Conflicts between MNEs and indigenous communities : the impact of host country development and community autonomy on conflict resolution within the Commonwealth of Nations

Conflicts between MNEs and Indigenous

Communities

The impact of host country development and community autonomy on conflict resolution within the Commonwealth of Nations.

Master Thesis University of Amsterdam MSc. Business Administration:

International Management Name: Marysé Jansen Student number:10246517

Date: 23-06-2017

Thesis Supervisor: Dr. Ilir Haxhi Second reader: Drs. Erik Dirksen

State of Originality

This document is written by Marysé Jansen, I declare to take full responsibility for the content of this document. I declare that the text presented in this thesis 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 of the University of Amsterdam is responsible solely for the supervision of completion of the work, not for the content

Abstract

Globalization facilitated a fast worldwide geographic expansion of Multinational Enterprises (MNE) operations, which affected numerous indigenous communities. Often the MNEs’ interaction with these communities resulted in conflicts involving firms operating mainly in resource extracting industries in countries with lower HDI. By focusing mainly on isolated cases of conflicts, previous studies have omitted the multi-level aspect of the resolution and dynamics of these conflicts. Therefore, in this study, by taking a multilevel (country,

community and MNE) approach, we explore the conflict resolution focusing particularly on the host country level of development and the community autonomy. Furthermore, we analyse to what extent the sustainability focus of the MNE moderates these relationships. For a sample of 237 cases within the Commonwealth of Nations, we analysed the factors

influencing conflict resolution, which is measured by the degree of violence and conflict severity. We argue that both the community autonomy and the host country HDI positively influence conflict resolution and that this relation will be positively influenced by

sustainability focus of the MNE. Our analyses show three main findings. First, a higher host country HDI can alleviate the length of conflict, second, a higher degree of community autonomy diminishes the degree of violence, and finally, the MNE sustainability focus plays an important role as a moderator in the relationship between country HDI as well as

community autonomy, and conflict resolution. Our contribution is twofold: First, while previous studies analysed conflict resolution by examining isolated cases, we use a

quantitative approach providing a comprehensive and geographically generalizable theory. Second, we analysed the unexplored moderating affect of sustainability focus. Consequently, our findings provide valuable insights for managers in the field of international management, by extending current knowledge on conflict resolution.

________________________________________________________________________ Keywords: Multinational Enterprise; Indigenous peoples; Indigenous Community; HDI; Conflict; Conflict resolution; Sustainability Focus

TABLE OF CONTENT

1.

INTRODUCTION

7

2. LITERATURE

14

2.1 Indigenous peoples 14

2.2.1 Community Autonomy 14

2.2 Conflicts Between MNEs and Indigenous peoples 16 2.3 National Context and Conflict Resolution 18 2.4 Sustainable Development 20

3.THEORETICAL FRAMEWORK.

23

3.1 Mechanisms Influencing Conflict Resolution 23

3.2 Host country HDI 25

3.2 Community Autonomy 26

3.3 MNEs sustainable focus 27 3.4 Conceptual Framework 30

4. METHODOLOGY

31

4.1 Sample & Data Collection 31

4.2 Variables 32 4.2.1 Dependent Variables 32 4.2.2. Independent Variable 33 4.2.3. Moderator Variable 34 4.2.4 Control variables 34 4.3 Methods 35

5. RESULTS & ANALYSIS

38

5.1 Descriptive Statistics 38 5.2 Correlation Analysis 40 5.3 Multicollinearity Test 42 5.4 Regression Analysis 43 Control Variables 44 Direct effect 45 Moderator Effect 46

6.

