• No results found

The link between corporate governance regimes and human rights violations

N/A
N/A
Protected

Academic year: 2021

Share "The link between corporate governance regimes and human rights violations"

Copied!
50
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The Link between Corporate Governance Regimes and

Human Rights Violations

[1]

Master Thesis

Sarah Schoorl

10727620

MSc. in Business Administration - International Management University of Amsterdam

Supervisor: Dr. M.K. Westermann-Behaylo Second reader: Dr. Johan Lindeque

(2)

[1] Etter, D. (2010). India´s child coal miners. The Christian Science Monitor. Retrieved November 2014, from: http://www.csmonitor.com/World/Global-News/2010/0920/India-s-child-coal-miners

(3)

Statement of originality

This document is written by Student Sarah SCHOORL who declares to take full responsibility for the contents of this document.

I declare that the text and the work 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 Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(4)

Acknowledgements

This thesis is the final stage of my Master in Business Administration in the track International Management. I want to thank everybody who helped me to complete my Master thesis.

My first and sincere appreciation goes to Dr. Michelle Westermann, my thesis supervisor, for her continuous help and support in all stages of this thesis. I would furthermore like to thank my family, especially my parents and sister, for always showing great interest, and for their perpetual love and support for my decisions. Last but not least, I would like to thank my friends Giorgos Karafotias and Pierre Sutra for their recommendations that helped me to complete my thesis.

(5)

Abstract

The purpose of this study is to explore the relationship between corporate governance regimes and the likelihood of extractive multinational enterprises (MNEs) to violate human rights. More explicitly, this research investigates how the difference of corporate governance regime, in the home country, and in the country where firms are primarily listed, can be associated with the involvement of the multinationals in labor and economic development violations. The hypotheses are explored using secondary data on companies’ details and human rights violations, provided by the Orbis and the CHRD databases. To establish and test the link, this research is of quantitative nature and the hypotheses are tested empirically by running a chi-square analysis. The results support the proposition that multinationals headquartered in, or primarily listed in, countries following the Rhineland regime of corporate governance are less likely to violate human rights, than multinationals headquartered in, or primarily listed in, countries following the Anglo-Saxon regime. These findings have implications for practitioners, policymakers, and academics around the globe. The ultimate goal of this paper is to help improve human rights in business.

Keywords: • Anglo-Saxon model • Corporate governance • Human rights violations •

(6)

Table of Contents

1. INTRODUCTION... - 7 -

2. LITERATURE REVIEW ... - 9 -

2.1HUMAN RIGHTS ... -9-

2.2LABOR RIGHTS ... -10-

2.3THE RIGHT TO ECONOMIC DEVELOPMENT ... -11-

2.4BUSINESS AND HUMAN RIGHTS ... -12-

2.5DUE DILIGENCE ... -13-

2.6CORPORATE GOVERNANCE ... -15-

2.7CORPORATE GOVERNANCE REGIMES ... -16-

2.8STOCK EXCHANGE INFLUENCE ON COMPANIES ... -20-

2.9GAP AND RESEARCH QUESTION ... -20-

3. THEORETICAL FRAMEWORK ...- 22 -

3.1CORPORATE GOVERNANCE REGIMES AND ECONOMIC DEVELOPMENT RIGHTS VIOLATIONS ... -23-

3.2CORPORATE GOVERNANCE REGIMES AND LABOR RIGHTS VIOLATIONS ... -24-

3.3CONCEPTUAL MODEL ... -26- 4. RESEARCH METHOD ...- 27 - 4.1DATA SOURCES ... -27- 4.2SAMPLE SELECTION ... -28- 4.3VARIABLES ... -30- 4.4METHODOLOGY ... -32- 5. RESULTS...- 34 - 5.1HYPOTHESES TESTING ... -34- 5.1.1 Cross-tabulations ... - 34 -

5.1.2 Chi-Square test outputs ... - 35 -

5.1.3 Strength of association ... - 36 -

5.1.4 Direction of association ... - 37 -

5.2CONCLUSION OF THE RESULTS ... -38-

6. DISCUSSION ...- 39 -

6.1IMPLICATIONS OF THE STUDY ... -39-

6.2LIMITATIONS AND DIRECTIONS FOR FUTURE RESEARCH ... -40-

7. CONCLUSION ...- 42 -

BIBLIOGRAPHY ...- 43 -

APPENDIX ...- 49 -

(7)

Index of Tables and Figures

Table 1: Characteristics of the two corporate governance models……….19

Figure 1: Conceptual model………..…..…26

Table 2: Descriptive statistics for sample 1………..………30

Table 3: Descriptive statistics for sample 2……….……….…30

Table 4: Variables and sources ……….31

Table 5: Crosstab for sample 1………..34

Table 6: Crosstab for sample 2………..………35

Table 7: Chi-square test output for sample 1………..35

Table 8: Chi-square test output for sample 2……….……….35

Table 9: Strength of association for sample 1……….36

(8)

1. Introduction

Nelson Mandela proclaimed in his speech `Make Poverty History´, the 3rd of February 2005, “overcoming poverty is not a gesture of charity. It is an act of justice.” Yet, nearly half of all human beings alive today are living in severe poverty, which is the lack of secure access to basic necessities for humans’ survival, such as food, water, shelter and medical care. In order to survive poverty some 250 million children between 5 and 14 years old do wage work outside their household, often as soldiers, prostitutes, domestic servants, or in agriculture, construction, and textile or carpet production (Pogge, 2007).

The potential for multinational enterprises, defined as firms with value-added activities in at least two countries (Rugman & Verbeke, 2001), to positively impact human rights, including lifting millions out of poverty and accelerating economic growth in their host countries, cannot be overlooked. However, as shown by the regular reports on the Business and Human Rights Resource Center (BHRRC) of abuses by companies, it seems that the converse is also true. If unchecked, businesses can exacerbate poverty and perpetrate human rights violations (Kibugu, 2014). Human rights violations include discrimination, sexual harassment, rape, torture, poverty, as well as any action that would harm people´s health, safety, freedom of association, freedom of expression, privacy, and access to food, water, education and housing (BHRRC, 2014).

To better understand the relations between human rights abuses and business, academics have begun to analyze the underlying risk factors and outcomes of corporate human rights violations. Most of these studies focus on the legal or political perspective of business and human rights (Cragg, 2012; Wettstein, 2012). Yet, hardly any research has been done on the managerial perspective. Additionally, the link between corporate governance and human rights violations has not been studied before, and literature about business and human rights is mostly of qualitative nature, thus has not been tested empirically. To fill these gaps in research, this paper quantitatively analyzes the link between corporate governance regimes and multinationals´ human rights practices. More explicitly, the aim of this thesis is to find out how the difference of corporate governance regime in the home country, and in the country where firms are primarily listed, is related to the likelihood of corporate human rights violations.

Firms´ compliance with human rights varies according to the industry they are active in. For instance, the extractive industry is especially associated with high human rights violation rates (Drimmer, 2010), and particularly with labor and economic development rights abuses (Ruggie, 2007). Therefore, to make this study as interesting as possible, the selected multinationals are from the oil, gas and mining sectors, and the two fundamental human rights that will be studied

(9)

are labor and economic development rights. Please note that the aim of this thesis is not to attempt to find blame, but rather to fill a gap in research and thereby hopefully advance human rights in business.

