The relationship between home country business ethics and values and the
type of company involvement in human right violations
Msc Business Administration – International Management Student: Lucas Poen
Student Number: 11157941
Supervisor: Michelle Westermann-Behaylo Second Reader: Mashiho Mihalache June 23th 2017
Statement of Originality
This document is written by Student Lucas Poen 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.
Abstract
Several scholars note that companies from countries with high values for human rights and a focus on strong business ethics are still involved in human right violations. Most of these violations happen in developing countries, which is due to the often-weak institutions and the absence of cross-border entities that can hold these companies accountable. This thesis looks at the influence of home country factors such as business ethics and national values on the level of complicity of firms in human right violations. Firms can be directly involved in violations, but there are also ways in which they are less directly involved, by for example supporting local governments in their unethical practices or by working together with a violating company somewhere along the supply chain. With the use of a database of around 450 alleged human right violations and different surveys concerning business ethics and national values, the relation between these home country factors and the level of involvement is tested. The study furthermore looks at the moderating role of the share of ownership to see if subsidiaries behave differently when they are wholly owned or a joint venture. The outcome of this thesis is that home country business ethics and national values do have a significant impact on the way in which multinational companies are involved in human right violations. This thesis also shows that the share of ownership indeed has a moderating role in this relation.
Table of Contents I Introduction 1 II Literature Review 4 A Degree of Complicity 4 B Business Ethics 9 C National Values 13 D Share of Ownership 15 E Research Question 18
III Theoretical Framework 19
A Home Country Effects 19
B Share of Ownership Effects 20
C Conceptual Model 21 IV Research Design 22 A Data Sources 22 B Variables 24 C Sample Selection 25 D Methodology 28 V Results 30 A Pre-test 30 B Analysis 31 VI Discussion 34
A Summary and Findings 34
B Implications for Managers 36
C Strengths and Limitations 37
D Future Research 38
VII Conclusion 40
1
I. Introduction
On the 31st of January 1999, the Secretary General of the United Nations Kofi Annan spoke the following words to the World Economic Forum: “I call on you -- individually through your firms, and collectively through your business associations -- to embrace, support and enact a
set of core values in the areas of human rights, labour standards, and environmental” (Annan,
1999). The speech pointed out the fragility and dangers of globalization for the victims of market failures. Research has shown that a lot of multinational companies are involved in human right violations (Ruggie, 2013). Even though, international entities like the UN actively promote human rights, many multinational enterprises (MNEs) still seem to favour profit over the human rights of the people in the countries where their products are being produced. The reason these atrocities still happen is because there is no effective international judicial entity that can convict human right violations, which leaves it to the governments in often lower developed countries who do not have the resources or priorities to convict these powerful corporations (McCorquodale and Simons, 2007).
Multinational enterprises and their relation to human rights is covered in a lot of literature, however MNEs can be involved in human rights in several ways and this aspect of levels of complicity is not sufficiently covered in the current literature. Multinational enterprises can for example be directly involved by deliberately violating human rights, but they can also be involved in more indirect ways. What is often seen is that human rights abuses are committed by governments or other organisations and that corporations in one way or another benefit from these violations (Clapham and Jerbi, 2000). A good example of the different complicity levels is the case of the multinational oil company Shell. Shell is accused of being involved in several human right violations in Nigeria. They are being accused of directly polluting the Niger Delta due to oil spills and air pollution, which can have serious health consequences for the locals (Okonta and Douglas, 2003). And they are also being accused of fuelling tensions between communities, with horrific violent consequences and the
2
death of peaceful protestors as a result (Zimmer, 2010). These tensions between the different communities and its consequences cannot primarily be blamed on the activities of Shell, but also on the government who sometimes violently forced the local minority ethnic groups to leave their traditional land (Zimmer, 2010). However, the presence of Shell and their support of public security forces have fuelled the tension making Shell complicit in the human right violations that followed (Pegg, 1999). The first accusation would be a human right violation in which Shell has direct complicity, while the latter accusation is an example of indirect complicity.
It is interesting to see that Shell, a company originating from a country with a free democracy and strong protected civil liberties, is nevertheless involved in human right violations. A phenomena that is also noted by scholars who also note that a lot of western companies from strong democracies with high levels of respect for civil liberties are involved in human right violations (Pegg, 2012; Fiaschi et al., 2012). Remarkably, other scholars often argue that home country values and ethics are transferred from the parent firm to the subsidiary (Christie et al., 2003; Matten and Moon, 2008; Robertson et al., 2008; Fiaschi et al., 2012; Noorderhaven and Harzing, 2003) which makes no sense if you see that companies from more developed countries are often involved in human right violations. Scholars also suggest that firms facing pressures for responsible behaviour in their home countries may seek to transfer socially irresponsible practices to overseas subsidiaries (Surrocca, et al., 2013). On another note, you would expect companies from lower developed countries to be even more often involved in human right violation. Scholars like Fiaschi, et al. (2012) however argue that lower developed countries are even more active in CSR practices. These discrepancies rais questions on the role of home country values and ethics on subsidiaries that are involved in human right violations. Furthermore, scholars note that corporations use sketchy ownership types to reduce the liability of the parent firm (Wouters and Ryngart, 2008; Thorsen and Meisling, 2006), this
3
might explain the absence of home country ethics and values. These contradicting views have led to the following research question for this thesis:
“How do home country business ethics and values affect the level of complicity of a firm in
human right violations and how is this relation affected by the share of ownership of the subsidiary?”
The contribution of this master thesis is that it studies the home country influences on subsidiary behaviour in a way that has not been done before. In the literature, we can see that scholars have previously done a lot of research on the effect of home country factors on subsidiary behaviour. This thesis builds on this concept, but looks at this influence on the level of involvement in human rights abuses, a perspective that has not yet received a lot of attention in the literature. Furthermore, unlike other studies which are most often based on case studies or small population quantitative studies, this thesis will use quantitative data from a large population of actual reported human right accusations from different parts of the world. The outcomes of this thesis can therefore be of great interest in the academic field, but also businesses, institutions and organisations can get new insights on how they can prevent human right violations from happening.
This thesis will be continued by a thorough literature review that covers all the variables involved. It will start with some background on human rights and international business, after which firm complicity, business ethics, cultural values and ownership shares will be thoroughly discussed. After the literature review a theoretical framework based on the literature is constructed and the different hypotheses and the conceptual model are presented. Subsequently, the methodology and data analysis will lead to discussion and a conclusion. Finally, recommendations for future research will be given.