DISCUSSION

51

6.5 Future Research 57

7. CONCLUSION

59

8. REFERENCES

61

List of Tables and Figures

Figure 1: Conceptual Framework_________________________________________________________________________________30 Table 1: Summary of Regression analysis ________________________________________________________________________ 37 Table 2: Descriptive Statistics___________________________________________________ Error! Bookmark not defined.0 Table 3: Correlations ____________________________________________________________ Error! Bookmark not defined.1 Table 4: Multicollinearity test: degree of violence _______________________________________________________________ 43 Table 5: Multicollinearity test: length of conflict _______________________________ Error! Bookmark not defined. Table 6 Regression Analysis: degree of violence _________________________________________________________________ 49 Table 7: Regression Analysis: degree of violence ________________________________________________________________ 50

1. Introduction

In the past two decades, there are several high profile cases of ethnic conflict caused by resource-extracting activities of major MNEs (Hook & Ganguly, 2000; Calvano, 2008). Natural resource development on indigenous land is typically characterized by distinct power imbalances, and many communities lack territorial rights. This results in the increase in conflicts that feature hostile interactions between indigenous peoples, MNEs and host country governments(Bruijn & Whiteman, 2010;Whiteman 2009). Indigenous communities can be resistant to any form of drilling or fundamental changes to the landscape, because of their special relation with the environment (Whiteman, 2009). It is crucial for both MNEs and the communities to find factors that could enhance conflict resolution. This study uses a neutral researcher perspective, which includes a multilevel analysis (country, community and MNE), to create a cohesive comprehension of conflict resolution.

Conflicts between MNEs and indigenous peoples have been on the research agenda for some time now. The first actor important for conflict resolution is the host-county

government. From a historical perspective, Acuña (2015) perceives the local context as root of indigenous conflict, since colonial patterns influence the socio-economic position of communities today. Colonial legacies, marginalization and unrecognized land rights enhance the current deficiency of the community in control over decisions, procedures, or forms of interaction that focused on the natural resource development. Many tribes are trying to have their traditional land rights acknowledged by national or international legal frameworks. Various international law conventions and a few state policies recognize the traditional land rights of indigenous peoples. However, both in developed and developing countries, there are only few examples of meaningful consultation of a regulatory framework protecting

Furthermore, previous literature identifies two ways in which the host country level of development negatively influences conflict resolution. First, developing countries might depend more on Foreign Direct Investment (FDI), and therefore grant the local governments free rein by favouring the demands of the MNE over the protection of the rights of the indigenous community. Second, lower levels of development are often related to weaker institutional and governmental structures. Lynch (2012) argues that conflict between indigenous communities and MNEs symbolize the absence of regulations that focus on development. In case low development is accompanied by low governance effectiveness, the community can turn to the MNE to execute government tasks, which causes the relationship to be altered (Esan, 2005).

In short, previous studies focusing on conflict situations predominantly analyse dynamics of conflict resolution form a qualitative design within a single-case perspective (Calvano, 2008). Although these studies give valuable insights on specific cases, they fail to present a multi-level perspective on these phenomena. Even though previous literature focussed on the role on the host country in conflict resolution, this factor is never analysed as a predictor for conflict. Consequently this study attempts to fill this gap by analysing the influence of host country development and community autonomy on conflict resolution from a quantitative perspective. Thus, by extending on previous research, in this current study we explore the extent to which the host-county HDI affects conflict resolution between

multinational firms and indigenous communities. Considering the relevance of the host country HDI, we address our first research question:

RQ1: To what extent does the host country development level affect the severity of the conflict between MNEs and indigenous peoples?

The purpose is to identify ways in which the host country HDI directly influences conflict severity. This knowledge contributes to FDI and location decision making. When we prove that the host country level of development positively influences conflict resolution, managers know that it is more beneficial to locate in more developed countries. Previous literature identified the host country context as an important factor for resolving conflicts. However, the precise role is still unexplained. With this question we attempt to contribute current literature, by examining the direct effect of the host country level of development on conflict resolution.

The second actor involved in the conflict is the indigenous community. Ballard and Banks (2003) highlight the neglect of indigenous rights and the absence of recognition as main cause of conflict. The lack of recognition of indigenous communities illustrates the difficult relationship between the host country and the community. A threatened community identity enhances the probability of conflict (Bruijn & Whiteman, 2010). Calvano (2008), argues that when the community requires autonomy, this diminishes the power inequality between the community and the MNE. This in turn, makes room for a more harmonious relationship between the two actors, which results in enhanced conflict resolution.

