Master Thesis
The Effect of Geographical Diversification on Corporate Social Performance:
Do Institutional Distance and Market of Origin matter?
Peter Uhlig S4106504
p.j.uhlig@student.rug.nl
Supervisor: dr. Kees van Veen Co-Assessor: dr. Andrea Kuiken
MSc International Business and Management
University of Groningen, Faculty of Economics and Business Date of Submission: June 29, 2020
Abstract
Interest in corporate social responsibility (CSR) has increased considerably, particularly in the last two decades (Aguinis & Glavas, 2012; Dorfleitner et al., 2015). The majority of research has been conducted in developed countries, while knowledge about CSR in emerging countries is still limited (Li et al., 2010). The diversification of companies has also been the subject of various studies, although these are mostly limited to the effect of financial performance. The aim of this study is to add to the existing literature by examining the relationship between geographical diversification and CSR performance. Starting from a stakeholder and institutional theory perspective, this study further examines possible moderating effects of institutional distance, as previous studies could already identify direct effects. An OLS regression is performed using a total sample of 150 companies from developed and emerging markets, using data from 2018. The results present mixed findings with respect to existing studies. A positive relationship between geographical diversification and CSR performance was found. Furthermore, increasing informal institutional distance negatively moderates this relationship. No significant results were obtained for moderating effects of formal institutional distance and possible effects of the market of origin. The results of this study partly confirm earlier findings and contribute to the existing body of literature.
Moreover, important practical implications regarding institutional distance as well as theoretical implications are provided.
Keywords: Corporate social responsibility (CSR), geographical diversification, formal institutional distance, informal institution distance, multinational enterprise (MNE)
Acknowledgements
I would like to thank my supervisor, dr. Kees van Veen for his support and advice during the process of writing this thesis. His feedback, advice and encouraging words have helped and been of significant value to me over the past months. I would also like to thank my family and friends who have supported me by listening to my concerns and providing words of encouragement.
Table of Contents
Abstract ... 1
Acknowledgements ... 2
Table of Contents ... 3
List of Figures ... 5
List of Tables ... 5
List of Abbreviations ... 5
1. Introduction ... 6
2. Theory ... 9
2.1 Literature Review ... 9
2.1.1 Corporate Social Responsibility ... 9
2.1.2 Multinational Enterprises and CSR ... 13
2.1.3 CSR in Developed and Emerging Markets ... 15
2.2 Hypotheses Development ... 17
2.2.1 Geographical Diversification ... 17
2.2.2 Developed and Emerging Market MNEs ... 19
2.2.3 Formal Institutional Distance ... 20
2.2.4 Informal Institutional Distance ... 21
2.3 Conceptual Model ... 23
3. Research Methodology ... 24
3.1 Data Collection ... 24
3.2 Sample ... 25
3.3 Variables ... 26
3.3.1 Dependent Variable ... 26
3.3.2 Independent Variables ... 27
3.3.3 Moderating Variables ... 28
3.3.4 Control Variables ... 29
3.4 Data Analysis ... 29
4. Results ... 31
4.1 Descriptive Statistics ... 31
4.2 Correlations ... 32
4.3 Regression Results ... 33
4.4 Robustness Test ... 36
5. Discussion ... 37
6. Conclusion ... 40
6.1 Theoretical Implications ... 40
6.2 Practical Implications ... 41
6.3 Limitations and Future Research ... 41
References ... 43
Appendices ... 52
Appendix A – Classification of Countries ... 52
Appendix B – Stakeholders influencing CSR ... 52
Appendix C - Social CSR and per capita GDP ... 53
Appendix D – Environmental CSR and per capita GDP ... 53
Appendix E – CSR Similarities ... 54
Appendix F – New Institutional World Order ... 54
Appendix G – Institutional Quality Ranking per Country ... 55
Appendix H – Shapiro-Wilk Test ... 55
Appendix I – P-Plot ... 56
Appendix J – White Test ... 56
Appendix K – Breusch-Pagan Test ... 57
Appendix L – Variance Inflation Factor ... 57
Appendix M – Regression Analysis Developed Markets ... 58
Appendix N – Regression Analysis Emerging Markets ... 59
Appendix O – Robustness Test 1 ... 60
Appendix P – Robustness Test 2 ... 61
List of Figures
Figure 1 – ESG Score Grade Example ... 12
Figure 2 – Conceptual Model ... 23
Figure 3 – Market Classification ... 26
List of Tables Table 1 – Descriptive Statistics ... 32
Table 2 – Correlation Matrix ... 33
Table 3 – Regression Analysis ... 35
Table 4 – Hypotheses Overview ... 40
List of Abbreviations
CSP Corporate Social Performance CSR Corporate Social Responsibility
DMNE Developed Market Multinational Enterprise EMNE Emerging Market Multinational Enterprise ESG Environment, Social, and Governance LOF Liability of Foreignness
LOO Liability of Origin MNE Multinational Enterprise
MSCI Morgan Stanley Capital International
1. Introduction
In recent years, public awareness has changed. An increasing number of consumers places value on socially responsible and environmentally conscious production. Through investigative journalism, the work of NGOs, and various climate-movements, an increasing part of the public has been made aware of the behaviour of companies, both locally and abroad. Social media has further increased the dissemination of information in this respect, by enabling the provision of information to a wide range of the population. This increased attention to the possible negative effects of business activities presents companies with the challenge of establishing and complying to certain standards that may go beyond legal requirements.