4
II. Literature Review
The literature review will focus on the different constructs that are used in this thesis. It will not only discuss the literature discussing each of the constructs, but it will also look at how these different ideas are related to each other in the current literature. I will start by (II.A) giving a brief introduction to the scene of human rights and business and firm complicity, after which I will look at (II.B) business ethics and how this affects business strategies and behaviour. Then the third section (II.C) will look at national values and how this influences business behaviour. Finally, the last section (II.D) will look at the share of ownership and how this influences subsidiary behaviour.
A. Level of Involvement Business and Human Rights
The end of the 20th century signified the rise of globalisation, the increasing integration of different markets and the rise of corporate power. Many people see this tightening of the global markets as a positive thing which has resulted in lower prices, economic growth and the availability of knowledge and technology (Aliber and Click, 1993). Nevertheless, there are many scholars that point out the other side of globalization. They argue that in some way it has made the rich richer and the poor poorer, as it has led to the exploitation of labour and nature and the increasing power of multinational companies in countries that have weak institutions (Enderle, 2012; Nault, 2011). Countries, companies and people that are able to take advantage of the process of globalization prosper, but signs of different human right violations are coming to light.
There is no overarching global entity that can control and enforce sanctions when a multinational company commits a human right violation. MNEs have to abide to the local national judiciary systems, but due to weak institutions in certain countries, MNEs can often benefit and make profits over the back of the people in the host countries (Porter, 1999). Already
5
in the 1990’s the UN acknowledged this issue and proposed several treaties to increase the responsibilities of the MNE and to introduce norms that will prevent new human right violations from happening. Unfortunately, these treaties were not really effective, due to different views of the actors involved on who should take responsibility (Ruggie, 2013). That is why in 2011 the UN human rights council introduced the “Guiding Principles on Human Rights”. These principles were supposed to be more effective and clearer on who should take which responsibility than the treaties that were presented by the UN earlier. The Guiding Principles on human rights revolve around three main anchors. The first one is the responsibility of the state to protect people from human right violations. The second revolves around the company who should respect human rights in the form of due diligence. Finally, the third principle focusses on ensuring the remedy for victims of human right violations (Ruggie, 2013). The guiding principles were received very positively and were adopted by the OECD, the EU and other overarching international standard setting bodies (Blitt, 2013). Nevertheless, other authors like Fasterling and Demuijnck (2013) argue that the guiding principles in the way they are formulated now reduce the effectivity of due diligence on the side of the corporations.
International law leaves the responsibility of prosecutions of human right violations to individual states. Corporations are being mentioned in many treaties in international law, but nevertheless international law is not able to hold a corporation liable for their human right violations, because these international treaties are often based on soft law and are not universally accepted (Kobrin, 2009). Kuuya (2008) identifies an often-occurring situation where the states, who should be able to hold corporations accountable, are sometimes complicit in the human right violations themselves. This situation can lead to some sort of carte blanche for the corporations, as there is no overarching international entity that can hold these companies accountable. McCorquodale and Simons (2007) provide a strong example where the international laws and treaties fall short in convicting human right violations, because they are
6
not sufficiently binding. Currently the responsibility of convicting and enforcing law on MNEs who are involved with human right violations lies with the host country government, who often do not have the resources, power or will (e.g. repressive regimes or failed states) to do so. According to McCorquodale and Simons (2007) a big responsibility lies with the home country governments, who according to the International Human Rights Law have the obligation to stand up against human right violations even if they do not occur in their own country, because the International Human Right Law is not restricted by territory. A theory that might explain the complicity of governments in human rights is the theory of “the race to the bottom”. This theory argues that governments are in a regulatory competition to lower their regulatory standards in order to attract more foreign investments (Porter, 1999). This leads to weakening of institutions in developing countries and eventually in an increase in human right violations (Spar, 1998). These violations can be seen in the area labour laws, but also in environmental regulation (Singh and Zammit, 2004; Spar, 1998; Porter, 1999).
Causes for human right violations
In this thesis, I will try to identify factors that influence the complicity of firms in human right violations. It is therefore important to find out what the current literature identifies as important factors that influence the occurrence of human right violations by corporations. Giuliani and Macchi (2014) did a thorough literature review and identified a range of factors that can influence the occurrence of human right violations. According to Giuliani and Macchi (2014) both internal and external factors to the MNE influence the occurrence of human right violations.
Looking at external factors you can think of host country factors and industry factors. There are several host country factors that can influence the presence of human right violations. Giuliani and Macchi (2014) argue that it is important to look at the state capacity, social capability and civil society of the host country. De Schutter (2006) and Eroglu (2008) for
7
example argue that the strength of the host country institutions and legal system plays a crucial role in the presence of human right violations. Other scholars focus more on the power of civil society and stress the vital role of NGO’s, local communities and activist groups in the shaping of corporate behaviour and increasing respect for human rights (Gereffi et al., 2001; Calvano, 2008).
Giuliani and Macchi (2014) also identify that the type of the industry the MNE operates in can play a role in the occurrence of human right violations. The data that will be used in this thesis is from the extractive industry. This particular industry is notorious for the often-occurring presence of human right violations along the production line. The extractive industries for example often get engaged in conflicts with indigenous people over land rights (Burger, 2014). Sama (2006) and Spar (1998) also looked at the role of competition in an industry. They found out that a high competition intensity can lead to more irresponsible behaviour of the firm. Other scholars looked at the role of the level of technological intensity. Blanton and Blanton (2009) found out that industries that require less technological intensity are more likely to commit human right violations in the field of labour rights and working conditions, while Papaioannou (2006) on the other hand argues that the necessity of natural resources plays a significant role.