The General Assembly of the United Nations introduced the “United Nations Declaration on the Rights of Indigenous peoples” (UNDRIP), which include: right to self-determination and autonomy; right to sovereignty over their land and resources and right to language and culture. The United Nations argue that the recognition and protection of indigenous peoples rights are a necessary condition for sustainable development (Belier & Préaud 2011). However, to our knowledge, there is no previous research that explores this relationship. We argue that the degree of autonomy influences the relation between the community and the firm, and thereby influence conflict resolution. By extending previous research, we thus explore how autonomy affects conflict resolution between the community and the MNE. Therefore, we present the second research question:

RQ 2: To what extent does the autonomy of an indigenous community affect the conflict resolution between an Indigenous community and a MNE?

This question is essential in order to explain whether the communities with a lower degree of autonomy are involved in more severe conflicts. This knowledge is valuable for MNEs, especially for their FDI strategies. If the results prove that communities with a higher degree of development are involved in less violent conflicts, managers know that it is more beneficial to locate in these areas. Additionally, this knowledge is useful for the community, the local government and third parties such as NGO’s that are involved in conflict resolution.

The third actor necessary for conflict resolution is the MNE, which is introduced in our study by focussing on their sustainability focus. The sustainability focus incorporates strategies of sustainable development into the business strategy, and comprehends a positive socioeconomic change that does not undermine the social and ecological systems were communities and societies depend on. Sustainable development meets the needs of the present without compromising the needs for future generations. Scholars emphasize the importance for sustainable development; they assert that it is unethical to affect and pollute the lands and lives of indigenous peoples (Lertzman & Vredenburg, 2005; Moldan,

Janouskova, Hák, 2012).

Lertzman & Vredenburg (2005) argue that companies need to become ‘part of the solution’, instead of being part of the problem. O’Faircheallaigh (2013) discovered a change in the corporate policies towards Indigenous peoples. A growing number of companies within resource extracting industries seek to enter agreements with indigenous peoples before

operating on their traditional lands. Calvano (2008) argues that when MNEs invest in the relation with the indigenous community, this will reduce conflict violence. The sustainability focus is important since it regulates the relation between the MNE and the indigenous

logically enhance conflict resolution. However, the effectiveness is unexamined as a moderator in academic literature. Thus, this study will address the unexplored moderating effect of sustainability focus on conflict resolution. Thereby, we address our third research question.

RQ3: To what extent does the MNEs sustainability focus moderates the effect on conflict resolution?

This question is essential to explore if conflict resolution enhances when firms act more responible and sustainable. If the results prove that a sustainability focus is beneficial for conflict resolution, this knowledge can improve managerial decisions. This study contributes to the encouragement of MNEs to adopt a sustainable strategy, especially when they are located in less developed countries and are involved in extracting activities.

Conflicts between MNEs and indigenous peoples have negative consequences for all parties involved. Extracting activities cause pollution, contamination, and environmental spills (Lertzman & Vredenburg, 2005). For indigenous communities these conflicts sometimes lead to serious injuries and even death (Bruin& Whiteman, 2010). For MNEs, these conflicts have negative reputational and financial consequences, due to the extensive media coverage (Calvano, 2008). Especially less developed countries seem to be prone to more violent conflict. For these countries, conflict can have negative influence on their economic development (Pehn,2009). We argue that deepening our understanding on conflict resolution is valuable to all parties. If we prove host country HDI and community autonomy have a positive influence on conflict resolution, all actors will benefit from this knowledge.

The research question will be addressed by conducting a quantitative research in which 237 cases are codified and analysed, within the Commonwealth of Nations. This decision is motivated first because of their shared colonial history and second because of the variance in their development level, through which the difference in host country context is

considered. With this study, we aim to assess the extent to which the host county

development level and the community autonomy influence conflict resolution. In addition, we argue that the MNEs sustainability focus has a positively moderating effect on conflict resolution.