In 2020, in response to the COVID 19 pandemic, multinational companies Adidas and H&M announced that they would stop paying rents to landlords for their stores in Germany, that had to be closed due to crisis regulations (“Coronavirus in Germany,” 2020). Although the Bundestag had approved a bill that would allow rent payments to be deferred, politicians strongly criticised the decision of these companies as they are in a healthy financial position and their decision could harm private landlords (“Coronavirus in Germany,” 2020). With the hashtag ‘#AdidasBoykott’ German consumers called for a boycott of the Adidas brand (Holtschneider, 2020). Public relations experts also criticised the company, as it took them three days to respond to the criticism (Theile, 2020). As this example demonstrates, socially responsible corporate behaviour is increasingly a necessity in order to comply with public demands and avoid negative publicity.
When companies grow and expand to foreign countries, the business environment becomes even more demanding, i.e. the political and cultural complexity increases when a company operates in more countries (Meznar & Johnson, 1996). This means that, due to increased geographical diversification, a firm has to cope with more stakeholders and different institutional environments. For example, environmental regulations can differ in terms of emission limits, energy-efficiency requirements for machinery or different standards for the disposal of hazardous waste (Stavins, 2005).
Based on Freeman (2010), stakeholders are individuals or groups of a population that have a general interest or a stake in a company and therefore its well-being. In addition, they are affected by the actions of a company and, in return, react to these actions, trying to influence the firm’s behaviour (Freeman, 2010). Based on this, an increased (geographic)
field of operation leads to an exposure to a higher number of stakeholders that the company needs to respond to (Kang, 2013).
Institutional differences also present a significant challenge. “Institutions are the rules of the game in a society” (North, 1990, p. 3), which means that they are the legal and regulatory framework of a country. Companies rely on these formal environments to distinguish between appropriate and inappropriate behaviour in order to gain legitimacy in the home and host markets (DiMaggio & Powell 1983; Scott 1987, as cited in Keig et al., 2019).
Accordingly, the exposure to an increased number of institutions poses an increased challenge to the firm.
Corporate diversification has been the subject of intense study (Chatterjee &
Wernerfelt 1991; Hoskisson & Hit 1990; Montgomery & Hariharan 1991; Wernfelt 1984, as cited in Kang, 2013). However, the research on the impact of diversification is mostly limited to the impact on corporate financial performance (Markides & Williamson 1994; Palich et al.
2000; Rumelt 1974, as cited in Kang, 2013). Kang (2013) notes that little attention is paid to the potential impact of corporate diversification on CSP. Moreover, the global diversification of multinational companies, which leads to increased interdependence, is increasingly cited as having a positive impact on a firms’ social performance, as Brammer et al. (2006) have noted.
The majority of the current studies suggest a positive relationship between geographical diversification and CSP (e.g. Attig et al., 2016; Brammer et al., 2006; Kang, 2013; Keig et al., 2019). However, those studies are limited with either regards to the number of observed firms or use of moderators, such as formal and informal institutional distance.
Therefore, some scholars suggest to perform a similar analysis with a larger number of firms and the implementation of different moderators (e.g. Brammer et al., 2006; Keig et al., 2019).
Additionally, while most research on CSR has been conducted in developed markets (Li et al., 2010), interest in research on CSR in emerging economies is increasing.
Researchers have found that companies in developed markets are more engaged in CSR activities, which leads to better CSR outcomes compared to companies from emerging markets (Baughn et al., 2007). Although some aspects of CSR in developed and emerging markets are similar, e.g. a focus on social media and publicity campaigns, the purpose of conducting CSR-related activities is different (Sharma, 2019). In this respect, when companies move geographically from a developed market to an emerging market and vice versa, they need to engage in organisational learning to adapt to the CSR focus and practices in the respective host market. Starting from the fact that both developed and emerging
countries differ in their CSR scores, this process could also have an impact on the CSR scores of respective companies.
Based on that, this thesis uses a stakeholder and institutional theory perspective to answer the following research question: Which impact does geographical diversification have on CSP, do MNEs from developed and emerging countries differ in this respect, and can this effect be explained by formal and informal institutional distance?
2. Theory
The following section introduces the term CSR, its definition, delimitation to CSP, and previously conducted research on possible relationships. Then the implementation of CSR activities by MNEs is discussed. Thereafter, the differences between emerging and developed markets are outlined and the differences in terms of their overall CSP are highlighted. Finally, hypotheses based on existing concepts and theories are developed and the conceptual model is presented.
2.1 Literature Review
2.1.1 Corporate Social Responsibility Relevance of CSR
The social concerns of companies have been studied especially during the last 50 years (SAGE Brief Guide to Business Ethics, 2012), but over the last two decades in particular, interest in corporate social responsibility (CSR) has grown considerably (Aguinis & Glavas, 2012; Dorfleitner et al., 2015). By and large, it can be understood as a dimension of corporate performance that reflects the extent to which a company implements socially sustainable principles and programs to meet the social interests of its stakeholders (Keig et al., 2019).
However, as Park (2018) noted, it is difficult to define a company as sustainable or unsustainable, and the underlying overall concept is controversial, since it is of subjective nature (Ehrenfeld, 2008). Nonetheless, sustainability can be divided into environmental, social and corporate governance issues (ESG) (Cubas-Díaz & Martínez Sedano, 2018; Fatemi et al., 2018). The ESG pillars are designed as a holistic concept of the term sustainable development (S. B. Park, 2018), that has been defined as “meet[ing] the needs of the present without compromising the ability of future generations to meet their own needs” (Brundtland, 1987, p. 16). However, according to Vurro & Perrini (2011) it is still being discussed whether CSR reporting also leads to better corporate social performance (CSP). It is therefore possible that the mere publication of CSR performance scores will not improve the current situation.