Internal factors that might influence corporate behaviour concerning human rights are more focussed on the corporation itself, its home country and subsidiary strategies. Many scholars argue that the role of home country factors should not be overlooked. Christie et al., (2003) for example found out that national business ethics play a role in the behaviour of MNE managers. While Pegg (2012) and Fiaschi et al., (2012) found out that there is no significant difference in unethical behaviour between developing and developed countries. Other internal factors that can potentially influence human right violations are concerned with the subsidiary. The level of autonomy of the subsidiary sheds an interesting light on the issue. There are for
8
example signs that MNEs try to outsource human right violations to minority-owned subsidiaries (Thorsen and Meisling, 2004; Clapham, 2006; Wouters and Ryngaert, 2009; De Schutter, 2006B; Surroca, et al., 2013). Also, the type of strategy of the subsidiary plays a significant role in the occurrence of human right violations and on the type of human right violations (Blanton and Blanton, 2009). Resource seeking subsidiaries tend to show less respect to environmental human rights, while subsidiaries focussing on efficiency are more likely to be connected to labour human right violations (Blanton and Blanton, 2009).
Complicity of firms
As shown in the part above there are multiple factors that influences occurrences of human right violations by MNES. This thesis, however, focusses not on the factors that mediate the occurrence of human right violations, but rather on the degree of complicity of firms. The specific focus on firm complicity is the result of the observation that lots most of the human right violations committed by companies are the result of complicity with another third-party actor (Kobrin, 2009). The legal meaning of complicity is; “knowingly providing practical assistance or encouragement that has a substantial effect on the commission of a crime” (Ruggie, 2008). The non-legal context of corporate complicity has become an important benchmark for investors, companies and organisations. Complicity allegations can possibly have severe consequences for the company’s reputation and can even lead to divestment (Ruggie 2008).
Clapham and Jerbi (2000) identify four distinct categories of corporate complicity. The first one is direct complicity which is when a corporation has a direct or causal contribution to a human rights violation. The second category is indirect complicity which is when a corporation does not directly contribute to a violation, but rather supports the violator to commit a human right violation. The third category is beneficial complicity which is a company that is not directly or indirectly involved with a human right violations, but the company knows that
9
it benefits for human right violations committed by a third party. The last category is silent complicity which is when the corporation itself is not involved in any human right violations but when they do not speak out on human right violations they have knowledge on committed by other parties.
Kobrin (2009) argues that a majority of the corporate human right violations are the result of indirect complicity with the perpetrator, which is most often the host country government. Scholars are therefore, calling for a guiding role for corporations. Wettstein (2010) stresses the political power of multinational enterprises. He argues that corporations have a lot of influence in global governance architecture, but also that they have a lot of influence on national government policies. Hsieh (2009) even claims that corporations have the duty to advice and encourage countries with weaker institutions to improve and strengthen their institutions. With the lack of an effective entity that can enforce international human right law many scholars argue that the companies should use their power to introduce their own values and ethics in the countries with weaker institution. As mentioned, there are different factors that play a role in the occurrence of human right violations. One of the factors that might have an impact on the way a company copes with weak institutions or a violating government is the strength of the business ethics in a company.
B. Business Ethics
One of the concepts I will focus on in this thesis is business ethics. Business Ethics is an interesting field of study and can, according to Crane and Matten (2016), best be defined as; “the study of business, activities and decisions where issues of right and wrong are addressed”.
The factor that makes business ethics such a difficult matter is that it is on the edge of the law. Business ethics can be seen as a grey area, the area of decisions between right and wrong that is not covered by the law (Crane and Matten, 2016). It is therefore often seen that as a consequence of a business ethics crisis, regulators respond with the implementation of new
10
legislation (Blodgett, 2012). Due to the increased power and significance of big companies in everyday society, business ethics have become a very important issue in society.
This has created a lot of pressure on companies to become more accountable and responsible, which has led to the emergence of codes of ethical conduct. Codes of ethical conducts are a voluntary way in which companies try to standardise ethical behaviour which can be enforced on different levels (Crane and Matten, 2016). Corporate codes can be implemented on for example an industry level, corporate level or by an organization (Crane and Matten, 2016). Much research has been done on the effectivity of codes of ethical conduct and the outcomes are mixed. Some researchers show that codes can improve ethical behaviour, but that the quality of these codes play a significant role in the effectiveness (Erwin, 2010) or further that the way the code is formulated is important (Cassell, et al. 1997). Nevertheless, there are also scholars that argue that these codes are not effective due to the absence of legal enforcement (McNeil & Li, 2006). Vogel (2010) argues that international corporate codes of conduct have resulted in improvements regarding labour, environmental and human rights, but that the effectiveness is still limited due to the lack of governmental enforcement. Nevertheless, Vogel (2010) argues that these private corporate codes are often stronger and more effective than the regulatory frameworks in developing countries, and are therefore an improvement. The opinions are however diffused as for example; Kaptein and Schwarz (2008) introduce another perspective and argue that such codes are often just a PR device to attract customers, but that the companies will keep doing business as usual.
The traditional study of business ethics covers three different levels; from micro (individuals), to macro (organizations) and meso (systems) (Enderle, 1997). Different companies and organizations have diverse types of ethical issues. For smaller companies, most ethical issues occur around employee issues, while in larger corporations ethical issues are more concerned with financial integrity and customer relations (Crane and Matten, 2016). The data
11
that will be used in this thesis is based on worker conduct (micro), which in turn tells us something about the organisation of business ethics in the specific country (macro, meso). Business Ethics and International Business
Globalization and the rise of multinational enterprises brought a whole new perspective to the traditional study of business ethics. The traditional three level conception of business ethics do not cover the new international dimension. All the three levels of conception need their own type of extension when looking at them from an international perspective (Enderle, 2015). The micro perspective for example gains a whole new cross-cultural aspect, while the macro perspective has to deal with new cross-national organisations. When looking at business ethics in an international context at the meso level, it is important to realize that the perceptions of business ethics are different depending on the culture (Gift, et al. 2013).
It is therefore interesting to take a close look at what happens when an MNE moves into a country that has different perceptions of business ethics than the perceptions of business ethics in the home country of the MNE. Donaldson and Dunfee (1994) also identified the differences in perceptions of business ethics around the world. They argue that adapting to host country business ethics should not cause any issues as long as companies respect the core human values, and local traditions. Palazo (2002) and Helin and Sandström (2008) agree with this view and argue that universal business ethics are hard to achieve. They argue that the way the business ethics are formulated and implemented is dependent on the country they are in. So, in order for MNES to have effective business ethics codes, MNES should adapt to the specific business culture in the different host countries.