To proceed, the next section of this study is a review of the existing literature on human rights and corporate governance. Thereafter a theoretical framework is developed to propose a link between corporate governance regimes and human rights violations. Section 4 presents the details about the research design, the data collection and the analysis method used in this paper. In section 5 the empirical findings of this research are reported, and section 6 discusses the implications and limitations of the study, and it provides potential future research subjects. Finally, section 7 presents the concluding remarks.

(10)

2. Literature review

The first part of this section reviews the literature on human rights and in particular on labor and economic development rights. Additionally, it explains how business is related to human rights, and it describes the due diligence process.

2.1 Human rights

The decades following World War II saw the development of an extensive body of international human rights law, which recaptured the morally appealing idea of adherence to shared standards of justice as a condition for full membership in international society (Donnelly, 1998). For instance, on the 10th of December 1948 the United Nations General Assembly adopted

the Universal Declaration of Human Rights (UDHR), which is the foundation of the international system of protection for human rights . The declaration proclaims that “all human beings are born free and equal in dignity and rights” (UDHR, 1948, Article 1), and that “everyone has the right to life, liberty and security of person” (UDHR, 1948, Article 3). Furthermore, during the cold war international human rights became as much a matter of internal as of external politics. Nowadays there is an infusion of positive universal human rights values into international politics, and a growing number of states seem to pursue sustained international human rights initiatives (Donnelly, 1998). Nevertheless, Donnelly (1998) argues that human rights represent a progressive late twentieth-century expression of the crucial idea that international legitimacy must rest in part on standards of just, humane or civilized behavior.

Not only states, but also non-governmental organizations, have been strong verbal advocates of human rights. The best known non-governmental organization in the field of human rights is Amnesty International, which works together with the United Nations (Thakur, 1994). Amnesty International (2013) defines human rights as “basic rights and freedoms that all people are entitled to regardless of nationality, sex, national or ethnic origin, race, religion, language, or other status.” Additionally, human rights include civil and political rights, such as the right to life, liberty and freedom of expression; as well as social, cultural and economic rights, including the right to participate in culture, to food, and to work and receive education (Amnesty International, 2013). According to legal scholars “human rights claim to what society is deemed required to do for the individual” (Kolben, 2009, p.6).

(11)

2.2 Labor rights

In contrast to human rights, which are possessed by all humans by virtue of their humanity, labor rights can be defined as “the set of rights that humans possess by virtue of their status as workers” (Kolben, 2009, p.6). Since these two elements have different definitions, the question whether labor rights are human rights needs to be answered. This question has attracted much interest in recent years among lawyers, academic scholars, trade unionists, and has given rise to heated debates. Some endorse the character of labor rights as human rights without hesitation, the ones following the positivistic approach, while others view that with skepticism (Mantouvalou, 2012).

Mantouvalou (2012) explains that one of the arguments the positivists give is that when looking at the UDHR several labor rights are human rights. For instance, the article 4 of the UDHR prohibits slavery and servitude; article 23 provides that everyone has the right to work and that everyone should work in a job freely chosen, that everyone should receive equal pay for equal work, that everyone should get decent remuneration to guarantee a dignified life for themselves and their family, and that everyone has a right to form and join trade unions; finally, article 24 guarantees a right to rest and leisure, including reasonable limitations of working hours, as well as holidays with pay. Mantouvalou (2012) states that another argument the positivists provide is that the International Labor Organization (ILO), the expert branch of the United Nations in the field of labor rights, predates all the human rights treaties and organizations (having been founded in 1919), which shows that labor issues became a matter of international concern before human rights. Moreover, the ILO developed in 1998 the Fundamental Declaration on Principles and Rights at Work, in which the ILO endorsed a list of labor rights as human rights (Kolben, 2009). The aim of this declaration was to promote opportunities for women and men to obtain decent and productive work, in conditions of freedom, equity, dignity and security (ILO, 2014). Finally, according to Mantouvalou (2012), a last argument positivists provide is that one of the key characteristics of human rights is that they are claims that prohibit grave moral wrongs. In this sense, labor rights are compelling, and qualify as claims that prohibit such moral wrongs. For instance, the right to privacy is a human right of central importance, which entails “a right to be left alone by the authorities in order to act and develop relations that each person finds meaningful, without external interference” (Mantouvalou, 2012, p.17). Similarly, protection of workers’ privacy against employer interference is an equally compelling claim. Positivists argue that it is a mistake to think that the right to privacy against state interference is more fundamental than the right to privacy against

(12)

employer interference. This study follows the positivistic approach and considers labor rights as being a human rights.

Ruggie (2008) states that core labor rights are the followings: the freedom of association, the right to work, to organize and participate in collective bargaining, to receive equal pay for equal work, to obtain just and favorable remuneration, to equality at work, to a safe work environment, to rest and leisure, and to family life. The aim of labor laws includes abolishing forced labor, and eliminating child labor and discrimination. According to the ILO (2014), these rights can be classified in four categories: freedom of collective bargaining and of association, abolition of forced labor, elimination of child labor, and freedom from discrimination. Any ILO member state is bound to `promote and to realize´ these fundamental rights, regardless whether it has ratified the related conventions (Kolben, 2009). These rights may be based on different foundations, such as freedom, dignity or capability (Mantouvalou, 2012).

Compa (1993) claims that differences in labor rights and standards among countries are growing in importance because they affect international trade and investment choices. For instance, from the investor's standpoint, countries with lower standards of labor rights become tempting targets for new investment aimed at cost-saving production systems. Therefore, the author argues that, strong, enforceable, and universal labor rights and standards are needed. Yet, international labor standards are hard to enforce because they impose considerable costs on multinationals, for instance costs related to minimum wage requirements, health standards and job security rules.

2.3 The right to economic development

The right to development was proclaimed by the United Nations in 1986 in the Declaration on the Right to Development. In Article 1, the Declaration defines such right as "an inalienable human right by virtue of which every human person and all peoples are entitled to participate in, contribute to, and enjoy economic, social, cultural and political development, in which all human rights and fundamental freedoms can be fully realized". The right to development includes: full sovereignty over natural resources, popular participation in development, self-determination, equality of opportunity, the creation of favorable conditions for the enjoyment of other civil, and political, economic, social and cultural rights (UNHR, 2014). For quite some time there was not an international recognition of the right to economic development. However, Sengupta (2004) argues that by 1993 there was an international consensus affirming the right to development as a human right, and that it can now be said to have general international recognition. Yet, there

(13)

remain major differences in the interpretation of the Declaration on the exact content of the right to development.

The Organization for Economic Co-operation and Development (OECD) Guidelines for Multinational Enterprises are recommendations addressed by governments to companies, operating in or from adhering countries. They provide non-binding principles and standards for responsible business conduct in a global context. The OECD (2011) states that enterprises should contribute to environmental, economic and social progress in order to achieve sustainable development in their host countries, and encourage local capacity building through close co-operation with the local community (i.e. business interests), as well as developing the enterprise’s activities in domestic and foreign markets. There are several ways in which multinational corporations can contribute to the development of host countries. For instance linkages created by multinationals with domestic firms may act as an incentive for the development of the local industrial sector (Markusen & Venables, 1999). Moreover, Blomström and Kokko (1998) provide three ways in which corporations positively affect host economies. First, they bring along new technologies, leading to efficiency benefits for local firms. Second, multinationals share their international networks with their domestic partner, which enables these firms to access foreign markets. Finally, domestic firms may feel threatened by the entry of multinationals in their industries, consequently they may want to strike back by becoming more efficient. This competition effect improves performance and subsequently fosters growth.