Defining CSR
Despite the growing interest and the increasing relevance of this topic, scientists have not yet been able to agree on a common definition of the term CSR (Dahlsrud, 2008; Sheehy, 2014;
Wan-Jan, 2006; Wood & Logsdon, 2019). Existing definitions have changed over time from
classical to normative, and to instrumental approaches, which shows the history behind this research and the continuous improvement of earlier concepts (Wood & Logsdon, 2019). Wan- Jan (2006) reviewed the various explanations and applications that scientists use to explain CSR, ranging from an excellent marketing instrument to a brand-enhancing tool, but also as something that is just the right thing to do. However, even researchers disagree on whether a common definition is both useful and necessary.
On the one hand, Wan-Jan (2006) argues that a correct definition of CSR is a necessity in order to conduct more in-depth studies, otherwise there is a risk of obtaining distorted results due to an incorrect understanding of the issue. Birch & Moon (2004) stress that CSR stems from national traditions, which vary from region to region, and involve the responsibilities of governments and companies. They therefore argue that there cannot be a single definition applicable worldwide, but rather that there must be several definitions that can be applied according to region and context. This is in line with the findings of Baughn et al. (2007), who examined various studies that distinguish 20 to 50 different aspects of CSR, based on nation-specific policies and variations in the business-society relationship.
In contrast, Dahlsrud (2008) analysed more than 35 existing definitions of CSR and discovered five dimensions, namely an environmental, a social, an economic, a stakeholder and a voluntariness dimension, which are present throughout almost all definitions. On the basis of these results, Dahlsrud (2008) argues that this strong congruency of definitions underlines that, contrary to the general assumption, the lack of a universal definition is not a problematic issue. Instead, he argues for a general understanding of the CSR construct in different contexts, rather than focusing on the search for a general definition. Wan-Jan’s (2006) research supports this by stating that there already exists a general definition in broad terms. A major convergence can be found in the perception of CSR as “[…] serving the needs of appropriate stakeholders” (Wan-Jan, 2006, p. 182).
Based on these results, CSR is considered in this paper as “a company's responsibility beyond return to shareholders to include an acknowledgement of its responsibilities to a broad range of stakeholders throughout society including employees, customers, business partners, communities and the environment” (Business for Social Responsibility, p. 1, as cited in Baughn et al., 2007). This broad definition covers all dimensions of CSR and can be adapted to different contexts.
CSR - CSP relationship
Furthermore, a distinction must be made between the concept of CSR and corporate social performance. While CSR mainly refers to the general belief that companies have a responsibility towards society that goes beyond the interests of shareholders, CSP extends this concept and focuses on the actual results of CSR measures taken by companies (SAGE Brief Guide to Business Ethics, 2012). Performance data therefore represent measures that a company is taking or has taken to address specific problems (Clarkson, 1995). The results relating to the relationship between the company and society must be observable [and measurable] accordingly (Wood, 1991). Furthermore, a lack of data on a particular issue indicates that no actions have been taken, as data is only available when a particular issue is considered important enough by relevant stakeholders to justify its management (Clarkson, 1995). It must also be noted that assuming social responsibility does not automatically lead to results that are beneficial for CSP (SAGE Brief Guide to Business Ethics, 2012).
Having distinguished between the terms CSR and CSP, it must be pointed out that the term CSP can be seen from different points of view. Liston-Heyes & Ceton (2009) examine the difference between perceived and actual CSP in their study. They argue that the perceived CSP ultimately influences the allocation of resources on a capitalistic market which drives companies to invest billions to improve their reputation. Furthermore, they outline that the appropriate tools to measure actual CSP are lacking. This creates a gap that leads to perceived and actual CSP (Liston-Heyes & Ceton, 2009).
There are good incentives for companies to be less transparent about their CSR activities and results, and there can even be manipulation of published data (Wood, 2010).
Ultimately, the perceived CSP is more important to companies than the actual CSP, which can affect company-behaviour, as it determines, inter alia, employer choice, investment selection and purchasing decisions (Liston-Heyes & Ceton, 2009).
Due to this problem, it is particularly important that ESG rating databases are very precise to reflect the actual CSP as accurately as possible. The overall problem with these databases is that there is no general consensus on a uniform approach to measuring CSP, on its quantification and on the aspects to be included in the assessment (Liston-Heyes & Ceton, 2009; Wood, 2010). In most cases, the scores are determined using positive and negative indicators based on company reports, and additional publicly available data is used (Dorfleitner et al., 2015).
This procedure is also used to determine the Thomson Reuters ESG Scores. First, the ESG Score Grade is calculated based on publicly available company reports (Thomson
Reuters ESG Scores, 2017). Then the ESG Controversies Score Grade is calculated, which is based on public reporting and thus also includes scandals (Thomson Reuters ESG Scores, 2017). The ESG Combined Score Grade, which is used in this paper, is then derived from both grades. Figure 1 shows the different scores for a selection of German companies. It can be seen that the scores based on corporate publications and those determined by media coverage can differ considerably. This is in line with the previous explanation.
Figure 1 – ESG Score Grade Example
In addition, it should be noted that the ESG controversy scores are used as an overlay on the ESG score to "[...] discount the ESG performance score based on negative media stories"
(Refinitiv, 2019, p. 7), in order to take into account ESG controversies in which companies might be involved (Refinitiv, 2019).
However, no guarantee can be given that the data used are completely accurate and reflect reality, as the values depend in part on the reporting of companies and the calculations are subject to approximation. It is therefore possible that the scores are above the actual CSP, as it is likely that not all company misconduct has yet been uncovered and discussed in the media.