Godiwalla (2012) agrees and argues that the role of the host country culture often has a high impact on business ethics in the subsidiary, but that the implementation of the type of business ethics for the subsidiary depends on the strategy of the parent firm. It depends on whether the parent firm is seeking to be recognized for their global social responsibility or
12
whether they benefit from diverse cultural standards around the world (Godiwalla, 2012). Donaldson (2002) poses that companies need to find a golden mean, between on the one had indulging in local culture and ethics, while on the other hand making sure that the home country core values are not reneged. Enderle (2014) looks at it differently and proposes that international relations play an important role in the influence of business ethics from the home country. He argues that the relations between the countries defines the way the way a company positions itself. As shown, different scholars stress the importance of the host country business ethics as important factors for subsidiary behaviour. While on another note, many scholars still
argue that the home country culture has a lot of impact on the MNE behaviour in foreign countries (Christie et al., 2003; Matten and Moon, 2008; Robertson et al., 2008; Fiaschi et al., 2012). Scholten and Dam (2007) agree and found out that location of the headquarters plays a very substantial role in the business ethics of the firm.
International Business Ethics Perceptions
Several case studies have been done, to identify the actual differences in business ethics or abidance of ethical conduct codes around the world. Matten and Moon (2008) and Winston (2003) look at specifically the differences in business ethics between the US and the EU. They identify differences in the way companies in these countries behave and how they comply to the Universal Declaration of Human Rights. They find that companies from the US have a lot more difficulties with compliance to standardised human right treaties, due to their embedded national business system.
Other scholars focus more on the differences between developed and emerging countries. Pegg (2012) for example argues that Chinese companies are not very active in promoting human rights when they are in countries where this would be necessary, but at the same time they are not very different in behaviour than companies from developed countries. Fiaschi et al. (2012) found out that BRIC countries are even more proactive in their CSR
13
activities than companies that come from more developed countries. According to Fiaschi et al. (2012) they do this in order to achieve more legitimacy and to build a reputation. These differences between the different degrees of business ethics lay in the cultural foundations of the different countries (Gift, et al,. 2013).
C. National Values
Another variable that I will look at are cultural values and their relation to human rights involvement. The core of culture, according to Hofstede (1990), lays in the values. These values are created in your early years and are determined by the environment you grow up in (Lundberg, 1985). Values define your judgements on for example what is good and what is bad or what is normal and what is abnormal in conscious and unconscious ways (Hofstede 1990). Culture can change due to modernization and economic development and so do values. Nevertheless, traditional core values are also very durable and they continue to reflect a society’s cultural heritage (Inglehart and Baker, 2000). Inglehart and Baker (2000) have seen that modernization indeed changes cultural values, but at the same time they argue that modernization does not lead to a determined direction. A lot of factors such as historical and societal context play a crucial role in the way cultural values change.
Cultural Values and International Business
Due to globalization, we see that values around the world become more and more homogeneous (Leung, et al., 2005). However, as pointed out by Inglehart and Baker (2000) traditional values, such as for example values for human rights, will keep reflecting the society’s cultural heritage. It is therefore interesting to see in what way home country cultural values influence the behaviour and strategies from MNE managers in foreign countries. Many studies have been done on the influence of home country cultural values and their effect on MNE behaviour, and the outcomes are very consistent. Tse (1988) for example identifies that ethnic culture plays a significant role in making business decisions. While, Geletkanycz (1997) used Hofstede’s
14
Cultural Dimensions with which he finds that cultural values play a role in strategy and leadership type adherence. Also, Schein (2010) argues that business leaders with a distinct cultural background will also have different views upon relationships among people or regarding the environment, therefore they are also expected to behave differently in specific situations.
Cultural Values and Human Rights
When looking at values of different countries on human rights, it would be no surprise that different countries with different values look differently at the concept of human rights. In 1948 the international community signed the Universal Declaration of Human Rights. This was an agreement on a universal perspective and recognition of human rights and achieving this consensus at that time was a breakthrough after years of worldwide division (Cerna, 1994). Nevertheless, the declaration, but more specifically the universality of human rights has led to some discussion afterwards.
Many non-western societies argue that the universal human rights are drafted from a very western perspective and they do not cover the needs, values and cultures of other parts of the world (Cerna, 1994; Berms, 2001). Donnelly (1982) for example argues that human rights in the way we know them now are really a western thing which mostly focusses on entitlement, while non-western civilizations, like in China, Africa, the Middle East and India, for centuries already emphasised the importance of human dignity. The International Federation for Human Rights (2015) particularly looks at the case of Asian countries where the interpretation of human rights is different. The International Federation for Human Rights for example found out that Asian cultures often revolve more around the community instead of individual rights. This debate about universality of human rights will remain a difficult subject, because different historical and cultural perspectives will interpret them differently.
15
Looking at the World Report by Human Rights Watch (2017) you can see that countries put different amounts of efforts in protecting and respecting human rights, which could be a result of how much value these countries have on human rights. The Human Rights Council (2012) introduces another perspective on the difference in human rights values by stressing the importance of the positive and negative effects of traditional values on the implementation of human rights. The positive effects are that many traditional values already promote human rights. The negative effects however, are that in many countries there are also traditional values that are not in line with human rights. Examples of these traditional cultural values are for example violence against women or the caste system in India (Donnelly, 2007). The differences in interpretation of business ethics and cultural values between different countries, might become an issue when companies from different parts of the world work together in shared ownership.
D. Share of Ownership
In this thesis, I will look at national business ethics and national values and how these home country factors influence the behaviour of subsidiaries in the host country. A factor that can moderate this relation is the share of ownership and how much influence the parent firm has on its subsidiaries.
Subsidiary mandate
Looking at the roles of subsidiaries in corporations we see that a lot has changed in the last decades. Where subsidiaries played a very passive role in the corporation in the past, Birkinshaw, et al. (1998) argue that the subsidiaries have become a lot more important as a strategic asset within the corporation. Subsidiaries are for example even used for asset seeking R&D (Nieto and Rodriguez, 2011). Birkinshaw, et al. (1998) identifies three concepts that play a role in the contributory role of the subsidiary. The first concept is the level of entrepreneurship and leadership from within the firm. The second concept looks at the level of autonomy and the
16
communication with the parent firm. The third concept they propose is the business environment of the subsidiary and level of competition. For this thesis, the focus will be on the role of the parent firm and more specifically how home country parent firm business ethics and national values play a role in subsidiary behaviour. Noorderhaven and Harzing (2003) did a study on the role of the home country culture and institutions on subsidiaries and concluded that they can play a strong role in subsidiary behaviour. The strength of the presence of such a “country of origin effect”, as they call it, is moderated by factors such as the homogeneity of the culture in the home country and the size of the home country.