2.4 Business and human rights

Historically, human rights standards were only applicable to governments, not to the private sector. Firms argued that their only obligation was to respect national laws, even when those laws failed to meet international human rights standards (BHRRC, 2014). Yet, over time, the respect towards human rights became increasingly important, and in the 1990s the issue of business and human rights became permanently implanted on the global policy agenda. The main reasons were the worldwide expansion of the private sector and the growth in transnational economic activity. These evolutions increased social awareness of businesses’ impacts on human rights and attracted the attention of the United Nations (UNHRC, 2011). To follow this trend, some companies tended to approach social issues through their corporate social responsibility (CSR) programs. However, the issue with CSR initiatives is that they are mostly based on what companies voluntarily choose to address (BHRRC, 2014).

(14)

In July 2005, Kofi Annan, the Secretary-General of the United Nations from 1997 to 2006, selected Professor John G. Ruggie to be the Special Representative of the United Nations Secretary-General on business and human rights (BHRRC, 2014). Ruggie proposed in June 2008 a policy framework, the “Protect, Respect and Remedy” framework, for better managing business and human rights challenges (Ruggie, 2009). The framework was based on three complementary and interdependent pillars: the state duty to protect against human rights abuses; the corporate responsibility to respect human rights; and greater access by victims to effective remedies (Ruggie, 2008). The second pillar of the Guiding Principles means that companies should act with due diligence (UNHRC, 2011). The due diligence process is explained in the next subsection. The Human Rights Council welcomed the framework and extended Ruggie´s mandate by three years with the task of operationalizing the framework. To operationalize the framework, Ruggie developed guiding principles for each of its three pillars (Ruggie, 2009). In March 2011, Ruggie handed his final report which presents the “Guiding Principles on Business and Human Rights: Implementing the United Nations ‘Protect, Respect and Remedy’ Framework”, for consideration by the Human Rights Council. The goal of the Guiding Principles was to implement the "Protect, Respect and Remedy" policy framework. A few months later the United Nations human rights council (UNHRC) endorsed the guiding principles. The council stated that "The responsibility to respect human rights is a global standard of expected conduct for all business enterprises wherever they operate […] exists over and above compliance with national laws and regulations protecting human rights" (UNHRC, 2011).

2.5 Due diligence

Due diligence is a process which requires companies to ask tough questions about the risks of major transactions, projects, and ongoing operations (Sherman & Lehr, 2010). According to the UNHRC (2011) the process enables firms to not only ensure compliance with national laws, but also to avoid the risk of human rights harm. In other words, due diligence represents the essential steps a company must take to become aware of, prevent, and address adverse human rights impacts. Moreover, the UNHRC (2011) argues that basic human rights due diligence process should include policies, impact assessments, integration and performance tracking. Companies need to adopt a human rights policy because an aspirational language which describes respect for human rights is not enough. Companies need a detailed guidance in specific functional areas to assess actual or potential adverse human rights impact either through their own activities, or as a result of their business relationships. The purpose is to understand the specific impact on specific people, given a specific context of operations. In order

(15)

to prevent adverse human rights impact, business enterprises should integrate the findings of their assessments across relevant internal functions and processes, and take appropriate action. Additionally, business enterprises should track the effectiveness of their response in order to know if the firm´s human rights policies are being implemented optimally, whether it has responded effectively to the identified human rights impacts, and to strive for continuous improvement. Lastly, companies should be prepared to communicate their results externally, and they should provide for, or cooperate in, their remediation through legitimate processes, if they identified that they have contributed to adverse impacts.

Sherman and Lehr (2010) point out that the answers to the questions that due diligence requires may reveal unwelcome facts, requiring the company to take action to avoid previously unappreciated risks. Consequently, the authors raise the following question: rather than reduce risks for companies, could human rights due diligence actually increase a company’s risk of liability? Their answer meets Ruggie´s standpoint: not conducting due diligence is too risky, for both business and society. Yet, the effectiveness of due diligence principal depends more on the moral commitment of corporations rather than on an obligation (Fasterling & Demuijnck, 2013). The reason is that the implementation of the United Nations Framework is justified by the foundation of self-interest, referring to the gap between the principles and the extent to which corporations are obligated to act due diligence (Cragg et al., 2012). Since, due diligence may reveal unwelcome facts (Sherman & Lehr, 2010) and depends on the moral commitment of corporations, firms strategic implementation of the guidelines is often very limited (Fasterling and Demuijnck, 2013).

To overcome the risk of corporate human rights violations, companies should implement standards of best practices and statutory rules regarding human rights. Best practices and statutory rules are the two broad regulatory mechanisms in corporate governance (Aguilera & Haxhi, 2012). The first mechanism, corporate codes of conduct, is considered to be a crucial instrument for addressing child labor (Spar, 1998). To better understand how corporate governance is related to human rights practices, it is necessary to review the literature on corporate governance. Therefore, the second part of this section provides a review of the literature on corporate governance and on the different corporate governance regimes. Furthermore, it explains how home country and stock exchange influence companies.

(16)

2.6 Corporate governance

The literature on corporate governance, which is a relatively new research topic, emerged in the United Kingdom and in the United States. Originally, the literature on corporate governance was focused on how to improve ‘shareholder value’ through raising share price (Koen & Mason, 2005). Shareholders are individuals that own shares in a business or company through the right of control or rights to participate in profit. The shares are regulated by the stock markets and diversification of portfolios (Schrader, 1996). Nowadays, there are many conceptualizations of corporate governance within and across disciplines, including economics and management, law, political science, and sociology, which creates significant challenges when taking studies of corporate governance to the comparative level. Corporate governance can broadly be defined as the study of power and influence over decision making within the corporation (Aguilera & Jackson, 2010). In other words, corporate governance refers to “the framework of rules, relationships, systems and processes within and by which authority is exercised and controlled in corporations” (Council, 2007, p. 3). As mentioned in the previous subsection, there are two broad regulatory mechanisms in corporate governance. First, regulation can take the form of statutory rules, for instance hard-law. Second, regulation can constitute standards of best practices, such as codes. Codes can be issued either by regulators, governments, directors’ associations, managers’ associations, professional associations, or stock exchanges (Aguilera & Haxhi, 2012). Codes aim to address gaps that exist because of lack of legislation through practices that govern the composition of the board, the relationship with shareholders, managers and other stakeholders, remuneration and control mechanisms (Fombrun, 2005). They can be distinguished from other forms of regulation in that they are formally non-binding and flexible in their application (Aguilera & Haxhi, 2012).

National systems of corporate governance are influenced by countries’ formal institutions, including political rules, judicial decisions, and economic contracts; and informal institutions, consisting of socially sanctioned norms of behavior, which are embedded in culture (Peng, 2002). For instance, the quality of corporate governance depends on the effective interaction and negotiation among concerned parties, which is highly influenced by culture (Rafiee & Sarabdeen, 2012). There exist two major systems of corporate governance: the Rhineland model and the Anglo-Saxon model (Koen & Mason, 2005). Both models will be elaborated in subsection 2.7.