Motivation to engage in CSR
Corporate social responsibility can have multiple benefits for the firm, when properly implemented. These potential benefits can be a motivation for companies to undertake CSR activities. Proper implementation means decision-making from a strategic management perspective, while the total costs of social undertakings are calculated and the corresponding returns are determined (Debski, 2009 as cited in Żychlewicz, 2015). According to Molteni
(2004, as cited in Patrizia, 2010), CSR can provide four types of benefits to the firm. It can (1) result in increased visibility and reputation, (2) positively affect innovativeness, (3) increase employee motivation, and (4) increase safety and prevent crises (Molteni, 2004 as cited in Patrizia, 2010). Żychlewicz (2015) confirms these possible benefits and adds that CSR activities can reduce costs due to increased health and safety standards, learning or development possibilities, and medical benefits for employees, that increase work efficiency and reduce employee rotation. Furthermore, increased revenue can be achieved by sustainable products which attract new customers or increase brand loyalty (Żychlewicz, 2015). Overall, business risks can be mitigated since effective CSR requires a continuous dialogue with stakeholders, which allows gathering information about the expectations of the different stakeholder groups (Żychlewicz, 2015).
2.1.2 Multinational Enterprises and CSR Defining MNE
The term multinational corporation was first introduced in 1960, although modern multinational corporations date back to the late nineteenth century (Kobrin, 2009). At a university conference, it was defined by Lilienthal (1960) as a company that is based in one country but also conducts business under the laws of a different country (Kobrin, 2009). Since then, varying definitions of the multinational enterprise (MNE) have been proposed, but none of them has become the generally used standard (Aggarwal et al., 2011; Qian, 2019).
Possible definitional approaches have been based on company size by turnover, the ratio of foreign sales or assets, the number of non-domestic subsidiaries and foreign employees (Aggarwal et al., 2011). For example, several authors define and use the definition of the MNE as an enterprise that controls production facilities in at least two countries, with a focus on cross-border entities and operations (Bouquet et al., 2009; Kobrin, 2009; Rangan &
Sengul, 2009). Eurostat (2019, para. 1) defines it as “[…] an enterprise producing goods or delivering services in more than one country”. Aharoni (1971, as cited in Aggarwal et al., 2011) further categorised the definitions into a performance definition - focusing on foreign sales, assets and employees; a structural definition - based on the number of operating countries and the nationality of management; and a behavioural definition - focusing on management's international strategic thinking.
Following the structural approach, in this paper an MNE is defined as a company that operates in at least two countries or has subsidiaries in at least one foreign country. However,
large publicly listed companies are used for the analysis, which ensures business activities in multiple countries.
MNEs and CSR
CSR is a strategic concern for companies (Albuquerque et al., 2019). An increasing number of companies worldwide are conducting CSR-related activities, which also leads to an increased amount of published information regarding social and environmental activities at company level (Tilt, 2016). The resources invested in CSR by MNEs are considerable and investments have increased over the last few years (Asmussen & Fosfuri, 2019). Moreover, global companies should have a CSR policy because of the attention they receive from stakeholders (“Do it right,” 2008). MNEs mainly disseminate their CSR actions through reports, websites and advertising campaigns (Asmussen & Fosfuri, 2019), with the potential aim of strengthening their brand personality (Huber et al., 2011).
MNEs operate in multiple locations, which differ in terms of their economic, political and CSR context, and therefore face the challenge of adapting to the different host countries (Munro, 2017). In addition, MNEs must ensure that CSR policies are enforced consistently in all subsidiaries, as a CSR breach in one subsidiary can destroy the social image of the entire MNE, even if all other subsidiaries have behaved responsibly (Asmussen & Fosfuri, 2019).
In the adoption process, MNEs must pay particular attention to the expectations of local societies, as there exists a sceptical view portraying MNEs as exploiters of resources (B.
I. Park & Ghauri, 2015). This is of particular importance, as MNEs need to gain and retain legitimacy to operate in the new environment (Kolk & van Tulder, 2010; Munro, 2017).
Furthermore, MNEs are confronted with the challenge of inconsistent and absent social and environmental international regulations (Kolk & van Tulder, 2010). Donaldson (1996) refers to this as a moral free space in which managers must shape their own policies, while paying attention to basic human values. By adopting CSR practices in these locations, MNEs act as a substitute for absent institutions (Rathert, 2016). However, the opposite is possible as well. It has been suggested that CSR can emerge in the presence of strong regulatory institutions and collective self-regulations by industries, as such an institutional environment enables stakeholders to demand enhanced social behaviour from the company (J.
L. Campbell, 2007; Gjølberg, 2010). However, this may not always be the case, as, for example, public pressure on CSR reporting is less in developing countries than in developed countries (Ali et al., 2017). Moreover, the quality of institutions varies from country to
country, and thus also the legal pressures on companies. Therefore, external pressures have an influence on whether the above described mechanism can actually occur.
2.1.3 CSR in Developed and Emerging Markets Categorisation and Definition
There are several approaches for classifying countries into developing, emerging and developed markets. For example, countries can be grouped on the basis of their gross domestic product per capita or on the basis of indicators such as the literacy rate (Niebel, 2018). Appendix A contains a list of one possible classification of countries based on gross domestic product per capita, as developed by Niebel (2018). Rostow (1960) proposes that societies can be grouped into five categories based on their economic development, namely
“[…] the traditional society, the preconditions for take-off, the take-off, the drive to maturity, and the age of high mass-consumption” (Rostow, 1960, p. 4). However, Tilt (2016) highlights that most of the existing literature only distinguishes between developed and developing countries. Accordingly, the developing countries, where the majority of today’s population lives, can be assigned to Rostow’s categories one to three (Tilt, 2016).