This study provides a contribution to the field of international management by extending on current knowledge on conflict resolutions in three ways. First we contribute to existing literature theoretically, by using a multilevel approach. Several studies examined the influence of isolated factors on conflict resolution. However, this study is distinctive in combining multiple perspectives on conflict resolution within one model. Previous literature failed to introduce multiple factors relating to conflict dynamics in one analysis. Furthermore this study is unique in examining the unexplored direct influence of the host country

development and community autonomy on conflict resolution. Additionally we extend on previous literature by measuring the effectiveness of a sustainability focus as moderating variable. Our second contribution is methodologically. Previous research failed to present relationships applicable across different geographical settings, because of their single-case research design. This study fills this gap by using a quantitative approach, which results in a compressive understanding of conflict dynamics. Lastly, form a practical point a view, our findings provide valuable insights for managers. This study contribute to the creation of better understanding of conflict dynamics. Consequently, this knowledge will contribute to improving location and FDI decision-making. The acquirement of this deeper understanding will contribute conflict prevention or a quicker resolving in the future.

In order to answer our research question, this thesis starts with a literature review on the main concepts of this study, namely: indigenous peoples, MNE conflict, host country factors and sustainability focus. Second, we develop a theoretical framework including the conceptual model and hypotheses. Subsequently, we discuss the research design, and present

and methods of analyse. The remaining chapters of the thesis are dedicated to the presentation and the discussion of the research results.

2. Literature Review

This chapter discusses the key concepts and thereby presents an outline of previous research. The first section includes the definition of indigenous communities and introduces the concept of community autonomy. Followed by the conflict dynamics between MNEs and indigenous peoples. Third, we discuss the importance of the national context for conflict resolutions. Lastly, we present the MNE characteristics by elaborating on sustainable development.

2.1 Indigenous peoples

“Many indigenous peoples live on traditional territories rich in natural resources. The United Nations estimates that there are between 350 and 500 million Indigenous People worldwide, distributed over 5,000 distinct groups” (Whiteman, 2009).

Scholars often use the term indigenous in association with: a particular place; an original or first inhabitant; a prior inhabitation; an attachment to land; historical continuity; a precedent habitation; and a sense of community. Even though there is no consensus about the correct definition for indigenous people, scholars agree that the concept concerns a special group of people, in need of special rights, including the right of self-determination (Esan, 2005). There is no universal definition of community in social science, but scholars across disciplines agree that it includes dimensions of social interaction, geography, and identity (Calvano, 2008).

2.2.1 Community Autonomy

Traditionally, many governments perceive indigenous peoples as second-class citizens (Bellier & Préaud 2012).

Around the world, there are many indigenous communities striving for autonomy and self-determination. Autonomy does not imply succession from the nation-state, it rather it includes the broadening of rights within the existing structures of the state (Stephen, 1997). The concept of autonomy can imply a situation of total sovereignty and political

independence. However, in the case of indigenous communities the term does not present total independence, but rather a continuing political and economic connection with a larger state. Arthur (2001) emphasizes that autonomy is not an absolute, but a relative status, which offers the possibility of power-sharing and internal self-governance within the institutions of the state. Esan (2005) identified autonomy as a right to self-determination; a way of life with control over their own social, cultural, economic and political development. Bellier and Préaud (2012), recognize a progress in recognition of the communities in recent years.

Historically, there is a difficult relation between the government and indigenous peoples. Since colonial expansion, the two actors are in conflict over indigenous rights, recognition and territorial ownership. Colonial power also executed the dispossession and exclusion of

indigenous peoples, which led to today’s reality where many indigenous peoples are installed in situations of entrenched poverty and socio-political exclusion (Bellier & Préaud, 2012). Acuña (2014) argues that socio-environmental conflicts that involve indigenous peoples are not only a result of ill-designed policies; they emerged because of the permanence of a colonial pattern of domination that denies indigenous ontologies. The scholar states that the dominance of western legal procedures in the governance of natural resource development is a product of colonialism. Colonial legacies, marginalization and unrecognized land rights enhance the fact that indigenous peoples do not have equal control over decisions,

procedures, or forms of interaction that focuses on the natural resource development.