National institutional domains do not operate independently but rather together, and the combined characteristics influence how firms will have to make choices about their governance.

(17)

Thus, macro level factors of a country impact the firms within that country (Aguilera & Haxhi, 2012). For instance, the capital institutional domain, which refers to the nature of the financial market and the system of property rights within a country, defines the nature of financial investments in a firm. Additionally, the management-labor institutional domain plays a vital role in the way a firm is internally run (Aguilera & Jackson, 2010). Consequently, national corporate governance regime influences firms´ corporate governance systems (Aguilera & Haxhi, 2012). Thus, according to their home countries´ corporate governance models, some firms follow the Anglo-Saxon model and other firms follow the Rhineland model of corporate governance.

Nachum (2001) argues that most multinationals maintain their headquarters in their home countries. The author mentions that there are some exceptions to this generalization, but that such examples are still rare because most attempts to move the headquarters to foreign countries have been unsuccessful. Nachum (2001) explains that further reasons are that most important managerial decisions regarding the strategic direction of the firm are taken in the headquarters, and that it is the spot where both the top management and the board of directors are based. Therefore, this study considers that a multinational´s home country refers to where its headquarters is located.

2.7 Corporate Governance Regimes

As famously supported by Friedman (1970), who supported the shareholder theory of corporate governance, the servicing of the shareholders’ interests in a company is the ultimate purpose. He believes it is not in a company’s best interest to invest in social activities as it may impede value maximization and be costly enough to make firms less financially successful. Friedman (1970) suggests that a firm should only abide the law and not take further social initiatives. Opposing to this is the view of Freeman (1984), who promoted the stakeholder theory whereby paying attention to stakeholders (attending to the interests of employees, local communities, government, consumers, etc.) leads to improved stakeholder relationships and consequently higher corporate financial performance. The better a firm manages its relationships, the more it gains a competitive advantage by minimizing contract costs. Scholars argue that the theories promoted by Friedman and Freeman have separated the world into two dichotomous models: the shareholder versus the stakeholder approach (Rönnegard & Smith, 2013). Moreover, the classification of countries according to their corporate governance model can be divided into these categories: the Anglo-Saxon tradition and the German tradition (García-Castro et al., 2008).

(18)

The Anglo-Saxon model sees corporate governance from a ‘shareholders perspective’, and the Rhineland (German) model sees corporate governance from a ´stakeholder´s perspective´ (Koen & Mason, 2005). Stakeholders can be described as parties who interact with the firm on a regular basis, such as the shareholders, employees, customers and suppliers. These are called ‘primary stakeholders’. Other stakeholders, who do not engage in direct economic exchange with the business, but are affected by its actions, are called the `secondary stakeholders´. They include the nearby residents and the natural environment, business support groups, and local and national communities (Clarkson, 1995). Differences between the two models have implications on the way companies are directed. For instance, in the shareholder model, or the `market-based model´, the financial needs of firms are fulfilled through the stock market, also called the capital market and stock exchange. In this model, corporate control is left to the markets. On the contrary, in countries which follow the bank-based, or the Rhineland model, capital markets play a less important role, and firms typically turn to banks rather than stock markets for finance (Koen & Mason, 2005). The two models will be further described in the following paragraphs.

The Anglo-Saxon model is characterized by strong shareholder rights, and in this system firms’ main goal is to increase value for their shareholders (Koen & Mason, 2005). Minimal government involvement and the residual welfare state are also characteristics of these economies, known as free market economies. Moreover, in the Anglo-Saxon model business relationships are short-term oriented, institutions are centered on the market and focused on competition, and top managers tend to be monitored by means of market-based rewards and penalties (Bebchuk & Roe, 1999). Finally, in this model, governance structures, such as the board of directors, union representation on the board of directors, and the legal superstructure, are used as mechanisms to ensure fulfilment of implicit and explicit contracts, all aimed at better serving the interests of shareholders (García-Castro et al., 2008). The shareholder model is dominant in the Anglo-Saxon cluster of countries, including the United States, Great Britain, Ireland, Australia and Canada (Thelen, 2004). This model is also dominant in Switzerland, Portugal, Sweden, Finland, Czech Republic, Argentina, Brazil, Colombia, Nigeria, Peru, Mexico, Venezuela, Chile, and Singapore (Rossouw, 2008).

The Rhineland model sees the corporation as a set of relationships between multiple stakeholders with an interest in the firm, and therefore a broader set of goals to be satisfied (Aguilera & Jackson, 2010). In the Rhineland model, it is considered that without investment from its stakeholders the firm could not function and survive (Clarkson 1995). Furthermore, the

(19)

focus on long-term relationships and institutional arrangements to limit the negative risks of market cycles, are central in this corporate governance regime. The stakeholder model originates from the economic model of society that was common in Germany (Koen & Mason, 2005). Nowadays, this corporate governance model is not only dominant in Germany but also in Austria, France, Italy, Belgium (Aiginger & Guger, 2006), Japan (Thelen, 2004), Denmark, The Netherlands, Spain, Lithuania, Poland, Romania, Slovakia, Hungary, India, China, Russia, South Africa, Ghana, Malawi, Mauritius, Kenya, Tanzania, Uganda, Zambia, Zimbabwe, and Turkey (Rossouw, 2008).

Table 1 provides the key characteristics of each corporate governance regime, as well as a list of the countries which follow these models.

(20)

Table 1 Characteristics of the two corporate governance models

MNEs attributes Anglo-Saxon model Rhineland model

Underlying theory Shareholder theory Stakeholder theory

Government involvement Minimal Moderate to strong

Firm aspects

 Goal: Maximizing shareholder return  Main Believe: Firms

distinguish the economic from the ethical

consequences and values  Financial needs fulfilled

through stock exchange  Short-term investments goals  Board of directors structure: single-tier  Goal: Sustainable stakeholder satisfaction  Main believe : Ethics and

economics cannot be separated

 Financial needs fulfilled through banks  Long-term investments goals  Board of directors structure: two-tier Labor aspects

 Short- term relationships  Employees try to control firms’ decisions externally (i.e. strikes)

 Performance-related pay schemes

Weak labor unions

 Long- term relationships  Ongoing co-operation  Basic rate schemes  Strong labor unions

Countries

United States, Great Britain, Ireland, Australia, Canada, Switzerland, Portugal, Sweden,

Finland, Czech Republic, Argentina, Brazil, Colombia, Nigeria, Peru, Mexico, Venezuela,

Chile, and Singapore

Germany, Austria, France, Italy, Belgium, Japan, Denmark, The Netherlands, Spain, Lithuania, Poland, Romania, Slovakia, Hungary, India, China, Russia, South Africa, Ghana, Malawi, Mauritius, Kenya, Tanzania, Uganda, Zambia, Zimbabwe, and

Turkey

Source: Acas, 2012; Aguilera & Jackson, 2010; Aiginger & Guger, 2006; Bebchuk & Roe, 1999; Brown et

al., 1998; Freeman, 1984; Freeman et al., 2004; Friedman, 1970; García-Castro et al., 2008; Gospel & Pendleton, 2006; Koen & Mason, 2005; Rossouw, 2008; and Thelen, 2004

(21)

2.8 Stock exchange influence on companies

Multinationals can be publicly traded or privately owned. A private company is a company whose stocks are not offered to the general public, on public exchanges, but rather owned and traded privately. Contrastingly, a publicly traded company issues stocks which are traded on a stock exchange. Shareholders constitute the owners of publicly listed firms and they have the final say in most decisions taken (Davis et al., 2007). Listed companies are subject to unique pressures to return value to shareholders which do not apply to private companies (Conway et al., 2008). Moreover, these firms have their stocks listed on a primary stock exchange and, if some requirements are met (i.e. company size and share liquidity), they can be listed on other exchanges. The primary stock exchange is the major stock exchange where a share is bought and sold, it is the main owner of publicly traded firms, and it is the stock exchange which most influences publicly traded firms (Gwartney et al., 2008). Therefore, this study only takes companies´ primary stock exchange into account.