Work by scholars suggests that the terms developing and emerging are being used interchangeably, when referring to markets and countries (e.g. Momin & Parker, 2013; Ye &
Zhang, 2011). Yet it must be noted that there are differences between the two terms.
Developing countries can be defined as “[…] countries [that] are in the process of industrialisation and are often characterised by unstable governments, higher levels of unemployment, limited technological capacity, unequal distribution of income, unreliable water supplies and underutilised factors of production” (Tilt, 2016, p. 2). Emerging countries (or markets) represent a specific form of this situation, being “[…] countries undergoing fast- paced turbulent change as a result of economic liberalization, rapid industrialization, and increased integration into the global economy” (Marquis & Raynard, 2014, p. 4). In summary, emerging markets are a special form of developing markets, which are characterised by higher growth and are moving more rapidly towards developed markets.
Described in more detail in the methodology section, emerging and developed markets are analysed in this paper, based on the classification of Morgan Stanley Capital International (MSCI).
Differences in CSR
The extent to which CSR is either promoted or impeded in each country depends on the overall business environment, according to existing research (Li et al., 2010). Local conditions on national level can have an influence on the degree of support for CSR (Baughn et al., 2007). Aguilera et al. (2007) argue that companies are influenced by individual, organisational, national and transnational actors to engage in CSR activities, e.g. governments, shareholders, consumers and employees (see appendix B). In addition, different socio-cultural environments influence the way individuals think about the concept of sustainability and therefore have an effect on the understanding of social responsibility (Tilt, 2016).
While CSR has been extensively researched in developed countries, the available knowledge about CSR in developing countries is limited (Li et al., 2010). Based on the studies of Baughn et al. (2007) and Welford (2004), it can be said that CSR implementation is less evolved in developing countries. Appendices C and D illustrate this, showing country ratings for social and environmental CSR in relation to gross domestic product per capita. In addition to the positive relationship between per capita GDP and CSP, Baughn et al. (2007) could find a positive effect between economic and political freedom, low levels of corruption, and CSP.
It has also been found that Asian companies lag behind Western companies in terms of CSR (KPMG, 2005; Welford, 2004), with the exception of Japan, Taiwan and Singapore (Baughn et al., 2007). According to these exceptions, generalisations about particular regions and their CSR should be avoided, inter alia, because of the diversity in economic development, politics and culture (Kimber & Lipton, 2005). In this context it should be mentioned that the CSR approaches of Western companies may not be applicable in Asia, due to cultural differences (Edwing & Windisch, 2007), which may force companies to adopt to new CSR practices when entering this region.
The reasons and motivation of companies to engage in CSR, as well as CSR expectations, differ in developing and developed countries. According to Sharma (2019), socially responsible companies in developing countries focus on local and marginalised communities and strive to improve their standard of living by, for example, providing education, food and shelter and protecting the environment. In this context, the main problems that prevent companies from successfully carrying out CSR activities are time constraints, resource costs and institutional problems (Possenti, 2012). Cooperation between companies and NGOs can be a solution, as the reputation of the NGO confers trust and authority (Lee, 2018). In addition, NGOs can act as intermediaries and provide intimate knowledge (Sharma,
2019). Moreover, the greater CSR needs in emerging countries require interaction with a larger number of actors, compared to developed countries (Munro, 2017). Ali et al. (2017) further note that public pressure on CSR reporting in developing countries is relatively low and reporting is mainly influenced by external stakeholders, like foreign investors and international media.
According to Sharma (2019), reducing environmental pollution, both domestically and in the rest of the world, is an important issue in developed countries. Sharma (2019) found that CSR is also conducted through people-led campaigns targeting discriminated groups in society.
Overall, the public supports CSR campaigns that are also beneficial for businesses. (Sharma, 2019). Additionally, there is a legal framework in place to punish companies for irresponsible behaviour (Sharma, 2019). The concerns of stakeholders such as government, investors, environmental activists and the media are given increased consideration when reporting on CSR (Ali et al., 2017).
There are also similarities between developing and developed countries with regard to CSR. For example, in both countries social media is increasingly used to reach and interact with the target audience (Sharma, 2019). An overview of the similarities can be found in appendix E.
However, as explained above, the demands and perceptions of CSR in emerging and developed countries continue to differ widely, creating challenges for MNEs. Possible implications of these differences will be discussed further in the following hypotheses section.
2.2 Hypotheses Development 2.2.1 Geographical Diversification
Based on Freeman (2010), stakeholders are individuals or groups of a population that have a general interest or a stake in a company, and therefore its well-being. In addition, stakeholders are affected by the actions of a company and, in return, react to these actions, trying to influence the firm’s behaviour (Freeman, 2010). Stakeholders are, for example, customers, employees, shareholders and governments (McWilliams & Siegel, 2001). In this respect, the financial success of a company depends on its ability to apply a strategy that enables it to manage the relationships with its stakeholders effectively (Donaldson & Preston, 1995;
Mitchell et al., 1997).
Kang (2013) further emphasises the impact that companies have on the welfare of stakeholders, and that the range of business activities has a strong influence on the number
and diversity of these stakeholders. In this context, various researchers point out that an increased geographical area of activity leads to a greater number of stakeholders with which the company is confronted (e.g. Brammer et al., 2006; Kang, 2013). According to Kacperczyk (2009), CSP data is a suitable indicator to measure the attention companies pay to their stakeholders, with higher scores implying greater attention. This is due to the fact that the CSP measures the engagement of companies regarding social and legal issues (Kang, 2013).
