance system has even been increased in the last years. In 2007, corporate governance has been well discussed in business (Larcker and Tayan, 2008). The media brought negative attention to inordinate compensation, corporate fraud, and other scandals. Central to these cases was the failure of the corporate governance system. The separation of owners and managers gave executives the chance to gain financial or other benefits for themselves. Improving corporate control and reducing conflicts of interest has therefore become more important. However, opinions are divided over which control mechanisms lead to effective corporate governance (Larcker and Tayan, 2008). This has led to differences in corporate governance systems across countries, which will be discussed in the next section.
Corporate governance models
There are two main types of corporate governance models: the Anglo-Saxon model and the Rhineland model. The Rhineland model can be seen as the stakeholder approach to corporate governance. A stakeholder can be defined as ‘’any group or individual who can affect or is affected by the achievement of the organization’s objectives’’ (Freeman, 1984, p. 46). Organizations have an ethical responsibility to stakeholders such as shareholders, customers, suppliers, employees and the community (Atkinson et al., 1997; Freeman and Philips, 2002). According to the stakeholder approach, corporate governance is ‘’a system that ensures that
the board and management of firms strike a balance between the interests of various stakeholders’’ (Rossouw, 2008). Thus, this model tries to integrate the collective interests. Firms in stakeholder-oriented countries are characterized by strong collective bargaining power and highly involved employees (García-Castro, Aguilera and Ariño, 2013). In addition, this model has a high cultural reliance on institutions such as employee associations, political parties, the government, and the unions (Matten and Moon, 2008). The stakeholder model is premised upon stakeholder theory, which is a useful theory to explain the complex relationships among different stakeholders (Freeman, 1984). This perspective describes the firm as a set of relationships between multiple stakeholders with an interest in the firm and thus a broader set of goals to be satisfied (Freeman et al., 2010).
The stakeholder-oriented Rhineland model is contrary to the Anglo-Saxon model whereby maximizing shareholder value is the primary goal of the firm (Aguilera and Jackson, 2003). Shleifer and Vishny’s (1997, p. 737) define corporate governance as ‘’the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment’’. This is typically a definition of a shareholder-orientated corporate governance model because the emphasis lies on satisfying shareholders. Firms in shareholder-oriented countries are characterized by a short-term orientation, weak collective bargaining and low levels of employment protection (García-Castro, Aguilera and Ariño, 2013). The goal of this corporate governance model is consistent with the agency theory as this theory suggests that firms are economic entities with the goal of maximizing shareholder value (Larcker and Tayan, 2008).
Researchers traditionally studied corporate governance within the framework of the agency theory. They viewed the firm as a collection of contracts between shareholders and managers (Aguilera and Jackson, 2003). However, according to Aguilera and Jackson (2003), the agency theory overlooks the various identities of stakeholders. Agency theory suggests that shareholders are the only stakeholders that are facing residual risk. Therefore, an important
assumption of the agency theory is that employee interests are only seen as an exogenous parameter (Aguilera and Jackson, 2003). This assumption is contrary to the recognition that employees are a key stakeholder group (Tylecote & Visintin, 2008; Westermann-Behaylo et al., 2010). Furthermore, agency theory overlooks how corporate governance is shaped by the institutional environment (Aguilera and Jackson, 2003).
Agency theorists also criticized the stakeholder theory. For example, Jenson (2001) argues that stakeholder theory makes it impossible for managers to make effective decisions. This is because stakeholder theory does not learn managers how to make necessary tradeoffs among contradicting interests of multiple stakeholders (Jenson, 2001). In addition, the lack of a clear objective function can lead to undermining mechanisms that are designed to foster managerial accountability to shareholders (Jensen, 2001). In summary, the two different corporate governance models brought up an interesting debate about how companies should respond to stakeholder expectations.
Governance practices and strategies of MNEs are generally aligned with the model that is dominant in the home country (Rossouw, 2008; García-Castro, Aguilera and Ariño, 2013). Rossouw (2008) notes that countries with an Anglo-Saxon model are: Argentina, Australia, Brazil, Chile, Colombia, Czech Republic, Finland, Ireland, Mexico, Nigeria, Peru, Portugal, Singapore, Sweden, Switzerland, United Kingdom, United States, and Venezuela. In contrast, countries with a Rhineland model are: Belgium, China, Denmark, France, Germany, Ghana, Hungary, India, Italy, Kenya, Lithuania, Malawi, Mauritius, the Netherlands, Poland, Romania, Russia, Slovakia, South Africa, Spain, Tanzania, Turkey, Uganda, Zambia, and Zimbabwe (Rossouw, 2008). Canada is considered a hybrid case because this country features both the shareholder and stakeholder orientation. The governance model of Japan moved from a shareholder model towards a stakeholder orientation that balances a variety of stakeholder
interests (Larcker and Tayan, 2008). Therefore, this research associates the corporate governance model of Japan with a stakeholder model.