According to Aguilera and Jackson (2010) stock markets corporate governance regimes are influenced by the national corporate governance regime of the country where the stock markets are based. Additionally, a country´s financial market, or stock market, influences how firms traded on that stock market will make choices about their governance. Consequently, the national corporate governance regime of the country where the firm is primarily listed strongly affects the firm´s corporate governance system. Put differently, companies listed in a country which follows the Anglo-Saxon model are more likely to follow the shareholder approach, and firms traded on a stock exchange in a country which follows the Rhineland model are more likely to follow the stakeholder approach.

2.9 Gap and research question

As mentioned earlier in this paper, firm´s corporate governance influences the human rights practices within that firm (Spar, 1998). For instance, codes of conduct, which is one of the regulatory mechanisms in corporate governance, can improve the working conditions for the employees (Egels-Zandén, 2013; Ruggie, 2007; Zadek, 2004). However, the Business and Human Rights Resource Center shows that many multinationals do not respect human rights, although most of them have codes of conduct. Therefore, it is relevant to question what factor really makes the difference between firms that perpetrate human rights violations and firms that do not. This paper argues that the answer lies in the way companies are run, and more explicitly it depends on which corporate governance regime they follow.

(22)

Thus, this study aims to answer the following research question: How do corporate governance regimes influence the likelihood of multinational enterprises to perpetrate human rights violations? This paper fills two gaps in research. The first gap is that the link between corporate governance and human rights violations has not been studied before. The second gap is that most of the literature about business and human rights is qualitative and has not been tested empirically. This study is of quantitative nature, and the hypotheses are tested empirically by running a chi-square analysis. The following section draws, based on the existing literature, the link between corporate governance regimes and human rights violations.

(23)

3. Theoretical framework

There is a multitude of human rights problems that are intrinsic to oil, gas and mining corporations (Kaeb, 2007). Corruption, for instance, is a central aspect of the poor governance and management record of the oil sector. In many oil states, the oil portfolio is under the direct jurisdiction of the president, which in turn worsens income distribution and seizes resources away from developmental and social welfare programs. Moreover, in the oil and gas industry, multinationals typically operate through local elites and are reluctant to engage the panoply of local stakeholders directly. Companies often only pay minimum service to local communities, offer irregular payments for the use of tribal and other lands, use systems of compensation that are unsystematic, and community development is minimal (Alagoa, 2001; ClearWater, 2003; Kemedi, 2002). For instance, employment opportunities have been few and development projects minimal and typically incomplete (i.e. paramedical facilities without staff or medicines) (Watts, 2005). Besides, companies in the mining sector have also frequently been accused of complicity in human rights abuses, for instance against local communities committed by private or public security forces (Kaeb, 2007). Finally, as Banerjee (2008, p.14) states, “Royalties, promises of jobs, putting one community against another are some strategies that have proved useful for mining companies.”

The oil and gas sectors typically employ both local workers and contract foreign workers, often from their home country. On the oil fields, where expatriate workers and host-country nationals work and live, something like apartheid has appeared (Watts, 2005). Simultaneously, according to Deshingkar (2009), mines employ a strong percentage of informal workers, specifically in the rural sector, where smaller, privately-owned mines are located. In practice, informal workers are not properly protected by law, for instance with regard to minimum wages, equal wages for men and women, laws covering contract labor and migrant labor. Another recurrent practice in the mining sector is the racially discriminatory hiring process (Kaeb, 2007). Finally, Fallon and Tzannatos (1998) state that mining companies are often responsible for child labor, and not less than five percent of child laborers are employed in this sector or in the export manufacturing sector. The ILO (2001) considers child labor in mines one of the worst forms of child labor.

To sum up, multinationals from the extractive industry, especially in the oil, gas and mining sectors, are regularly accused of all sorts of economic development and labor rights violations (Ruggie, 2007; Drimmer, 2010). It is therefore interesting to analyze how corporate governance regimes affect extractive multinationals´ practices, regarding these two human rights.

(24)

3.1 Corporate governance regimes and economic development rights violations

Multinationals can have a positive spillover effect in their host country. For instance, a multinational from a home country with higher human rights standards can positively influence its host country (Cragg et al., 2012). Moreover, Fort and Schipani (2003) state that multinationals can contribute to sustainable peace. They can do this by fostering economic development, particularly for the marginalized, adopting external evaluation principles, such as transparency, nourishing a sense of community both within the company and in the areas in which the company is located, and mediating potentially conflicting parties and redirecting those parties toward a common goal, even if that goal is only that of profitability. The authors (Fort & Schipani, 2003) argue that there are a number of ways in which multinationals contribute to more stable societies. The first contribution they make toward stability is providing jobs to residents of the country. The second corporate contribution results from the benefits they can provide to the local population. For example, multinationals can pay for training and educational programs for abused youth or for recreational programs for children. The third contribution they make is the transfer of technological development or development of human managerial capabilities. For instance, by building a manufacturing plant multinationals provide technological know-how to the locals. A fourth contribution is the simple task of paying taxes.

This thesis argues that the companies which are most likely to try to contribute to sustainable peace by fostering economic development in their host countries are the ones that have a stakeholder perspective of the firm. The main reason is that in the stakeholder theory there is the assumption that values are necessarily a part of doing business, thus ethics and economics cannot be separated (Freeman et al., 2004). According to this theory, managers are agents of all stakeholders and have two responsibilities: to ensure that the ethical rights of no stakeholder are violated, and to balance the legitimate interests of the stakeholders when making decisions (Smith, 2003). On the contrary, many proponents of a shareholder view of the firm distinguish the economic from the ethical consequences and values (Freeman et al., 2004). Consequently, this paper also claims that multinationals with a shareholder view of the firm will less likely focus on contributing to sustainable peace in their host countries.

The capital institutional domain, which refers to the nature of the stock exchange within a country, defines the nature of financial investments in a firm (Aguilera & Jackson, 2010). Consequently, companies from one country will make other financial investments than firms from another country. For instance, in countries with the Rhineland regime of corporate governance, companies work as a community and focus on stakeholder value with a long term

(25)

perspective in mind (Koen & Mason, 2005). Thus, it is most likely that these firms will make long term investments in the host countries where they operate, for instance by developing the local economy. On the opposite, firms in countries which follow the Anglo-Saxon model focus mostly on short term value maximization for shareholders (Bebchuk & Roe, 1999). Consequently, it is more likely that these firms will make short term investments to increase profit for shareholders, rather than doing long term investments in order to foster host country economic development. It is therefore argued in this thesis that multinationals with a stakeholder perspective of the firm are more likely to respect economic development rights than firms with a shareholder perspective.