Subsidiary role in human rights
Nieto and Rodriguez (2011) argue that often subsidiaries become more and more strategically important entities in the MNE. According to Surroca, et al. (2013) this would be a positive development for the reduction of human right violations, since he argues in his paper that subsidiaries that are focussed on R&D and entrepreneurship are less likely to be part in corporate social irresponsibility, in comparison to subsidiaries that are focussed more on seeking efficiency and resources. When looking at the subsidiary and their autonomy from the parent firm many scholars have identified that more independent subsidiaries are more likely to become more and more influenced by the local politics, culture and economy (Iršová, and Havránek, 2013; Giroud, 2007). This would suggest that more independent subsidiaries would be more likely to become involved in irresponsible and unethical business conduct, if the host country institutions and culture are weak and have less strong values for human rights (Giuliani and Macchi, 2014).
Share of ownership
The opportunities that globalization brought for companies drove companies to expand their businesses into other countries. When companies decide to expand their business, there are several options for a firm to do this. They can either chose for a contractual agreement
17
(franchises, licenses, export) or when they choose to physically move their company to a foreign country they can choose between shared ownership (joint venture (JV)) and full ownership (wholly owned subsidiary (WOS)) (Brouthers and Hennart, 2007). Most of the scholars look at the different types of entry modes as a sequence of increasing control, risk and commitment (Anderson and Gatignon, 1986; Erramilli & Rao, 1990; Hill, et al., 1990)
This thesis will particularly look at the share of ownership of the parent firm and will therefore look at joint ventures and wholly owned subsidiaries. The rationale behind firms choosing for a joint venture is most often either to bridge the disadvantages compared to local competitors (Dunning, 1988), to get access to new technology and resources (Luo and Tung (2007) or to reduce the liability of foreignness (Campbell, et al., 2012). Engaging in a joint venture can in some cases be necessary to get access to the resources the company wants. According to Roording and de Vaal (2010) it can be of great advantage for MNES to work together with local companies to get a better access to resources.
The share of ownership in a joint venture is therefore a major decision because of for example; the risk of knowledge spill overs when you have a high share of ownership (Talay and Cavusgil, 2009), or the risks of relational problems in a joint venture (Parkhe, 1993). These relational problems are often rooted in cultural differences (Parkhe, 1993). Other scholars like Farkas and Avny (2005) are also able to show that cross cultural differences can have a negative impact on JV performance. Different scholars point out the importance of both the host country national culture as well as the home country based organizational cultures in the organizational culture of international joint ventures (Lane and Beamish, 1990). Barger (2007) argues that joint ventures with parents that have different values and norms are bound to have tensions and misunderstandings. Lee (2009) agrees and argues that the type of ownership structure plays a significant role in the amount of effort the company puts into corporate social responsibility.
18
Which seems logical as Giuliani and Macchi (2014) argue that subsidiaries that are more autonomous are more likely to indulge in unethical behaviour.
Surroca, et al. (2003) point out that MNES in some cases choose to use minority share in a joint venture to avoid liability for the parent firm. In this way corporations enter a grey area of law where the minority ownership reduces the liability of the parent firm (Wouters and Ryngart, 2008; Thorsen and Meisling, 2006). This is attractive for MNEs because in this way they can outsource unethical practices that are not allowed in the home country (Clough, 2005).
E. Research Question
The different perspectives as show by the literature review has led to the following research question:
“How do home country business ethics and cultural values affect the complicity of a firm in human right violations and how is this relation affected by the share of ownership of the subsidiary?”
19
III. Theoretical Framework A. Home Country Effect
The literature review has shown that the occurrence of human right violations is influenced by a lot of different aspects. The focus in this thesis is on the degree of complicity of firms and how this concept is influenced by different home country factors. The first variable this thesis focusses on is the degree of business ethics in the home country. You would assume that when a company comes from a country with a strong degree of business ethics it will be less likely for this company to indulge in unethical behaviour. The question is however, whether these levels of business ethics are passed through to the subsidiaries if these subsidiaries are for example situated in a completely different institutional environment or in a country with lower developed regulations. As shown in the literature on business ethics and its influence on MNEs it shows that home country business ethics often do have an effect on subsidiary behaviour (Scholten and Dam, 2007). However, as some studies showed, home country business ethics do not always result in the predicted behaviour of the firms’ behaviour (Pegg, 2012; Fiaschi, et al., 2012). The paradox lies in the empirical fact that many firms from countries with strong business ethics are still complicit in human right violations. The assumption is therefore that home country business ethics do play a role in human right violation involvement, but that it does not necessarily mean that it reduces the amount of human right violations, but rather the type of involvement in these human right violations. Companies with stronger business ethics would most logically be less directly involved in human right violations.
This had led to the following hypothesis:
H1: Stronger home country business ethics negatively affects the degree of complicity of a firm in human right violations.
The second variable in this thesis is the national value concerning human rights in the home country. When the national values concerning human rights in a country are strong, you would
20
assume that this an impact on the managers and therefore also on the behaviour and strategies of the firm. Many scholars agree that home country culture influences the strategy and business decisions from managers from MNEs and their subsidiaries (Geletkanycz, 1997; Tse, 1988; Schein, 2010). The literature review furthermore shows that values concerning human rights differ across the world and human rights cannot easily be seen as a universal concept (Cerna, 1994; Berms, 2001). If the national values in the home country of the parent firm influence the behaviour of the subsidiary, it seems logical that subsidiaries from MNEs originating from a country with high values concerning human rights would be less likely to be complicit in human right violations. It is however also noted that the host country institutional and cultural factors can also have a significant impact on subsidiary behaviour, and might lead to unethical behaviour from the subsidiary (Giuliani and Macchi, 2014). The assumption is therefore that national values concerning human rights influences the way a firm is involved in human right violations. This has led to the following hypothesis:
H2: Strong national values in the home country concerning human rights negatively affects the degree of complicity of a firm in human right violations.