MNEs operating in these countries are embedded in different institutional environments, which influences their corporate governance system (Williamson, 2000; Aguilera and Jackson, 2004; Doh and Guay, 2006; Aguilera and Jackson, 2010). To better understand why MNEs behave as they do toward stakeholders, this research proposes that understanding of institutional environments is important. Therefore, the next section sheds lights on institutional theory.
Institutional environment
Institutional theory
According to North (1991, p. 97), institutions are ‘’the humanly devised constraints that structure political, economic and social interaction’’. These constraints consist both of informal constraints (such as sanctions, taboos, customs, and traditions) and formal rules (such as constitutions, laws and property rights). Scott (1995) developed three pillars of institutional order that build the institutional environment. First, the regulative pillar represents the formal rules as imposed by the government. Second, the normative pillar defines what is appropriate or expected in a social situation, which influences the behavior of organizations. Third and last,
cognitive pillars are the cultural aspects that guide the understanding of the nature of reality.
These three pillars build the institutional environment that provides stability and meaning to social life (Scott, 1995). This thought goes in line with the recognition that institutions have been established to create order and reduce uncertainty in exchange (North, 1991). Both definitions provide a basis for legitimacy. However, North (1991), as an economist, lays emphasis on the rules and norms that guide how individuals, organizations, and markets interact with each other, while Scott (1995), as a sociologist, emphasizes the regulative and
normative pillars. In the end, these definitions complement each other. Therefore, incorporating both definitions will lead to a better understanding of the concept.
Institutional theory is based on the premise that institutions are stable structures that guide behavior (Meyer and Rowan, 1977. DiMaggio and Powell, 1983). This signifies that organizations do not always make choices as individuals, but rather conform to institutionalized rules to increase their legitimacy. Marano and Kostova (2016) also recently emphasized that organizational practices reflect the institutional environment in which the firm operates. Laws, rules and social norms exert pressures on organizations to adopt institutionalized structures, processes, and practices. As a result, organizations become increasingly similar (DiMaggio and Powell, 1983). Global institutional pressures even contribute to cross-national isomorphism, which explains why there is some convergence in the adoption of governance structures across countries (Aguilera and Cuervo-Cazurra, 2004).
Organizations operating across different fields are less institutionally embedded in a single context but are embedded in multiple environments (Kostova et al., 2008; Marano and Kostova, 2016). Since these environments differ from one another on a number of institutional dimensions, this leads to increasing complexity firms have to deal with (Berry et al., 2010; Marano and Kostova, 2016). Balancing both home and host country requirements has therefore generated attention in recent years (Marano and Kostova, 2016). Some MNEs adopt many requirements of the home country, while others solely adopt formal aspects or even do not adopt local requirements at all. This choice will depend on the attractiveness of the institutional environments of host countries because not all institutional environments contribute in a positive way to organizational practices (Khanna, Palepu and Sinha, 2005; Muller 2006; Marano and Kostova, 2016). The next section explains this more in depth.
Institutional environment shaping CSR requirements
Some host countries are more favorable than others with regard to CSR (Muller, 2006; Marano and Kostova, 2016). The more favorable countries contribute in a positive way to the adoption of a practice through laws, regulations, and rules (Marano and Kostova, 2016). For example, some countries have established laws that ensure strong employment and environmental protections. MNEs that have established subsidiaries in favorable countries can benefit from such laws. This indicates that firms are not constrained within the national environment where the headquarter of the MNE is located with regard to CSR, but are instead exposed to a variety of CSR requirements across numerous host countries (Marano and Kostova, 2016).
Luo and Tung (2007) note that emerging market multinational enterprises (EM MNEs) use international expansion as a springboard to acquire strategic resources and reduce their constraints at home. Marano et al. (2016) also argue that MNEs from developing countries often internationalize to escape the poor institutional conditions at home. Thus, they might also use the stringent CSR requirements of host countries in developed countries in which they operate as a strategic resource to overcome their latecomer disadvantage or compensate for the poor home-country image.
Marano and Kostova (2016) note that in their research American MNEs were mainly influenced by CSR requirements of other developed countries because developed countries have more stringent CSR requirements. These findings are similar to the results found by Muller (2006), who argues that when the local context is a developing country with lower CSR requirements and less public pressure, MNEs will not adopt local CSR requirements. MNEs then adopt the lower rather than the higher standards that may be expected from them in their home country.