In general, stakeholder expectations of environmental and social performance have increased significantly in recent years, but to a varying extent in different countries (Hoffman, 1999; Sharfman et al., 2004). This divergence is due to the different social issues stakeholders face, based on their country, region, and cultural or religious background (Kang, 2013). The different pressures to act in a socially and environmentally responsible manner are leading MNEs to adopt global standards to ensure that they meet these different expectations (Kostova & Zaheer, 1999; Sharfman et al., 2004).
In terms of stakeholder theory, geographically diversified firms face a greater number of stakeholders with both power and legitimacy than focused firms (Kang, 2013; Mitchell et al., 1997). This is due to the fact that stakeholder demands and social issues can vary greatly from country to country (United Nations Development Programme, 2003) and a geographically diversifying company must deal with these demands and issues accordingly (Kang, 2013).
Interestingly, Kang (2013) mentions that the relationship between diversification and CSP could also be negative if the company decides to ignore increasing stakeholder demands.
This could be the case if a company only meets the legal requirements in the country where it operates but does not address any additional demands from stakeholders. However, the literature provides several reasons that indicate a positive relationship.
First, diversification leads managers to pursue a risk-minimising strategy regarding the allocation of resources (Hayes & Abernathy, 1980). Risk reduction and safe decision making can be achieved by responding appropriately to stakeholder demands (Kang, 2013).
Secondly, diversification, including geographical diversification, reduces the employment risk of managers, as it reduces the risk of bankruptcy of the company and makes managers less interchangeable as their skills are needed and replacement is more expensive (Kang, 2013; Shleifer & Vishny, 1989). According to Kacperczyk (2009), reducing employment risks allows managers to focus on long-term rather than short-term results, leading to greater attention to all stakeholders rather than just shareholders.
In addition, the involvement of internationally diversified companies in CSR activities can create a positive brand image and benefit the entire organisation (Kang, 2013). An improved brand image in one country could thus also benefit operations in other host countries. The potential benefits of CSR, as explained above, could therefore be transferred across the organisation, which might be an incentive for MNEs to engage in CSR activities.
Based on these findings, it is assumed that an increase in geographical diversification of a company leads to an overall increase in CSP.
Hypothesis 1: Increased geographical diversification of MNEs leads to increased corporate social performance.
2.2.2 Developed and Emerging Market MNEs
Cai et al. (2016) found that firms from developed countries have significantly higher CSP ratings than firms from emerging markets, when comparing the median for both country groups. They argue for a strong link between a country's economic development and CSP rating, as factors such as per capita income and political stability, but also cultural orientation, influence the costs of conducting CSR activities and the potential benefits that can be achieved. Baughn et al. (2007) found similar results in their study.
Additionally, Keig et al. (2019) elaborate on the differences in organisational learning that occur due to the differences in stringency levels between countries, with regards to CSP.
In this way, MNEs entering a foreign environment with a stringency level below that of their home country can more easily make use of previously acquired knowledge. However, if the level of stringency in the host country is higher than in the home country, MNEs cannot use their previously acquired knowledge and have to develop new strategies and skills to cope with the new environment (Keig et al., 2019). In conjunction with the different average CSP ratings, this implies a higher learning curve for emerging market multinational enterprises (EMNEs), i.e. their organisational learning outcomes are relatively higher than for developed market multinational enterprises (DMNEs).
It has also been found that with increasing dependence on foreign sales, EMNEs place higher emphasis on corporate sustainability because they have to pay more attention to global stakeholders, that have higher concerns regarding sustainability compared to the local stakeholders of the EMNE (Kang, 2013; S. B. Park, 2017 as cited in S. B. Park, 2018). This is especially the case if the EMNE starts to operate in more developed countries.
Moreover, the study of Fiaschi et al. (2015) shows that firms from BRIC countries that are entering new markets with more stringent regulations engage in CSR reporting, in order to overcome the liability of foreignness (LOF) and liability of origin (LOO). LOF has been defined as “[…] costs associated with a foreign firm’s distance from the cognitive, normative, and regulatory domains of the local institutional environment” (Zaheer, 2002, p. 352). It can be understood as a disadvantage that a foreign company faces because it is not local (Madhavan & Gupta, 2014). LOF is something that all MNEs face (Madhavan & Gupta, 2017) and must cope with in order to do business successfully. In addition to the LOF, EMNEs can also be confronted with the LOO, which can be seen as a disadvantage a firm faces in a host country because of where it is from (Madhavan & Gupta, 2017). Because of that, EMNEs must take additional measures to demonstrate their CSR commitment, such as the increased reporting, as mentioned above.
Moreover, multiple scholars have found a positive relationship between the depth of a firms’ CSR reporting and its overall CSP (e.g. Graafland & Smid, 2016; Vurro & Perrini, 2011 as cited in Tashman et al., 2017). Tashman et al. (2017) find additional support for a positive relationship between EMNEs’ intensity of CSR reporting and CSP. Therefore, increased CSR reporting by EMNEs caused by LOF and LOO may also lead to an increase in CSP.
Hence, it is assumed that geographically diversifying EMNEs experience higher levels of CSR-related organisational learning, leading to a greater increase in CSP.
Hypothesis 2: Geographical diversification of MNEs from emerging markets leads to a higher increase in firm-level CSP than the geographical diversification of MNEs from developed markets.
2.2.3 Formal Institutional Distance
“Institutions are the rules of the game in a society […]” (North, 1990, p. 3), meaning that they provide the legal and regulatory framework of a country. According to institutional theory, companies rely on these formal environments to distinguish between appropriate and inappropriate behaviour (Scott, 1987) in order to gain legitimacy in the home and host markets (Keig et al., 2019). The institutional distance between two countries increases the difficulty of correctly understanding the institutional requirements (Kostova & Zaheer, 1999), i.e. understanding the requirements in the host country, which is different from the home
risk and uncertainty, thereby increasing the difficulty of operation and hence leading to an increased need for organisational learning (Xu & Shenkar, 2002 as cited in Keig et al., 2019).