B. Share of Ownership Moderation
The share of ownership determines the degree of control, risk and responsibilities. The literature has shown that home country culture influences the behaviour of subsidiaries. The higher level of ownership would logically lead to stronger influence of home country business ethics and values. Nevertheless, some scholars argue that it is more interesting for companies to have a minority share in a subsidiary to reduce the liability of the parent firm. The following hypotheses will look at the moderating role of the ownership share and its strength on the relation between home country business ethics and values, and the degree of human right complicity. The logical assumption would be that the higher the ownership share is of a parent firm in its subsidiary, the stronger the home country influences would be. The possible role of
21
MNEs deliberate decision to have a minority share to reduce liability might introduce an interesting side note.
H3: The share of ownership of the subsidiary positively influences the relation tested in H1
H4: The share of ownership of the subsidiary positively influences the relation tested in H2
C. Conceptual Model
The literature review and the theoretical framework have led to the formulation of the following conceptual model. The degree of business ethics (H1) and the level of national values for human rights (H2) have a positive influence on the degree of complicity of a firm in human right violations. This interplay is positively moderated by the share of ownership (H3 + H4).
22
IV. Research Design
To test the relationship between the home country factors and the involvement of MNEs in human right violations the different hypotheses were drafted. To test these hypotheses, I use cross-sectional data from secondary sources. This section will show the data sources, how the variables are constructed and how the hypotheses will be tested.
A. Data Sources
The data used for looking at the dependent variable, the degree of firm complicity in human right violations, comes from the Corporations and Human Rights Database Project (CHRD). The quantitative data provided by the CHRD is based on qualitative data on Company Abuse Allegations (CAA) that are collected from the website of a non-profit organisation called The Business and Human Right Resource Centre (BHRRC). The BHRRC has become a very important player in the business and human right advocacy. The organisation does not only publish human rights allegations, but it also gives companies the ability to respond. Which according to Marr Page (2016) in some cases have led to more pressures to increase accountability.
The CHRD project is an initiative to quantify the CAA’s that are available on the website in order to make this information useful for academic research. By making the data quantitative the CHRD wants to increase the understanding of the different aspects that influence business behaviour and human rights. Quantifying the different cases is done through a coding process which focusses on different aspects from abuse allegations to responses and judicial actions. Due to the big amount of data that is available on the BHRRC website the data is coded by different coding teams who had a coding training to increase cohesiveness and validity.
This thesis particularly focusses on the degree of firm complicity (or degree of company involvement) which is one of the data variables in the survey. The CHRD makes a distinction
23
between three types of involvement. The first one is direct involvement, the second one is indirect involvement and the third one is involvement by supporting a government that is committing a human right violation. The database consists of 1247 cases in the extractive and textile industries and almost entirely consists of cases that happened between 2000 and 2014.
For the first hypothesis concerning business ethics data provided by the Global Business Ethics Survey (GBES, 2016) will be used. The survey was taken in 2015 and looks at worker conduct and workplace integrity, giving an interesting insight into national levels of business ethics. The survey covers 13 countries and tries to identify the aggregate levels of national business ethics per country. The variable in this survey that this thesis is focussed on is the amount of observed misconduct at the country levels. The data tells something about the national business ethics and can therefore be used to compare different countries. Apart from this survey there is not a lot of comparative empirical data on business ethics in different countries.
For the second hypothesis data concerning national values data from the World Value Survey (WVS) is used. The WVS is a database collected by social scientists concerning changing values and their effect on culture and politics. The WVS data bank has a very large data bank with a wide variety of values for almost every country. For this thesis, I will particularly focus on how people perceive the respect for individual human rights in their country. The data used is gathered between 2010 and 2014. The respondents had to answer the following question: “How much respect is there for individual human rights nowadays in this country?” The World Value Survey is an often used and respected source of data concerning value data (Inglehart and Abramson, 1999).
The third hypothesis on the share of ownership will be retrieved by looking at the Orbis database. The Orbis database has data on more than 200.000 private firms in the world. The Orbis database can give insight on the relationship between the parent firm and its subsidiaries
24
and can therefore also be used to identify the share of ownership. Furthermore, it provides the information on the home country of the parent firm which is also necessary for determining which home country values to look at.
B. Variables
The different datasets were prepared in excel and were later transferred and analysed in IBM SPSS Statistics 24. The first step was to test the data of the independent variables to find out whether they are normally distributed. The Kolmogorov-Smirnov test has showed that the datasets were not normally distributed. I therefore decided to code these datasets into categories to make sure that they can be used in non-parametric testing.
Dependent variable: Level of involvement
This variable defines the level of involvement of the firm in a human right violation and comes from the CHRD database. The level of involvement is determined for each case by dividing them as nominal categorical data into “direct involvement” (1), “indirect involvement” (2) and “supporting a violating government” (3).
Independent variable: Home country business ethics
The data for this variable comes from the Global Survey of Business Ethics. The variable looks at how much the respondents observe workplace misconduct in their company or organisation. By misconduct the survey means: “a violation of the law, an organization’s values or principles and/or universal ethical principles, e.g., respect, fairness, honesty” (GBES, 2016). This data is
coded in an ordinal categorical way in the following order; “countries in the sample with highest amounts of misconduct” (1), “countries with higher amounts of misconduct” (2), “countries with lower amounts of misconduct” (3) and “countries with lowest amount misconduct” (4). Independent variable: National values concerning human rights
The world values survey shows the perceived values for human rights as ranked by the people from that specific country. The categories based on the sample go from countries that have the
25
highest values concerning human rights “1” to countries that have the lowest values for human rights “5”.
Moderator variable: Share of ownership
For the moderator variable, the data was quantified by creating a dichotomous database based on the share of ownership of the subsidiary. For the testing, wholly owned subsidiaries are marked with “0” and joint ventures are marked with “1”. This variable is used for both hypotheses 3 and 4. The datasets used for these hypotheses are different due to the different home-country restrictions of hypotheses 1 and 2.
Table 1 Variable Description
C. Sample Selection
In order to be able to test the relations between the different variables it is important to make sure that the different variables cover the same sample. The availability of data ultimately defines the sample and the scope of this study.