In general, CSR requirements in developing countries are relatively weak because the institutional backgrounds for CSR in developing countries are established relatively late
(Gugler and Shi, 2009). Marano et al. (2016) also note there is underdevelopment of institutions in developing countries. This means that there are no effective mechanisms that prevent corruption or protect property rights for example, which in turn makes managers of MNEs in developed countries less susceptible to imitate firms in developing host countries (Marano and Kostova, 2016). However, EM MNEs are catching up because they are also embedded in developed institutional environments. This explains inconsistency between practice and the literature. To decrease this inconsistency, more research needs to be done on this topic.
Existing gap
As described above, MNEs are under increased pressures from stakeholders to provide a high level of social responsibility (Marano and Kostova, 2016). In addition, they are embedded in multiple complex institutional environments (Kostova et al., 2008), which makes it difficult to manage all these stakeholder expectations. Developing a CSR strategy can help managers to address such challenges and improve business performance (Husted and Allen, 2006; Porter and Kramer, 2006; Muller, 2006). There has, however, little research been done on this concept. Therefore, this research tries to explore the concept of CSR strategy further. Corporate governance is expected to provide contributions to the understanding of CSR strategy because corporate governance is all about decision making within a company, and a company needs to make decisions regarding CSR. In addition, the degree of internationalization and the strength of CSR requirements of the home country are expected to have a role in this relationship. However, this has not been researched yet. Therefore, this research addresses the following research question: What is the effect of corporate governance on the CSR strategy of
multinational enterprises and what are the roles of the degree of internationalization and the strength of CSR requirements of the home country herein? The next section elaborates on the
discussed theories of the literature review, which leads to the development of hypotheses. This conceptual framework provides a theoretical answer to the research question.
Conceptual framework
Corporate governance and CSR strategy
As described above, firms have different perspectives on the concept of corporate governance. In this research, corporate governance is defined as ‘’the study of power and influence over decision making within the company’’ since this definition is quite neutral (Aguilera and Jackson, 2010, p. 487). This study focuses on the difference between shareholder and stakeholder orientation because these contrasting ideal-types contain differentiating institutional characteristics. Since governance practices and strategy are generally aligned with the national model (García-Castro, Aguilera and Ariño, 2013), the comparative corporate governance research is expected to provide valuable contributions to the concept of CSR strategy.
The shareholder corporate governance model seeks to maximize shareholder value. Firms in shareholder-oriented countries focus mainly on pleasing one stakeholder group, the shareholders, rather than multiple stakeholders (Rossouw, 2008; García-Castro, Aguilera and Ariño, 2013). Shareholder-oriented firms tend to focus on the economic layer of Carroll’s pyramid (1979), which is consistent with the goal of maximizing profits. Conversely, the
stakeholder corporate governance model sees the firm as a set of relationships between
numerous stakeholders with an interest in the firm and thus a broader set of goals to be satisfied (Rossouw, 2008; García-Castro, Aguilera and Ariño, 2013). This corporate governance model reflects Carroll’s pyramid (1979) as a whole because the stakeholder approach also integrates the collective interests.
When the shareholder corporate governance model is dominant in the home country, it is more likely that the MNE adopts a global CSR strategy. This is because a global strategy is centrally coordinated and therefore less focused on satisfying multiple stakeholders in different host countries (Muller, 2006). Although Friedman (1970) notes that social related initiatives lead to increasing profit margins, we still can expect that firms in shareholder-oriented countries implement a global strategy since these firms typically focus on shareholder expectations rather than multiple stakeholder expectations.
Because the stakeholder corporate governance model seeks to balance the interests of different stakeholders that are located in multiple host countries, MNEs with this corporate governance model will probably try to meet all these expectations. Therefore, it is expected that MNEs with a stakeholder corporate governance model implement a local CSR strategy.