De-colonial perspectives emphasize that even though formal colonialism has ended in legal and political terms, a colonial mind-set is still part of our societies. This is apparent in

power inequality, colonial ontology, and epistemology present in contemporary society. Consequently, regulative aspects of society, for example the organization of economic and social relations, are still built upon colonial ideas (Acuña, 2015; Bellier & Préaud, 2012). This is apparent in the violation of the human rights of indigenous people, for example their right to autonomy. Recognition is granted to indigenous communities in some cases, but these rights are often perceived as inferior to any other type of property titles within domestic law; therefore, they are subject to legal insecurity and can be overridden by other rights or acts of states (Bellier & Préaud, 2012). Recognized communities of indigenous people obtained a special status under international law, which grants them protection of their fundamental human rights (Esan, 2005).

2.2 Conflicts Between MNEs and Indigenous peoples

Conflicts between communities and companies have evolved in the last decades, with both environmental and economic grievances as a result (Acuña, 2015). Calvano (2008) argues that the resource-rich lands of indigenous communities are the new battleground on which MNE activities are contested, with enormous consequences of both financial and reputational nature. MNE activities, in particular within the extractive sector, can cause friction between different actors. Since indigenous peoples rely heavily on their land and resources, they are vulnerable to environmental damage from extraction, which can threaten viable livelihoods They also lack valid political influence due to discrimination and social disadvantage (O’Faircheallaigh 2012; Acuña, 2014; Hook & Ganguly, 2000).

Differences in ethical principles, reasoning or behaviour between two or more parties can result in cross-cultural ethical conflict. Communities might view that MNEs impose their will on them, contrasting the MNEs that may view the communities as ungrateful for the benefits of economic development. This perception gap is facilitated by the contrasting

ideology of both actors regarding development (Calvano, 2008). Indigenous peoples perceive their ancestral land to be sacred and view oil exploitation as a profane act, which defiles their land, and consequently the community does everything within their power to prevent oil exploration activities. Indigenous peoples express a different political ontology, based on a different conceptualization of the relation between human beings and the natural

environment. This differing worldview makes consequent legal and political arrangements a necessity (Esan, 2005). The cultural context may influence the organizational legitimacy of the MNE in the eyes of the indigenous community. Organizational legitimacy is defined as the acceptance of an organization by its milieu. The lack of organizational legitimacy enhances conflict (ibid.).

Indigenous communities may feel that their homeland is being exploited for benefits of outsiders, since they experience little financial and economic advantage. MNE activities may result in the dispossession of land, health, habitat and way of life of for indigenous communities as consequences of conflict (Acuña, 2014). Additionally, the environmental damage resulting from extracting activities can cause conflicts over the diminished

environmental resources within communities, or can stimulate large-scale migration from the affected location to richer areas. Such situations would be most destructive when the affected group draws its living directly from the land (Hook & Ganguly, 2000). MNE activities can desecrate indigenous sacred sites, which in turn can lead to conflict (Whiteman, 2009)

There are both benefits and costs related to MNE activities. On the positive side, extracting activities can bring valuable infrastructure development. This includes

transportation systems, electricity, roads, as well as improvement in education and medical facilities. On the contrary, resource development can result in significant negative change of indigenous traditional lands. Common negative effects are: the modification of community relations and indigenous social structure, resulting in a trend towards individualism within communal societies; a shift to a more class-based or elitist system; the alteration of traditional

power structures and hierarchies; and increases in violence and human rights abuse. The ecological degradation can therefore damage cultural identity (Whiteman, 2009).

2.3 National Context and Conflict Resolution

Globalization changed the relation between states and corporations across the globe. Scholars agree over the importance of the local context for conflict resolution (Bellier & Préaud, 2012; Acuna, 2015). North (1991), makes a distinction between formal and informal institutions, which together structure the economic, political and social interaction. Therefore, the state has the power to influence the activities of the MNE on local grounds. Governments have the power to regulate corporate activities and to apply sanctions in case of non-compliance. State agencies in both developed and developing countries occasionally ignore indigenous rights and interests in favour of the wishes of the MNE because of their dependency on FDI (O’Faircheallaigh 2012).