Variable Variable type Values Source
Level of complicity Categorical 0: Direct 1: Indirect
2: Supporting violating government
CHRD
Business Ethics H1 Categorical 1: Lowest amount of misconduct (Germany, Japan, Spain)
2: Low amount of misconduct (South Korea, UK, USA)
3: High amount of misconduct (China, France, Italy, Mexico)
4: Highest amount of misconduct (Brazil, Russia, India)
GBES
National Values H2 Categorical 1: Highest values (Germany, Australia, South Korea)
2: High values (Netherlands, India, USA, Japan, Ecuador)
3: Medium Values (South Africa, Chile, Spain)
4: Low values (Mexico, Russia, Brazil) 5: Lowest values (Colombia, Peru)
WVS Share of Ownership H3/H4 Dichotomous 0: WOS 1: JV ORBIS
26
The data for the complicity variable covers cases of human right violations that happened in South America, Africa and Asia. This data is therefore only based on human right violations that were committed in these areas. Furthermore, the database on the human right violations covers the extractive and the textile industry. For this thesis only the data from the extractive industry will be used. This results in a reduction from 1247 cases to 963 cases. For the first two hypotheses, I will look at the influence of home country factors and therefore the cases that have multiple owners (JV’s) are removed for the analysis of the first two hypothesis. The cases that concern a joint venture are used for the third and fourth hypothesis.
Another restriction on the CHRD data is the time span, as all the cases that are from before 2000 are removed. This is done, because the data on business ethics and national values will probably be a little less representative with cases older than 15 years. Using the year 2000 as a decision point is justified as business ethics and national values take a lot of time to change (Kotter, 2008). Looking at business ethics we see that this particularly developed in the 80’s and 90’s, but that by the start of the new millennium its importance was already acknowledged (Kilcullen and Kooistra, 1999). Also, the role of businesses concerning human rights was already clear by that time as shown by the opening quote of Kofi Annan (1999). It is furthermore noted that the survey for the business ethics was taken in 2015, but had no restriction as to when the misconduct had been observed. As a vast majority of the respondents have a work experience of at least 15 years, the 14 years decision point seems justified. Concerning national values Inglehart and Baker (2000) argue that core values are strongly embedded in historical and societal context and will therefore not change very drastically. This is also noted when looking at the same question in the WVS survey in two earlier time periods (1999-2004, 2005-2009). These results would lead into a similar distribution of the countries among the categories. The outcome of these surveys shows very consistent data compared to the dataset used of the time period (2014). Only the dataset from the time period of
2010-27
2014 is used, because this dataset consists of more countries that are relevant to this study than the previous ones.
For the first hypothesis, I only look at cases of WOS that have been committed in one of the 13 countries that are discussed in the GBES survey. After removing the cases from before 2000 this dataset ended up with 429 cases.
For the second variable another dataset is used, since the countries that are available from the World Value Survey cover different countries than the countries available in the GBES dataset. For this survey 17 countries are selected that represent 449 WOS cases in the CHRD database. Due to the availability of the data for the different data sources that are used for hypothesis 1 and 2, the different hypotheses are based on different datasets with different home countries.
The third and fourth hypotheses look at the home countries of MNEs participating in joint ventures that are in the CHRD database. Looking at the home countries of the companies involved in joint ventures it became clear that almost all the home countries involved are from countries that are performing strong in both the surveys used for H1 and H2, it is therefore that only the two best performing categories of both the surveys are used in order to have the highest N and to reduce bias. For hypothesis 3 I will only use companies originating from the two categories with the lowest observed misconduct and for hypothesis 4 I will only look at the two categories that show the highest values. For hypothesis 3 the WOS and JV’s of the top 6 strong performing countries that have the lowest in the amount of observed misconduct are used. Resulting in 292 cases for hypothesis 3. For hypothesis 4 the WOS and JV’s are based on the 8 best performing countries that have the highest values for human rights. Resulting in 247 cases for hypothesis 4.
28 Table 2 Sample Description
D. Methodology
Since both the dependent variable and independent variables are categorical the most logical and most effective way, according to Field (2014), is to use the Pearson’s chi-squared for hypothesis 1 and 2. The Pearson’s chi-squared is a test that looks at how well the observed frequencies of the distinct categories fit with the expected frequencies if these categories were independent. When the observed frequencies are different than the expected frequencies and therefore do not fit the model, it means that the observed data is dependent (Field, 2014). After identifying whether or not there is an association it is also important to identify in which Level of
complicity
Share of Ownership
Direct Indirect Supporting violating government H1: GBES 13 countries (Brazil, China,
France, Germany, India, Italy, Japan, Mexico, Russia, South Korea, Spain, USA, UK)
WOS 318 93 18
H2: WVS 17 countries (Argentina, Australia, Brazil, Chile, Colombia, Ecuador, Germany, India, Japan, Mexico, the Netherlands, Peru, Russia, South Africa, South Korea, Spain, USA)
WOS 351 85 13
H3: Share of Ownership
6 Highest Ranked (Germany, Japan, Spain, South Korea, US, UK)
WOS JV 142 59 67 10 12 8 H4: Share of Ownership 8 Highest Ranked (Germany, Australia, South Korea, Netherlands, India, Japan, Ecuador, USA)
WOS JV 125 51 53 8 2 2
29
direction and how strong the specific association is. To figure out in which direction there is an association it is a good idea to look at the standardized residuals in the contingency table that also come up in the SPSS output if requested. The standardized residuals furthermore function as a z-score and tells which specific categories have a significant association (Field, 2014). To identify the actual effect size of the association many people use the Cramérs V test, Field (2014) however argues that another more useful way to measure the effect size by using the Odds ratio. The Odds ratio can however only be performed on 2 x 2 contingency tables. In order to still be able to use the Odds Ratio I have decided to create a 2 x 2 table by combining some categories in a logical way.
For hypothesis 3 and 4 the Pearson’s chi-squared can also be used as the dataset consists of the dependent categorical variable and the dichotomous moderator variable consisting of “0” Wholly owned subsidiaries and “1” joint ventures. The direction and size of the association between these two variables will be tested in a similar way as described for hypotheses 1 and 2.
30
V. Results A. Pre-tests
The first step was to test whether the datasets for this study are normally distributed. As mentioned before the datasets turned out not to be normally distributed. The study therefore needs a non-parametric test. One of the most used non-parametric test is the Pearon’s chi squared test (Field, 2014). The assumptions for using the Pearson’s chi-squared test are that the data must be independent, that no more than 20% of the expected counts are less than 5 and for contingency tables with more than 2 categories all the expected counts should be higher than 1 (Field, 2014). The different sets of data are all independent, so this assumption has been met for all the variables. The second assumption that is necessary to perform the Pearson’s chi squared shows that all the different tests are below the 20% maximum of expected frequencies below 5. Furthermore, there are no expected counts that are below 1. This means that for all the different hypotheses the Pearson’s chi-squared is an appropriate test for this data.