H1a: MNEs with a shareholder corporate governance model are more likely to adopt a global
CSR strategy
H1b: MNEs with a stakeholder corporate governance model are more likely to adopt a local
CSR strategy
Moderating role of internationalization
As mentioned before, internationalization refers to ‘’the process by which the firm increases its involvement in international markets’’ (Dutot, Bergeron and Raymond, 2014, p.674). Firms operating in many different countries have the opportunity to utilize location-specific advantages such as local CSR standards to better exploit their CSR practices and to satisfy stakeholder expectations (Muller, 2006). Internationalization enables firms to customize their CSR practices for different markets. Research indicates that internationalization is associated with increased activities in CSR, proposing that the importance of addressing stakeholder
expectations increases with the degree of internationalization (Attig et al., 2016). Therefore, it is expected that the degree of internationalization moderates the relationship between corporate governance and CSR strategy. Moderation can be defined as ‘’the combined effect of two variables on another’’ (Field, 2014, p. 395). For example, when a stakeholder-oriented MNE has achieved the majority of sales in the home country, we can expect that the MNE also adopts mainly CSR requirements of the home country because they are less confronted with stakeholder expectations from host countries. This may weaken or change the relationship between corporate governance and CSR strategy because it was expected that it is more likely that stakeholder-oriented MNEs adopt local CSR strategies. Therefore, the following hypothesis is developed:
H2: The degree of internationalization moderates the relationship between corporate
governance and the CSR strategy of MNEs
The outcome of this hypothesis is expected to explain the relationship between corporate governance and CSR strategy further. Moreover, the outcome of this hypothesis is expected contribute to the long-standing debate on internationalization by incorporating a new dimension.
Moderating role of the strength of CSR requirements of the home country
As explained before, developed countries contribute in a positive way to the adoption of CSR practices through rules, laws, and regulations because they have more stringent CSR requirements (Muller, 2006; Gugler and Shi, 2009; Marano and Kostova, 2016). Therefore, Muller (2006) argues that the probability that an MNE from a developed country adopts local CSR requirements of a developing country diminishes. This is because the MNE then will
target the lower rather than the higher standards that are expected from the home country. In contrast, the likelihood that an MNE from a developing country uses CSR requirements of developed countries increases because they are expected to use international expansion as a springboard to reduce their constraints at home (Luo and Tung, 2007). This might partially explain why EM MNEs are catching up (Marano et al., 2016).
Based on these theories, we expect that the strength of CSR requirements of the home country where the MNE is headquartered moderates the relationship between corporate governance and CSR strategy. For example, when the MNE is stakeholder-oriented and the CSR requirements in the country where the MNE is headquartered are weak, we expect an increase in the probability that the MNE adopts a local CSR strategy. This expectation is based on two reasons. First, since the MNE is stakeholder-oriented, the MNE probably considers stakeholder expectations of host countries in the CSR strategy. Second, the MNE will use the CSR requirements of host countries to reduce their constraints at home. In contrast, the likelihood that the MNE adopts a local CSR strategy diminishes when the MNE is shareholder oriented because shareholder-oriented MNEs focus mainly on shareholder expectations (Muller, 2006). Therefore, the following hypothesis is developed:
H3: The strength of CSR requirements of the home country where the MNE is headquartered
moderates the relationship between corporate governance and CSR strategy of MNEs
Researchers have begun investigating the varying effects of organizations’ operations in strong and weak institutional environments relative to the home country of the firm. Dau (2013), however, argues that there is more research needed to understand these differences. In this research, we will contribute to this debate by looking further into the relationship between institutional environments and organizations responses, using the example of the strength of
CSR requirements. Moreover, given the increasing importance of firms in developing countries in the global economy (Marano et al., 2016), the result of this hypothesis is expected to provide valuable insights. In addition, testing this hypothesis will help to explore the CSR strategy literature. It will add to previous work by considering a novel concept.
A visualization of the conceptual framework is provided below.
Methods
Research strategy
The objective of this research is to test existing theory. More precisely, this study tests whether corporate governance indeed influences the CSR strategy of multinational enterprises because theory suggests that differences in corporate governance may lead to the development of different CSR strategies. A deductive approach is used because deduction involves the testing of a theoretical proposition (Saunders and Lewis, 2012). More specifically, an explanatory type of study is chosen since this method is most suitable for testing relationships between variables (Saunders and Lewis, 2012). Explanatory studies are limited by the variables the researcher puts in the model. Impacts of other potential variables are therefore not taken into account. However, since there is a need to test the influence of corporate governance on the CSR strategy of MNEs, this type of study seems most appropriate. Moreover, given the lack of especially quantitative studies in the research domain of CSR strategy, a quantitative method is expected to provide valuable outcomes.