Gervey & Newell (2005) state that a healthy relationship between the state is

necessary for the recognition of the community and the protection of their indigenous rights. Politically marginalized communities are often not supported by governments, which instead focus on building more powerful coalitions and constituencies, directed at the protection of the MNEs. The relation between the community and state is important in determining the level of state protection the community receives. The policies and legal framework of the government need to protect and promote the rights and responsibilities of both communities and firms. Nonetheless, the realization of these rights dependents on the willingness and the ability of the state to implement this. On the grounds of lacking technical skills and literacy, indigenous peoples are often excluded from negotiations, which reduces their ability to engage in meaningful dialogue with corporations (Garvey & Newell 2005). Preuss and Brown (2012) state that the reliance on host country duties alone may not be sufficient to

broadly protect human rights. When the host country is highly dependent on a specific

industry, a lot of effort is made not to jeopardise FDI, which illustrates the access of MNEs in decision-making processes of local governments. However, Bellier and Préaud (2012) argue that sometimes, indigenous peoples might be better protected by the state in developing countries, in development countries, the international community is involved in monitoring the local conditions.

Countries with a low level of development and low quality institutional and regularity structures are referred to as fragile states. The fragility of a state is measured with a score on the dimensions of effectiveness and legitimacy. In the eyes of the citizens, fragile states have ineffective or illegitimate governments. Low developed countries are weaker in enforcing regulation on MNEs (Bebbington, 2008). Pehn (2009) notes that countries in conflict are fragile, but not all fragile states are necessarily in conflict. Furthermore, Ballard and Banks (2003) argue that weak states are less capable of providing communities with social services such as, justice, education, health and security.

Situations of violent conflicts can create risks for firms, to which they might respond with exciting the country and moving to safer locations. Esan (2005) shows that resource-rich countries as Angola, Indonesia, Nigeria and Colombia experienced a decline in FDI in the extractive sector due to various conflicts, political unrest and ethnic clashes. Price

Waterhouse Coopers report, issued in 2001, shows that conflicts deter investment. In this research, 78% of the mining firms surveyed, had withdrawn from investment or refrained from investing because of violent conflicts or political instability (Esan, 2005). Pehn (2009) describes violent conflicts as “development in reverse’’ because of their disastrous

consequences: conflicts can ruin a country’s economic and social systems; curtail

productivity; destabilize neighbouring countries through negative spill-over effects; and they enforce the destruction of productive capacities.

Pehn (2009) emphasizes that the main goal of firms operating in conflict zones is profit and not the reduction of poverty, conflict, or fragility. Consequently, sometimes these firms exacerbate conflict and fragility by doing business with oppressive states or excluding local communities. The activities in resource extracting industries are known for causing environmental spills, pollution and contamination, while also affecting traditional livelihoods of indigenous communities (Lertzman & Vredenburg, 2005).

According to Kolk and Tulder (2006), the UNCTAD report shows that

overdependence on MNEs has its drawbacks. FDI is concentrated in a few developing countries, and this can crowd out local competitors that cannot meet the MNE standards. Furthermore, the OECD Guidelines for Poverty Reduction indicate that the political and economic policy, combined with the lack of interest in long-term development are the main corporate influence on poverty. Local conditions need to be taken into consideration when poverty policies are composed. International organizations emphasise the importance of collaboration between all organization involved in eradicating poverty. The state is necessary for poverty reduction interventions, since it is the provider of the critical public goods such as national security. Therefore, external parties should work together with the government to encourage accountability rather than to create an alternative system that could undermine the state (Pehn, 2009).

2.4 Sustainable Development

Sustainable development is the positive socioeconomic change that does not undermine the social and ecological systems upon which communities and societies depend. Natural resources should be used in a manner that does not eliminate or degrade them or otherwise diminish their usefulness for future generations (Lertzman & Vredenburg 2005; Moldan, Janouskova, Hak 2009). Increasing population growth, deep worldwide poverty, and limited

resources within an already threatened ecosphere, come together in a difficult ethical

dilemma. On the one hand, significant economic activity is necessary to meet the basic needs of an increasing human population, on the other hand, economic growth itself is responsible for much of the environmental degradation (DesJardines, 1998).