3.2 Corporate governance regimes and labor rights violations

The corporate governance literature, which is mainly written by American researches, largely neglects employees (Parkinson & Kelly, 2001). This neglect reflects weak employee participation in the United States relative to countries like Germany or Japan, where labor participation is politically important (Brown et al., 1998). In the stakeholder theory, followed by countries such as Germany and Japan, the continuity of labor relations is seen as the most important pillar and firms and employees focus on an ongoing co-operation (Koen & Mason, 2005). On the contrary, co-operation and trust between management and labor is often fragile in countries which follow the Anglo-Saxon regime of corporate governance. García-Castro et al. (2008) argue that in the United States, for instance, most firms would opt to cut back on labor in order to sustain current profitability. Under this system, employees often find it difficult to trust top management, as their behavior is subject to constant market scrutiny. Moreover, in this system a main believe is that shareholders are the only bearers of risks, and thus employees’ interests are treated only as an exogenous parameter. Finally, in countries with the Anglo-Saxon system, employees seek to control firms’ decisions externally with the threat of collective action (i.e. strikes) (Aguilera & Jackson, 2003). Consequently, it seems that employees from companies which follow the stakeholder model more easily protect their rights than employees from firms which follow the shareholder system.

In Germany firms with more than 500 employees are required to utilize a two-tier board structure, with a supervisory board providing general corporate strategy and a management board providing more of a day-to-day management oversight function. The two-tiered board structure is an explicit representation of stakeholder interests other than interests of shareholders. In firms with the stakeholder perspective no major decisions are made without the input of employee representatives (García-Castro et al., 2008). Representation rights can be

(26)

established by statutory law or direct state intervention, as in Germany. Representation rights can also be established contractually through collective bargaining, such as in Japan, where extensive joint consultation practices are written into collective agreements (Aguilera & Jackson, 2003). On the contrary, in the United States, the board structure is single-tier and labor has weak rights of representation (Fort & Schipani, 2003). Therefore it is reasonable to assume that firms with the stakeholder view are more likely to respect representation rights, such as the right to organize and participate in collective bargaining, than firms with the shareholder view.

In countries which follow the Anglo-Saxon regime corporate performance is measured in terms of share price and most firms have performance-related pay schemes to increase income for shareholders (Koen & Mason, 2005). Several writers claim that performance-related pay scheme is unethical for two reasons (Ryan, 2013). First, such a system tends to place increased business risk onto employees. Second, it undermines collective bargaining systems, and reduces the power of unions. Contrastingly, in Rhineland countries, firms use a combination of systems. For instance, they use the basic rate scheme, which is job-based (i.e. the rate for the job). The basic rate scheme offers much more stability to the employees, because pay only varies when moving up a scale, skill development, promotion to another grade, or a general uprating of pay levels (Acas, 2012). It is therefore possible to claim that firms with an Anglo-Saxon corporate governance regime probably violate the right to a safe work environment and the right to participate in collective bargaining more easily than companies with the Rhineland model.

As mentioned earlier in this paper, in the Anglo-Saxon model financial needs of firms are fulfilled through stock exchanges and corporate control is mainly left to financial markets (Koen & Mason, 2005). Consequently, in that corporate governance system, stock exchanges are very powerful, and they can easily put pressure on firms in order to obtain better conditions for shareholders, such as higher dividends. This, in turn, could take away money that would otherwise have been used to answer other stakeholders´ needs, such as a more secure work environment for employees. Contrastingly, in the Rhineland model, financial markets play a less important role and firms typically turn to banks rather than stock markets for finance (Koen & Mason, 2005). Therefore, in this model, stock exchanges are probably not as powerful as in the Anglo-Saxon regime. Additionally, in the Rhineland system, companies typically try to respect all the stakeholders´ rights (Smith, 2003). Taking this into account, it is reasonable to assume that multinationals publicly traded in Rhineland countries are more likely to take into account all the stakeholders´ needs and rights than multinationals publicly listed in countries which follow the Anglo-Saxon system, which are likely to focus only on the shareholders´ interests.

(27)

3.3 Conceptual model

As explained in the literature review, a multinational´s corporate governance system is both influenced by its home country and its stock exchange (provided the company is publicly traded) corporate governance regime. The links between home country corporate governance regime, stock exchange corporate governance system, and the likelihood of corporate human rights violations, lead to the following hypotheses and conceptual model (Figure 1).

Hypothesis 1 (H1): Multinationals whose home countries follow the Rhineland model are less

likely to violate labor and economic development rights than multinationals whose home countries follow the Anglo-Saxon regime of corporate governance.

Hypothesis 2 (H2): Multinationals publicly traded in countries that follow the Rhineland system

are less likely to violate labor and economic development rights than multinationals primarily listed in countries that follow the Anglo-Saxon model of corporate governance.

Figure 1 Conceptual model

In order to find out if the above hypotheses reflect the reality, they will both be empirically tested in section 5. In the section which follows (section 4) the research methods used to test the hypotheses are described.

•H1

Home Country Corporate Governance Regime

•H2

Stock Exchange Corporate Governance Regime Labor and Economic Development Rights Violations

(28)

4. Research method

The purpose of this study is to explore the relationship between corporate governance regimes and the likelihood of extractive multinational enterprises to violate human rights. I introduced two hypotheses that explain this correlation (see subsection 3.3). This section outlines the methodology and research methods used to test the hypotheses. It describes in detail how the data was collected and analyzed in order to investigate the working propositions.

The data used in this study are analyzed quantitatively and they are collected through secondary sources. There are several limitations to quantitative research. For instance, (1) it is less flexible, (2) it allows for less individual input than qualitative research, (3) it often misses contextual details, (4) it is vulnerable to statistical error, and (5) it is carried out in an artificial environment so that a level of control can be applied to the exercise. However, the quantitative approach also has strong benefits. The benefits include the followings: (1) it allows for a broader study (i.e. a greater number of subjects), (2) a greater objectivity and accuracy of results, and (3) the data collection procedure is less dependent on individual respondents so that personal bias can be avoided (Labaree, 2013).

4.1 Data sources

To test the above hypotheses, I used cross-sectional data on companies’ details and human rights violations, provided by the Orbis and the CHRD databases. The CHRD database uses the data available from the Business & Human Rights Resource Centre (BHRRC) to document allegations of human rights abuses by companies around the world from 2000 to the present. According to Olsen (2014) the BHRRC is the most comprehensive online archive. The information on the website is systematically organized using the online tool Qualtrics. Furthermore, the unit of analysis in the CHRD database is a Company Abuse Allegation (CAA). A CAA is an instance in which some group and/or individual accuses a company of a human rights violation. There are five main ‘violation types´: Labor, environment, health, abuse and economic development. This study focuses on labor and economic development CAAs. It is important to note that the CHRD database does not yet cover all the corporate human rights allegations available on the BHRRC website, thus it still has missing data. In order to contribute to the building of this dataset I coded a portion of the data. My contribution to the CHRD database is further explained in subsection 6.1.