Data collection procedure
Secondary data is used to investigate the relationship between corporate governance and CSR strategy. Data is obtained from UNCTADstat, a free to use platform that provides data of companies from both developed and developing countries (UNCTAD, n.d.). This research uses two datasets. First, the top 100 non-financial MNEs from developing economies of 2014 ranked by foreign assets. Second, the world’s top 100 non-financial MNEs from developed economies of 2015 ranked by foreign assets. This were publicly available datasets that could be used to test the hypotheses of this research, which will be briefly explained in the next section. The initial dataset consists of 200 records. The first step was to exclude the MNEs that cannot be categorized as an ‘Anglo-Saxon’ or ‘Rhineland’ model according to Rossouw
(2008), a number of 164 records remained. The second step was to exclude countries that cannot be categorized as a developed or developing country according to the MSCI ACWI and frontier market index, a number of 162 records remained. MSCI is an organization that provides research-based indexes and analytics (MSCI, n.d.). The third step was to exclude the MNEs that do not use the GRI Standards in their CSR reports because these standards will be used to determine which CSR strategy the MNE developed. GRI Standards are the first global standards for sustainability reporting, which can be considered as synonymous with the term corporate social responsibility. The standards help companies to measure, understand and manage their economic, environmental and social performance. GRI’s reporting standards are foundational to the success of CSR reporting, and therefore, important to take into account (GRI Standards, n.d.). In the end, anumber of 120 records remained that met all the criteria.
Description of the sample
As mentioned above, the sample of this research consists of the top non-financial MNEs from developed and developing countries. 120 MNEs are investigated and the following information about the companies is known: home country, foreign assets, total foreign assets, and industry. This sample is appropriate for this research since the dataset makes it possible to determine which corporate governance model is most dominant in the home country, which is needed to test hypotheses 1a and 1b. The MNEs in the sample are categorized as ‘shareholder’ or ‘stakeholder’ corporate governance model by Rossouw (2008). Table 1 provides information on the corporate governance models of the countries in this sample. Furthermore, it is possible to determine whether the firm is headquartered in a ‘developed’ or ‘developing’ country, which is needed to test the third hypothesis. Table 2 categorizes countries in the sample as a developed or developing market according to the MSCI ACWI and frontier market index.
Table 1. Counts of MNEs according to corporate governance by country
Shareholder model Stakeholder model
Country Frequency Country Frequency
Australia 1 Belgium 1
Brazil 3 China 13
Mexico 2 France 9
Sweden 2 Germany 9
Switzerland 4 India 8
United States 20 Italy 3
United Kingdom 14 Japan 10
Netherlands 2 Russian Federation 2 South Africa 7 Spain 3 Total 53 67 N = 120
Table 2. Counts of MNEs according to the type of market by country
Developed Developing
Country Frequency Country Frequency
Australia 1 Brazil 3
Belgium 1 China 2
France 9 India 8
Germany 9 Mexico 13
Italy 3 Russian Federation 2
Japan 10 South Africa 7
Netherlands 1 Singapore 7 Spain 3 Sweden 2 Switzerland 4 United Kingdom 14 United States 20 Total 85 35 N = 120
In addition, all firms published annual or CSR reports, which makes it possible to determine which CSR strategy they have developed. Lastly, information about foreign assets and total assets is given for each MNE, which makes it possible to test the second hypothesis.
Measures
Dependent variable. The dependent variable in this research is the ‘CSR strategy’. Due to the
fact that quantitative CSR studies are mainly focused on concepts such as CSR performance and CSR dimensions, the concept CSR strategy is underdeveloped. There are some researchers that used a qualitative approach to investigate CSR strategies, but there is a lack of quantitative measurement. Therefore, a quantitative measurement was created to determine whether the MNE adopts a local or global CSR strategy. The theory of Muller (2006), Carroll’s pyramid (1979) and the GRI Standards were used to operationalize the concept CSR strategy. Economic, environmental and social responsibilities are main categories in the GRI Standards. These three categories are linked to the four categories of Carroll (1979). More than fifty CSR reports were scanned before the measurement was created. By searching on words such as ‘local’, ‘regional’, ‘international’ and ‘global’, it turned out that many MNEs report about local economic development, local laws, local environmental legislation, local HR policies and local communities. These subjects also fit with the categories of Carroll (1979) and GRI Standards and are therefore used to determine the CSR strategy of MNEs. The following measurement is eventually created:
1. Economic: Indicator takes value 1 if the report describes that the MNE supports local economic development, otherwise it takes a value of 0
2. Legal:
a. Indicator takes value 1 if the report describes that the MNE adapts to general local laws and regulations, otherwise it takes a value of 0
b. Indicator takes value 1 if the report describes that the MNE complies with specific relevant local environmental legislation, otherwise it takes a value of 0
c. Indicator takes value 1 if the report describes that the MNE adapts to local HR policies, otherwise it takes a value of 0
3. Ethical: Indicator takes value 1 if the report emphasizes that the MNE works with local governments and communities to improve the lives of those communities, otherwise it takes a value of 0
4. Discretionary: As mentioned before, this category is hard to find. According to Carroll (1979), examples of voluntary activities might be ‘making philanthropic contributions, conducting in-house programs for drug houses, training the unemployed, or providing day-care centers for working mothers. If a business does not engage in these activities, it is not considered unethical per se (Carroll, 1979). These discretionary activities are voluntary and subjective. This means that it is not part of the GRI Standards. Therefore, this category is excluded from the measurement.