Overall, formal institutional conditions vary from country to country (Kogut, 1991).
Kuncic (2012) carried out a cluster analysis based on multiple institutional indicators, from which three groups of formal institutions emerged: legal, political and economic. Based on this analysis, a World Institutional Quality Ranking was developed, in which countries are categorised on the basis of their performance in these three groups (Kuncic, 2012). The according map and data can be found in appendix F and G. It can be seen that the formal institutional quality can differ greatly from country to country.
It follows that the CSP standards of the MNE's host countries can be higher than those of the MNE's home country, and vice versa (Keig et al., 2019). Kostova & Zaheer (1999) point out that by operating in these different environments, MNEs gain organisational experience with regard to the expectations in and requirements of the different environments.
However, if the formal institutional stringency of the host country is higher compared to the home country, the MNE will need to acquire new skills with regard to CSR regulations (Strike et al., 2006).
However, it has been found that when confronted with externalities that are very different from previous experiences, this difference can prevent the MNE from assimilating and using the new knowledge (Keig et al., 2019; Lane et al., 2006). J. T. Campbell et al.
(2012) also argue that a greater formal institutional distance regarding the distance between home and host country of the MNE requires greater effort to understand the social regulations of the host country, which increases the cost of CSR.
Accordingly, the previously outlined possible relationship between geographical diversification and CSP could be negatively affected by a high formal institutional distance within the MNE's subsidiary portfolio.
Hypothesis 3: Higher formal institutional distance within an MNEs subsidiary portfolio negatively moderates the relationship between geographical diversification and CSP.
2.2.4 Informal Institutional Distance
Stakeholders' expectations of social responsibility often derive from informal social norms which are based on national culture, varying from country to country (Keig et al., 2019).
National culture can be defined as “[…] the assumptions, values, norms and beliefs shared by individuals in a society” (J. T. Campbell et al., 2012, p. 87). Matten & Moon (2008) note that
meeting the expectations of those stakeholders can help MNEs to overcome the LOF by gaining local legitimacy.
However, learning about the culture of the host country and developing a strategy for coping with cultural differences is time and resource intensive (J. T. Campbell et al., 2012; Wilkinson et al., 2008), especially when home and host country are very different (Lane et al., 2006).
Cultural distance, the “[…] differences in social norms, religions, languages and ethnicities”
(J. T. Campbell et al., 2012, p. 87), thus represents a challenge since every country has unique culture and customs (Keig et al., 2019). Moreover, a greater cultural distance can lead to cultural conflicts, regardless of the degree of geographical diversification of the MNE (Popli et al., 2016).
In particular, knowledge must be obtained about the differences in conducting business (Keig et al., 2019), that can be rooted in different cultural backgrounds. This cultural background also has an impact on ethics (Ho et al., 2012). It has been found that ethical attitudes, sensitivity, judgement and perception are influenced by culture (Ho et al., 2012), which could affect stakeholders' perceptions of CSR. MNEs therefore need to understand not only which CSR issues are relevant, but also how they should address these issues in the respective cultural environment. Failure to conduct these activities adequately could have a negative impact on the perception of the company's operations or even negatively affect CSP.
Additionally, opposed to formal institutions, knowledge about informal institutions acquired due to organisational learning cannot automatically be used for immediate application in another country (O’Grady & Lane, 1996). This is due to the different history and culture of the individual countries (Keig et al., 2019). This argument can be further supported by the fact that cultures can differ even within a country, which is reflected in sub- cultural regions that can differ widely in terms of political attitudes, religion and economic development (Dheer et al., 2015). Because of that, MNEs operating in culturally very different countries could be at disadvantage, compared to MNEs with lower variety in terms of cultural difference (J. T. Campbell et al., 2012).
Moreover, the more diverse stakeholders an MNEs faces, the more likely it is that the expectations of these stakeholders will be contradictory, will conflict with each other and therefore require more administrative resources to manage, which can adversely affect the MNE’s CSP (Barnett, 2007; Keig et al., 2019).
It can be said that operating in a geographically diverse environment always requires MNEs to engage in organisational learning, as knowledge about informal institutions is not easily transferable; but some cultures are even more diverse and require even more resource investment to be managed properly. Therefore, the relationship between geographical diversification and CSP is expected to be negatively moderated by increasing informal institutional distance between the home country of an MNE and the host countries of its subsidiaries.
Hypothesis 4: Higher informal institutional distance within an MNEs subsidiary portfolio negatively moderates the relationship between geographical diversification and CSP.
2.3 Conceptual Model
Based on the preceding literature review and hypothesis development, the following conceptual model can be derived.
Figure 2 – Conceptual Model
3. Research Methodology
This section explains the research methodology that is used to test the previously developed hypotheses. First, the data collection process is explained. Then the obtained sample is described in detail. Finally, the conducted data analysis is outlined.