(29)

The complement secondary data are derived from the Orbis database. Orbis contains comprehensive and up to date information on companies worldwide, on both listed and unlisted companies. Moreover, it includes information on over 65,000 listed companies and on 120 million private companies (Orbis, 2014). The database is used to research individual companies, search for companies by profile and to analyze companies. For this study Orbis is used to collect data on extractive firms active in the countries of my population (which is described in the next subsection). In more details, the data collected is on these firms´ headquarters´ locations (sample 1), and the location of the main stock exchange where they are listed (sample 2). To find information about multinationals active in the population, I selected all the companies shown by Orbis to own at least one subsidiary in at least one of the six countries of the population. Furthermore, as explained earlier in the paper, the home country is where the multinational headquarters is located (Nachum, 2001). Orbis provides the city and the country where multinationals are headquartered. Lastly, the primary (or main) stock exchange is the exchange which most influences publicly traded firms (Gwartney et al., 2008). Therefore, this study only focuses on primary stock exchanges. In the Orbis list of stock exchanges, the first one listed is the primary exchange on which multinationals shares are listed and the following names are other exchanges where the shares can be traded (called secondary stock exchanges).

4.2 Sample selection

The population that this research explores is the following: all the oil, gas and mining multinationals active in Peru, Colombia, Mozambique, Nigeria, China and India. These countries were selected because this study tries to cover the developing world. The developing world is typically mirrored by Africa, Asia and South America. I selected Mozambique and Nigeria to represent Africa, India and China to represent Asia, and finally Colombia and Peru to represent South America. This variety in geographic location increases the representativeness of the sample and the generalizability of the results (Eisenhardt & Graebner, 2007). These countries were also chosen because out of all of the countries in the developing world, these six countries are the ones which are best represented in the Orbis and the CHRD databases. Additionally, they are the host of a considerable number of extractive multinationals. The population is composed of 679 firms in total. The time range of the data is 2000 to the present, because the CHRD database contains data on company human rights violations in that specific period.

The selected firms for the samples are all multinational enterprises. These companies operate in at least two countries and are managed from one home country (Dunning & Lundan, 2008). The multinationals vary from medium sized companies, to large and very large

(30)

companies. However, both sample 1 (84 %) and sample 2 (89 %) are mainly composed of very large companies. Appendix 1 provides the criteria related to each size category.

To test the first hypothesis, I selected all the public and private multinationals whose home countries clearly follow either the Anglo-Saxon or the Rhineland model of corporate governance (these countries are listed in Table 1). The sample to test hypothesis 1 (sample 1) contains 197 firms (N = 197) of which 125 are headquartered in countries which follow the Anglo-Saxon regime and 72 are headquartered in countries following the Rhineland regime. This sample is summarized in the descriptive statistics provided in Table 2. To test the second hypothesis, only publicly traded multinationals were selected, because privately owned multinationals are not affected by any stock exchange governance (Davis et al., 2007). The selected firms for hypothesis 2 are all primarily listed in countries which clearly follow either the stakeholder or the shareholder regime of corporate governance. The sample to test hypothesis 2 (sample 2) contains 173 (N = 173) companies of which 116 are primarily traded in countries which follow the Anglo-Saxon regime and 57 are listed in countries which follow the Rhineland system (these countries are listed in Table 1). The descriptive statistics in Table 3 provide a summary about sample 2.

The samples contain firms that were accused of violations and firms that were not. In sample 1 (Table 2) 48 firms violated human rights and 149 firms did not and, in sample 2 (Table 3) 43 companies perpetrated human rights abuses and 130 did not. The firms that were accused of human rights violations were accused of either labor or economic development rights abuses. Note that some multinationals perpetrated a human right violation in one of the selected host countries but did not in another host country from the population. In this situation I coded the firm as being a violator. Finally, it is also important to note that a few companies are present in both samples. Yet the two samples contain distinct information about the companies. Sample 1 incorporates data on firms´ headquarters´ locations and sample 2 encompasses data on firms´ major stock exchanges´ locations.

(31)

Table 2 Descriptive statistics for sample 1 Corporate Governance Total Rhineland Anglo-Saxon Human rights violation No Frequency 64 85 149 Percent 88,9% 68% 100% Yes Frequency 8 40 48 Percent 11,1% 32% 100% Total Frequency 72 125 N = 197 Percent 100% 100% 100% Source: Author

Table 3 Descriptive statistics for sample 2

Corporate Governance Total Rhineland Anglo-Saxon Human rights violation No Frequency 50 80 130 Percent 87,7% 69% 100% Yes Frequency 7 36 43 Percent 12,3% 31% 100% Total Frequency 57 116 N = 173 Percent 100% 100% 100% Source: Author 4.3 Variables

For both hypotheses there is one dependent variable: the Company Abuse Allegation, and one independent variable: the corporate governance regime. However for hypothesis 1 the independent variable is the corporate governance regime of the country where the multinational is headquartered. For hypothesis 2 it is the corporate governance regime of the country where the multinational is mainly exchanged. Companies which are headquartered (hypothesis 1) or exchanged (hypothesis 2) in countries following the Anglo-Saxon regime were coded as “1”, and companies headquartered or exchanged in countries following the Rhineland regime were coded as “0”. Moreover, to distinguish between multinationals that did violate labor or economic development rights (CAA) and multinationals that did not, I made use of a dummy variable (0/1). Companies which have a CAA were coded as “1” and companies which do not were coded as “0”. Consequently the dependent and independent variables are both categorical variables.

(32)

Table 4 Variables and sources

Variables Code Name Definition Values Source

Dependent Is CAA Is company abuse allegation The company is accused of a labor and/or economic development human rights violation, or not 0 = No 1 = Yes

The Corporate & Human Rights Database (CHRD, 2014)

Hypothesis 1 Home Country country Home Country where the company is

headquartered -

The Orbis Database https://orbis.bvdinfo.com Independent H1 CG governance Corporate The home country follows the Anglo-Saxon or the Rhineland regime of corporate governance 0 = country follows the Rhineland model 1= country follows the Anglo-Saxon model Table 1:

Acas, 2012; Aguilera & Jackson, 2010; Bebchuk & Roe, 1999; Brown et al., 1998; Freeman,

1984; Freeman et al. 2004; Friedman, 1970; García-Castro

et al., 2008; Gospel & Pendleton, 2006; Koen & Mason, 2005; and Rossouw,

2008.

Hypothesis 2 Country SE exchange Stock country

Country where the company is

primarily listed -

The Orbis Database https://orbis.bvdinfo.com Independent H2 CG governance Corporate The country where the company is primarily listed follows the Anglo-Saxon or the Rhineland regime of corporate governance 0 = country follows the Rhineland model 1= country follows the Anglo-Saxon model Table 1:

Acas, 2012; Aguilera & Jackson, 2010; Bebchuk & Roe, 1999; Brown et al., 1998; Freeman,

1984; Freeman et al. 2004; Friedman, 1970; García-Castro

et al., 2008; Gospel & Pendleton, 2006; Koen & Mason, 2005; and Rossouw,

2008.