A total score of 5 can be realized. When the MNE scores 0, 1 or 2, we assume that the MNE developed a global CSR strategy. A score of 3, 4 or 5 indicates that the MNE developed a local CSR strategy.
Independent variable. The ‘corporate governance’ variable is based on the home country of the
MNE. The overview of countries that are categorized as Anglo-Saxon or Rhineland by Rossouw (2008) is used to determine which corporate governance model is dominant in the home country of the firm (see table 1). A ‘0’ is reported when the Anglo-Saxon model is dominant in the country where the MNE is headquartered, while a ‘1’ is reported when the Rhineland model is dominant in the country where the MNE is headquartered.
Moderators. There are two moderators in this research that may affect the relationship between
corporate governance and CSR strategy. The first moderator is the variable ‘degree of internationalization’. Foreign assets are divided by total assets because this is a common way to measure this variable (Rugman and Li, 2007).
The second moderator is the ‘strength of CSR requirements in the home country’. Due to lack of data availability, the difference between ‘developed’ and ‘developing’ countries are used as
a proxy for this variable. Developed countries clearly signal for relatively strong institutional backgrounds for CSR, while developing countries are known for relatively weak institutional backgrounds for CSR (Gugler and Shi, 2009). The MSCI ACWI and frontier markets index is used to classify the home country where the MNE is headquartered as a ‘developed’ or ‘developing’ country (see table 2). A ‘0’ is reported if the home country where the MNE is headquartered is a developed country, while a ‘1’ is reported if the country where the MNE is headquartered is a developing country.
Control variables. This research includes control variables to make sure that the relationship
between the above-mentioned variables is not confounded by other variables. The control variables in this research are size and industry. Size is based on the total assets of the firms in the sample. Size is included as a control variable because larger firms face greater public scrutiny over their social and environmental impacts and thus are likely to invest more in CSR (Christmann and Taylor, 2001; Wickert et al., 2016). Industry is included as a control variable because a lot of studies show that it might influence CSR initiatives (Arora and Dharwadkar, 2011; Marano et al., 2016; Marano and Kostova, 2016). The standard industrial classification system (SIC) is used to categorize the firms in the sample by industry because SIC is commonly used in government and private data (SIC code, n.d.). An overview of all the variables and their measures is provided in Table 3.
Table 3. Overview variables and sources
Variable Variable
type
Definition Values Source
CSR strategy Dependent There are no or not enough indications that the MNE adopts CSR standards of the host countries (global)
versus MNE adopts
CSR standards of the host countries (local)
0 = Local 1 = Global Annual reports and CSR reports Corporate Governance
Independent The corporate
governance model that is dominant in the home country of the MNE Anglo-Saxon Rhineland Database UNCTAD, based on Rossouw (2008) Degree of internationalization
Moderator Foreign assets as a percentage of total assets Outcome will be a percentage Database UNCTAD Strength CSR requirements home country
Moderator Home country of the MNE classified as developed or developing country (Developed country = stringent CSR requirements, developing country = less stringent CSR requirements) Developed country Developing country Database UNCTAD, based on MSCI ACWI & frontier markets index
Industry Control The industry in which the MNE operates
1= Mining 2= Construction 3= Manufacturing 4 = Transportation, Communications, Electric, Gas & Sanitary service 5= Wholesale 6= Retail Trade 7 = Finance, Insurance & Real Estate
8 = Services
Database UNCTAD and SIC codes
Size Control Total assets of the MNE
Value of assets Database UNCTAD
Data analysis
IBM SPSS is used for statistical analysis. First of all, there has been a check of frequencies to examine if there were any errors in the data. There were no errors found. Furthermore, all variables under investigation were checked for missing data before running descriptive statistics, the correlation matrix and the statistical tests. The study does not include an inter-reliability test such as Cronbach Alpha since all variables are based on single item measures. Cronbach’s Alpha, for instance, would be useful when a variable consists of multiple Likert-scale items (Field, 2014).