Assumption 1: Independence
Assumption 2:
No more than 20% of expected counts below 5 Assumption 3: No expected counts below 1 H1 Business Ethics Independent 16.7% YES H2 National Values Independent 20% YES H3 Share of ownership Independent 16.7% YES H4 Share of ownership Independent 16.7% YES
31
B. Analysis
To test hypothesis 1, the relationship between levels of involvement in human rights and home country business ethics have to be determined. To do this a cross tabulation has been performed. The outcome of the test shows that the null hypothesis, that the variables are independent, can be rejected. The test shows that there is a significant association between the level of involvement and the home country business ethics (6) = 29.054, p < .001. This shows that there is a very strong significance for this association. To find out whether there is a positive or negative association I have looked at the standardized residuals. This shows that countries with lower observed misconduct are significantly more likely to be indirectly involved in a human right violation. To find out the size of the effect I look at the Cramér’s V and an odds ratio is performed. The Cramér’s V shows a significant weak effect size φc =.184, p < .001. To use the odds ratio, I have combined categories from both the variables. On the one side, I combined “indirect involvement” and “supporting a violating government”, since they are both more indirect ways of involvement. And on the other side I combine the two categories with the higher amounts of observed misconduct and the two categories with the lower amounts of observed misconduct. Based on this odds ratio the odds of a company being indirectly involved in a human right violation is 2.71 times higher when they are from a country where there is lower observed misconduct.
H1 Value Significance
Pearson’s chi-squared
29.054 (6) p < .001
Cramér’s V .184 P <.001
Table 2 Hypothesis 1 Results
To test hypothesis 2, the relationship between the level of involvement and the national values in the home countries of the firms have to be determined. The cross tabulation shows that the null hypothesis can be discarded because the test shows there is a significant association
32
between the two variables (8) = 29.462, p < .001. This means that the level of involvement in a human right violation is definitely associated with the degree of values concerning human rights in the home country. The standardized residuals show that countries that have higher values concerning human rights are more likely to be indirectly involved. The standardized residuals furthermore tell us that there are two associations between variables that are significant. It shows that Looking at the effect size it is noted that the Cramér’s V shows a weak significant effect size φc =.181, p < .001. To use the odds ratio, I again combine different categories. On the one side, I combined “indirect involvement” and “supporting a violating government” again. On the other side the two categories with the highest values and the two categories with the lowest values are combined. The odds ratio showed that the odds that a company is indirectly involved is 2.15 times higher when the company is from a country that has higher values concerning human rights.
H2 Value Significance
Pearson’s chi-squared
29.462 (8) p < .001
Cramér’s V .181 P <.001
Table 3 Hypothesis 2 Results
For hypothesis 3, I test whether the share of ownership can act as a moderating value that has an influence on the level of involvement of a company. In this test, the difference between WOS and JV’s is tested in relation to the level of involvement. For this test, I only look at WOS and JV’s that are from the two lowest levels of observed misconduct. At first again a cross tabulation was conducted which performed the Pearson’s chi-squared. This showed that the null hypothesis meaning that the different variables are independent can be rejected. The Pearson’s chi-squared test shows that there is a significant association between
33
between the level of involvement and the share of ownership. To find out in which way there is an association I look at the standardized residuals. This shows me that JV’s are significantly less likely to be indirectly involved. The Cramér’s V shows that there is weak significant effect size φc =.175, p < .05. For the odds ratio, the observation of “indirect involvement” and “supporting a violating government” are again combined. Resulting in a 2 x 2 contingency table. The odds ratio showed that the odds that a company is directly involved is 2.31 time more likely when it is works through a joint venture.
H3 Value Significance
Pearson’s chi-squared
8.907 (2) p < .05
Cramér’s V .175 P <.05
Table 4 Hypothesis 3 Results
The test conducted for hypothesis 4 is very similar to the test of hypothesis 3. The difference for this test is that I only look at WOS’s and JV’s from the two highest categories in national values. Running the Pearson’s chi-squared shows a significant association between the level of involvement and the share of ownership (2) = 6.253, p < .05. To find out whether this is a positive or negative association I look again at the standardized residuals. These show that JV’s are more likely to be directly involved compared to WOS. Looking at the Cramér’s V it shows that there is a weak significant effect size between the variables φc =.159, p < .05. With a similar set up as used in hypothesis 3 the odds ratio was calculated. The odds ratio shows that the odds for a company to be directly involved is 2.49 time more likely if this company operates through a joint venture.
H4 Value Significance
Pearson’s chi-squared
6.253 (2) p < .05
Cramér’s V .159 P <.05
34
VI. Discussion A. Summary and Findings
Looking at the database provided by the CHRD project you see that over a period of 14 years around 1200 cases have been reported, where businesses are in one way or another involved in human right violations. Taking into account that not even all the cases are reported, this is in line with the argument by Ruggie (2013) that businesses are often involved in human rights and that they could do more to prevent human right violations. Kobrin (2009) argues that companies are most often indirectly involved in human right violations. The data on the different complicity levels, however shows that companies are by far more often directly involved. It is furthermore noted that a far majority of the cases in the database happened in a country outside of the home country. Which stresses the importance of the relation between parent firms and its subsidiaries. The weak effect sizes of the different hypotheses, that are shown with the Cramér’s V, suggest that there are more factors that play a role in the involvement of companies in human rights.
For the first hypothesis on business ethics, the focus is on the amount of observed misconduct. Observed misconduct is explained by the survey as the violation of laws or ethical conducts on all levels in an industry. The data shows that misconduct is observed in every country that is covered in the survey and therefore also in countries that have strong regulations. This could be in line with the literature arguing that codes of ethical conduct lack effectivity (McNeil & Li, 2006). The test of hypothesis 1 has shown that companies from home countries with stronger business ethics will be more likely to be indirectly involved in human right violations. This is in line with the literature arguing that business ethics can be passed through to the subsidiary (Scholten and Dam, 2007). It however also shows that having a strong business ethic environment in the parent firm does not necessarily mean that the company or its subsidiaries are less likely to be involved in human right violations. The only conclusion that