(33)

4.4 Methodology

To test the hypotheses I ran, for each hypothesis separately, a chi-square (χ²) cross-tabulation (crosstab) analysis using IBM SPSS Statistics (version 21). This statistical method is also called the Pearson´s chi-square test, the chi-square test for independence or the chi-square test of association. In this method, the numbers of cases that fall into each combination of categories, called the observed count, are compared to the expected counts of the cells, which are the frequencies expected if the null hypothesis (see below) is true. In other words, it measures the discrepancy existing between observed and expected frequencies (Field, 2013). Based on this method, the following hypotheses are analyzed:

• The null hypothesis (H0): The two categorical variables are independent.

• The alternative hypothesis (Ha): The two categorical variables are related.

The difference between expected and observed counts is large when the significance value, or “p value”, is small. Significance levels are expressions of the likelihood of the relationship between two variables being due to sampling error. For instance, a significance level of p = .05 means that the likelihood of a relationship between two variables being due to sampling error is 5 out of 100. If the chance of sampling error is less, for example 1 out of 100 (p = .01) or 1 out of 1000 (p = .001), than the significance level is higher. In case the significance value is small enough (p must be less than .05) it is possible to reject the null hypothesis in favor of the alternative one (Field, 2013).

The Pearson´s chi-square test is reliable only when the two following assumptions are met: first, data should be independent, which means that each case (person, item or entity) should contribute to only one cell of the contingency table (tables which have two categorical variables and each variable has only two categories). Second, the expected frequencies should be greater than 5 for two by two contingency tables (Field, 2013). Furthermore, for this statistical method, SPSS calculates automatically significance levels using the asymptotic method. The issue with that is that the asymptotic method may fail to produce reliable results when the data set is sparse, small, contains many ties, or is unbalanced (Mehta, 1989). This is why it is recommended to use the Pearson chi-square test only when the expected frequencies in each cell are greater than 5. When this criterion is met, the sampling distribution is probably close enough to a perfect chi-square distribution to rely on the results (Field, 2013). When the criterion is not met, Fisher (1935) suggests the use of exact p values. Another limitation of the Pearson chi-square test is the difficulty of interpretation when there are large numbers of categories (20 or more) in the independent or dependent variables (McHugh, 2013).

(34)

Nevertheless, the Pearson chi-square test is the most adequate statistical method for this study. First because the two assumptions are met, consequently it is possible to rely on the results (Field, 2013). The first assumption is met because the selected firms either follow the Anglo-Saxon model or the Rhineland model of corporate governance, and they either perpetrated a human right violation (CAA) or not. Consequently, each company falls only in one cell of the contingency table. The second assumption is met because the smallest expected count is 17.5 for sample 1 (Table 5) and 14.2 for sample 2 (Table 6), and both values exceed 5. Second, the difficulty of interpretation when there are large numbers of categories (McHugh, 2013) is insignificant because in this study there are only two categories in both the independent and dependent variables. Additionally, this statistical test has been chosen because it has strong advantages. For instance, (1) it tells us whether a relationship found in a sample can be generalized to the total population from which the sample was drawn (Nzssds, 2010), (2) it makes no assumptions about the distribution of the population, unlike many other statistics which assume certain characteristics about the distribution of the population such as normality, (3) it is easy to compute, which decreases the chance of miscalculations, (4) detailed information can be derived from the test, and (5) it is flexible in handling data (McHugh, 2013).

(35)

5. Results

In this section, an overview of the results of the empirical analysis will be provided and based on these results conclusions about the hypotheses will be drawn.

5.1 Hypotheses testing

5.1.1 Cross-tabulations

A crosstab table is a two-way table which contains the observed count and the expected frequency for each cell (Field, 2013). The contingency table for sample 1 (Table 5) indicates that when multinationals were headquartered in countries which follow the Rhineland model, a lower number of multinationals violated human rights (no = 8) than was expected (ne = 17.5),

however when the multinationals were headquartered in countries which follow the Anglo-Saxon model, a higher number of multinationals violated human rights (no = 40) than would be

expected by chance (ne = 30.5). Additionally, the contingency table for sample 2 (Table 6) shows

that when multinationals were primarily listed in countries which follow the Rhineland model, a lower number of multinationals violated human rights (no = 7) than was expected (ne = 14.2), but

when multinationals were primarily listed in countries which follow the Anglo-Saxon model, a higher number of multinationals violated human rights (no = 36) than would be expected by

chance (ne = 28.8). Consequently, it seems that depending on their corporate governance

regime, multinationals are more or less likely to violate labor and economic development rights. However, the question is whether the difference is statistically significant. This question is affirmatively answered in the following section.

Table 5 Crosstab for sample 1

Corporate Governance Total Rhineland Anglo-Saxon Is CAA No Observed 64 85 149 Expected 54,5 94,5 149 Yes Observed no 8 40 48 Expected ne 17,5 30,5 48

Total Total Observed 72 125 197

Total Expected 72 125 197

(36)

Table 6 Crosstab for sample 2 Corporate Governance Total Rhineland Anglo-Saxon Is CAA No Observed 50 80 130 Expected 42,8 87,2 130 Yes Observed no 7 36 43 Expected ne 14,2 28,8 43

Total Total Observed 57 116 173

Total Expected 57 116 173

Source: Author

5.1.2 Chi-Square test outputs

Table 7 indicates that there is a significant association between the type of corporate governance regime in the home country and whether or not the multinationals violated human rights, χ² (1, N = 197) = 10.82, p = .001. Furthermore, Table 8 shows that there is a significant association between the type of corporate governance regime of the country where multinationals are primarily listed, and whether or not the multinationals violated human rights, χ² (1, N = 173) = 7.20, p = .007.

Table 7 Chi-square test output for sample 1

Value df Asymp. Sig. (2-sided)

Pearson chi-square 10.817 1 .001

N of valid cases 197

Source: Author

Table 8 Chi-square test output for sample 2

Value df Asymp. Sig. (2-sided)

Pearson chi-square 7.197 1 .007

N of valid cases 173

Referenties

GERELATEERDE DOCUMENTEN

This is the first study that evaluates the differences in gene expression of the chondrocytes before they are used in ACI to assess the chondrogenic potency of the inserted cells

• Circa 97% heeft in 2013 toegang tot het internet. • In Europa gaat Nederland aan kop met IJsland. In de wereld met Zuid Korea. • Verzadigingspunt fysieke toegang in Nederland

Until now, a few clinical studies showed that BKPyV miRNA levels in renal transplant patients can be detected in plasma and urine and in recipients with BKPyVAN 104,105..

Several techniques which have been used to increase the performance of the metal oxide semiconductor field effect transistor (MOSFET) are also applied to the FinFET; such as

Proposition 7: Instances of environmental abuse or labor violations that received negative media attention regarding the entire industry are likely to have a damaging

Bij de vraag of uitzendfrequenties ook eigendom zijn stelde het Hof dat de vraag die moet worden onderzocht, is of de omstandigheden bij de zaak als geheel aan belanghebbende

Alternatieve gassen en aansprakelijkheid: De Nederlandse gasketen in een geliberaliseerde markt: contractuele en buitencontractuele aansprakelijkheid van groen-gasinvoeding

Crystal structures of glutamate transporters in multiple different conformations have been solved, but most structures were determined at relatively low resolution, with