A binary logistic regression model is used to test the relationship between corporate governance and CSR strategy. This type of regression analysis is most suitable because the dependent variable is dichotomous in nature. Logistic regression does not make many of the key assumptions of linear regression like linearity, normality, and homoscedasticity. It is important that the dependent variable is dichotomous, observations are independent, and that there are no high (>0.7) intercorrelations among independent variables (Field, 2014).
Before running the binary logistic regression analysis, crosstabs command is used to obtain a crosstab of the two variables. A crosstab depicts the number of times each of the possible category combinations occurred in the sample data (Field, 2014). This is needed to determine whether MNEs with a shareholder corporate governance model are indeed more likely to adopt a global CSR strategy and MNEs with a stakeholder corporate governance model are more likely to adopt a local CSR strategy. To test whether the relationship is significant, the binary logistic regression analysis was performed. A dummy variable was created for the dependent variable CSR strategy because this variable is categorical in nature. A dummy-coded variable can be treated as numerical even though it is a categorical variable in nature, which is needed to run a regression model. The Wald chi-square test, the Omnibus tests of model Coefficients and the Hosmer-Lemeshow are widely used for logistic regression
and thus the outputs of these tests are used to support or reject hypothesis 1a and 1b. After running the binary logistic regression, this research controlled for the variables size and industry. The control variables were entered in the same way as the independent variable ‘corporate governance’ and thus another binary logistic regression was performed.
In order to test the moderating effects, the Process macro written by Andrew F. Hayes (2013) for SPSS was used. More specifically, model number 1 represents the moderation analysis. Therefore, this model is used to test hypothesis two and three. The two moderators are tested in two separate models with both the same predictor variable. When the result of the moderation analysis was significant, a closer inspection of the conditional effects has been done.
Results
This section provides an overview of the descriptive statistics, the result of the binary logistic regression analysis and the moderation analysis.
Descriptive statistics
Table 4 below shows descriptive information and correlations between the variables in the total sample included in this study.
Table 4. Means, Standard Deviations, Correlations
Means, Standard Deviations, Correlations
Variables M SD 1 2 3 4 5 6 1. CSR strategy 0.35 0.48 -
2. Corporate governance 0.56 0.50 0.512* -
3. Degree of internationalization 0.62 0.22 0.045 -0.088 -
4. CSR requirements 0.28 0.45 0.198* 0.373** -0.139 -
5. Firm size (millions of dollars) 111.67 111.36 0.072 0.047 -0.237 -0.237** -0.239** 6. Industry 4.02 1.77 0.045 0.121 -0.055 -0.095 0.07 - ** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.05 level (2-tailed)
The results demonstrate that Corporate governance positively correlates with CSR strategy (r = 0.51, p < 0.05). Moreover, this relationship is significant, which means that it is likely that
Corporate governance influences the CSR strategy of MNEs. As mentioned before, if
correlations between some of the independent variables are high (> 0.7) and significant, you should be mindful of potential multicollinearity problems (Field, 2014). Table 4 indicates that there are no high potential multicollinearity problems in this research.
Binary logistic regression
Before reporting the results of the binary logistic regression analysis, the frequency distribution of the variables is shown. Table 5 below provides a basic picture of the interrelation between corporate governance and CSR strategy.
Table 5. Crosstab corporate governance and CSR strategy
Corporate governance
CSR strategy Anglo-Saxon Rhineland Total
Global 49 29 78
Local 4 38 42
Total 53 67 120
The results suggest that 49 out of 53 (92.5%) MNEs with a shareholder corporate governance model adopt a global CSR strategy. Thus, it seems more likely that MNEs with this corporate governance model adopt a global CSR strategy. 38 out of 67 (56.7%) MNEs with a stakeholder corporate governance model adopt a local CSR strategy. Thus, more than half of stakeholder-oriented MNEs adopt a local CSR strategy. The overall accuracy of the classification is 72.5%. This signifies that 72.5% of the model is correctly specified.
Next, binary logistic regression analysis is performed to examine whether the relationship between corporate governance and CSR is significant. Before performing the analysis, assumptions of logistic regression that are described in section 3.5 were checked. None of the assumptions were violated and thus the Wald chi-square test, the Omnibus tests of model Coefficients and the Hosmer-Lemeshow tests were performed to test hypothesis 1a and 1b. Table 6 provides an overview of the outcomes of these tests.