3.1 Data collection
The relationship between geographical diversification and CSP, as well as possible moderating effects of formal and informal institutional distance, is tested using a quantitative approach. The corresponding secondary data is taken from databases available to students at the University of Groningen and from publicly accessible databases. The data for the dependent variable CSP is taken from the ESG database of Thomson Reuters Eikon. The database provides information at company level on three different pillars, namely environmental, social and governance. The overall CSP rating based on the three pillars is used. The data for the independent variable geographical diversification, measured as the number of countries where an MNE has subsidiaries, is taken from the database Orbis. The independent variable market of origin, which distinguishes between developed and emerging markets, is also obtained from Orbis data. The data for the moderating variables is from the World Bank and Beugelsdijk et al. (2016). For the moderating variables, a mathematical approach is used to calculate the weighted distance from the MNE's home country to its subsidiaries in different host countries. The Worldwide Governance Indicators (WGI), which measure the moderating variable formal institutional distance, consist of six different measurements at country level, whose average is used for the calculation. Beugelsdijk et al.
(2016) provide data for the moderating variable informal institutional distance. In their paper they outline an approach to combine data at country level from the Hofstede, Schwartz and GLOBE frameworks. The data for the control variables firm size and firm age are taken from the database Orbis. The control variables were selected on the basis of data availability, and because other studies on similar topics stressed the importance of testing them and also included them in their analysis (e.g. Brammer et al., 2006; J. T. Campbell et al., 2012; Kang, 2013; Keig et al., 2019).
3.2 Sample
The sample used consists of MNEs from both developed and emerging markets. The country in which the MNE's headquarter is located is defined as its home country. The classification of countries is based on the MSCI ACWI index, as shown in figure 3. The index replicates the performance of the majority of stocks from 49 countries and classifies them into 23 developed and 26 emerging markets (MSCI ACWI Index, n.d.).
For the regression a total of 7 variables is used. The rule of thumb indicates that at least 10 events per variable (EPV) should be selected, but it was found that problems can still occur at this size and at least 17 EPV should be used (Vittinghoff & McCulloch, 2007).
Following this approach, a minimum of 119 observations should be taken for the regression.
To obtain data for both developed and emerging markets, the five countries with the highest weighting within the MSCI World Index (United States, Japan, United Kingdom, France, Switzerland) and the MSCI Emerging Markets Index (China, Taiwan, South Korea, India, Brazil) are used. This is because listed companies are legally obliged to publish corresponding company data, which ensures the availability of the respective data. Moreover, large companies are more likely to be geographically diversified (Markides & Williamson, 1996, as cited in Kang, 2013). However, a list of the individual constituents of the MSCI indices is not publicly available. Hence, based on the approach of Kang (2013), the largest 100 companies per country are selected on the basis of market capitalisation. In order to reduce possible bias in the results, 15 companies from the previous selection are randomly taken using the random sampling function of the Orbis database. However, in order to ensure that the minimum number of observations is not undercut in the case of any missing information, 30 companies per country are selected at random in the first instance.
Figure 3 – Market classification (MSCI ACWI Index, n.d.)
To ensure both timeliness and availability of the data, the observations of the year 2018, respectively the end of 2018, were used for the analysis. Eventually, the data of 75 companies from developed markets and 75 companies of emerging markets is used for the analysis. In addition, the total number of 150 observations exceeds the minimum requirement of 119 observations, as explained above.
3.3 Variables
3.3.1 Dependent Variable
There are several methods and models to assess CSP at company level (Rahdari & Anvary Rostamy, 2015). For this thesis, the Thomson Reuters ESG database is used for three main reasons. First, the University of Groningen offers its students free access to the complete database. Secondly, the database is managed by over 150 research analysts from several countries, covering more than 400 ESG measures, and is updated every 14 days (Thomson Reuters ESG Scores, 2017). This ensures the most accurate and up-to-date data. Thirdly, the evaluation system of the database was rated by experts in a survey published in 2013 as one of the most credible (GlobeScan/SustainAbility, 2013). In addition, Dorfleitner et al. (2015) compared two of the most important providers of ESG ratings, namely ASSET4 from Thomson Reuters (the predecessor of Thomson Reuters ESG Database) and KLD from MSCI, as well as Bloomberg Sustainability, and conclude that “[…] ASSET4 provides a much higher database depth and therefore a more differentiated evaluation” (Dorfleitner et al., 2015, p.
459). Since the Thomson Reuters ESG Database is “[…] an enhancement and replacement to the existing ASSET4 ratings […]” (Thomson Reuters ESG Scores, 2017, p. 3), it is assumed that these findings are still valid.
The database evaluates companies according to 3 pillars: environment, social and governance. The first pillar environment includes resource consumption, emissions and innovation. The social pillar considers workforce, human rights, community and product responsibility. The third pillar governance includes management, shareholders and the CSR strategy. Each company receives a grade for each pillar and an overall grade. As this study is designed to include the environmental, social and governance performance of the companies, the overall grade is used.
Nonetheless, it should be noted that the corresponding CSR score calculated by the ESG database is only an approximation of the actual score. In particular, the values for the ESG controversy score are based on media coverage. However, it cannot be guaranteed that all incidents have been reported appropriately and it could therefore be that various irresponsible practices have been carried out without being represented in the company's CSR score. As a result, the CSR score could be above the actual CSP of the respective company.
3.3.2 Independent Variables Geographical Diversification
Following the approach of Brammer et al. (2006), the independent variable geographical diversification is measured by the total number of countries where the MNE has principal subsidiaries in. In other studies, measures such as the foreign sales ratio were used to determine the international diversification (e.g. Kang, 2013). However, some bias is attributed to this approach, as both the incidence and consistency of such reporting for each region within the company breakdowns are insufficient (Brammer et al., 2006).
Using the database Orbis, the number and country codes of the existing first-level subsidiaries are retrieved and then processed accordingly to arrive at a final value for the analysis. The final value is a natural number that represents the actual number of operating countries with subsidiaries.
Market of Origin
The independent variable market of origin is based on the market where the headquarter of the MNE is located. A dummy variable was created to investigate possible differences