Table 6. Binary logistic regression analysis
B SE Wald df p Exp (B)
Constant 0.619 0.191 10.462 1 0.001 0.538
Chi-square
Overall model evaluation Omnibus Tests of
Model Coefficients 35.356 1 0.000
Goodness-of-fit test
Hosmer & Lemeshow 7.941 8 0.439
Note. Nagelkerke pseudo R2 = 0.355, p <0.001
The Wald chi-square is used to test whether the b coefficient for a predictor in logistic regression is significantly different from zero (Field, 2014). The table presents a p-value of 0.001. A value smaller than the critical value of 0.05 indicates that the result is significant. Therefore, we are able to reject the null hypothesis. This means that the coefficient is not simultaneously equal to zero and therefore Corporate governance has predictive capacity.
The results of the Omnibus tests of model Coefficients indicate the overall model evaluation (Field, 2014). The p-value is smaller than the critical value of 0.05, which means that the overall model is statistically significant. The Hosmer-Lemeshow statistic presents a p-value of 0.439. A p-p-value greater than the critical p-value of 0.05 indicates that the model is correctly specified and that the observed and expected observations are close to each other (Field, 2014). In summary, the results are significant and the overall goodness-of-fit is correctly specified. Thus, there is sufficient evidence at the alpha level of significance to support hypothesis 1a. There is, however, less support for hypothesis 1b because only 56.7% of stakeholder-oriented MNEs adopt a local CSR strategy. Overall, we can conclude that
explained by the model is 35.5% (Nagelkerke pseudo-R2 = .355). This means that corporate governance explains 35.5% of the variation in CSR strategy.
In order to isolate the effects that could have an impact on the relationship between
Corporate governance and CSR strategy, there were two control variables incurred in the
model: size and industry. The coefficients of these control variables show, however, insignificant findings. This means that both size and industry do not influence the studied relationships.
Moderation
Finally, the moderating effects of Degree of internationalization and the strength of CSR
requirements of the home country on the relationship between Corporate governance and CSR strategy were analyzed. Table 7 provides the result of the second hypothesis.
Table 7. Moderator analysis degree of internationalization
Coefficient SE Z p Intercept i1 -1.01 0.30 -3.43 0.0006 CG (X) c1 1.47 0.30 1.15 0.2482 Degree of internationalization (M) c2 0.35 0.32 4.59 0.0000 Degree of internationalization*CG (XM) c3 -0.18 0.33 -0.56 0.5781 Nagelkerke pseudo-R2 = 0.364, p<0.05
The results demonstrate that the interaction effect is not significant because the p-value is greater than the critical value of 0.05 (B = -1.01, p>0.05). Therefore, the second hypothesis is not supported. This means that there is not enough evidence to conclude that the Degree of
internationalization moderates the relationship between Corporate governance and the CSR strategy of MNEs. Table 8 provides the results of the third hypothesis.
Table 8. Moderator analysis strength CSR requirements of the home country
Coefficient SE Z p
Intercept i1 0.86 0.29 2.93 0.0004
CG (X) c1 1.34 0.26 1.72 0.0145
CSR requirements home country (M) c2 0.45 0.32 4.24 0.0004
CSR requirements home country*CG (XM) c3 0.69 0.29 2.44 0.0147
Nagelkerke pseudo-R2 = 0.398, p <0.05
The interaction effect presented in this table is significant because the p-value is smaller than the critical value of 0.05 (B = 0.86, p<0.05). Therefore, we can conclude that there is enough evidence to support the third hypothesis. This means that the strength of CSR requirements of
the home country moderates the relationship between Corporate governance and the CSR strategy of MNEs. It indicates how much the effect of corporate governance on CSR strategy
is different between MNEs from developed and developing countries (B = 0.69, p<0.05). Moreover, the total variance explained by the model increased with 4.3% when the strength of
CSR requirements of the home country was introduced (Nagelkerke pseudo-R2 = 0.398, p <0.05). This means that this solution explains 39.8% of the variation in CSR strategy. A closer inspection of the conditional effect indicates that the moderating role of the strength of CSR requirements is only significant within the group of MNEs that are from developed countries (B =1.77, p<0.0000). Table 9 below provides the result of this closer inspection.
Table 9. Conditional effect moderator analysis
Conditional effect of CG (X) on CSR strategy (Y) at levels of CSR requirements of home country (M)
Effect SE Z p
Developed 1.77 0.39 4.482 0.0000
Thus, we conclude that having a shareholder-oriented corporate governance model increases the probability of the development of a global CSR strategy, but only if the MNE is from a developed country. In addition, having a stakeholder-oriented corporate governance model increases the probability of the adoption of a local CSR strategy, but only if the MNE is from a developed country. Table 10 gives a quick overview of the hypotheses that have been tested and the corresponding results.
Table 10. Summary hypotheses
Hypothesis p-value Supported
H1a&1b 0.001 Yes
H2 0.5781 No