The concept of sustainable development encompasses three fundamental dimensions: economic, environmental, and social development, which are complementary and

interrelated. Economic sustainability focusses on the implementation of sustainability

principles and safeguards an optimal amount of general capital for future generations (Ciegis, Ramanauskiene & Martinkus, 2015). Second, environmental sustainability, seeks to improve human welfare by protecting the sources of raw materials used for human needs in order to prevent harm to the environment (Moldan, Janouskova, Hak, 2009). The vitality of sub-systems is essential in the critical view of global stability of the total ecosystem. The importance of preserving biological variety is emphasized, in order to secure a balanced nature, flexibility of ecosystems at a global level and their ability to adapt to changes in biosphere (Ciegis, Ramanauskiene & Martinkus, 2015). Specific cross-border land use

impacts such as noise, smell, increased traffic and ground water pollution can lead to conflict. On the contrary, successful land use coordination can result in joint economic development opportunities and more liveable communities. (Nelles & Alcantara, 2014). Third, the social-cultural dimension includes the well-being of the community. This dimension argues for a responsible strategy towards communities whilst performing FDI (Ciegis, Ramanauskiene & Martinkus, 2015).

Bellier and Preaud (2012) emphasize that indigenous peoples are not opposed to development, but rather seek to utilize their right to self-determination within development initiatives. The political and economic fragility of the community often obscures and constrains their ability to oppose major development projects. The community wants to define its own criteria of what constitutes sustainable development. Furthermore, scholars

emphasize that indigenous peoples’ attitudes regarding land rights and sustainability are far from homogenous, building on different conceptions of autonomy and self-determination

Hamilton& Knouse (2001) argue that the fact that MNEs can afford the litigation costs of toxic waste emissions or are able to pay for the cooperation of a foreign government in depriving indigenous people of their livelihood, does not make these actions ethical. Lertzman and Vredenburg (2005) add that it is immoral to sacrifice the viability of

indigenous cultures for the purpose of industrial resource extraction. Furthermore, Acuña (2014) states that it is only right to engage with indigenous peoples when this is consistent with their wishes and needs.

Reasons for MNEs to adopt a sustainability focus is because they want to do good or because of isomorphic pressures, created by a new industry standard, increase shareholders’ investment (DiMaggio & Powell, 1983). Moldan, Janouskova and Hak (2012) argue that the adoption of a sustainability focus is beneficial to firm performance. A widely used method is to acquire a membership of an index series that is designed to measure the performance of companies demonstrating strong Environmental, Social and Governance (ESG) practices.

The sustainability focus should include two things: it should address the entire range of environmental and ecological issues affected by business decisions in a way that might actually turn the tide of environmental and ecological deterioration, and it should be capable of influencing business policy (Desjardines 1998). Kolk and Tulder (2006) raise the question on why not all MNEs execute active policies regarding poverty alleviation.

3.Theoretical Framework.

In this chapter we elaborate on conflict dynamics. First, we present the mechanisms that influence conflict resolution. Second we elaborate on the effect of host country development, community autonomy and sustainability on conflict resolution, followed by the proposed hypotheses.

3.1 Mechanisms Influencing Conflict Resolution

Even though there is extensive literature on isolated factors relating to conflicts between multinational firms and indigenous communities, there is a lack of analysis on conflict

resolution (Calvano, 2008). A deeper understanding of conflict resolution is beneficial for the all actors. Bellier and Préaud (2012) argue that the recognition of indigenous peoples for international law impacts the whole global system, since it affects the relationship between multiple economic and political powers.

We identified three factors essential for conflict resolution, namely: community characteristics, MNE characteristics and the local context. We argue that we can only comprehend an cohesive understanding of conflict resolution, by introducing all three perspectives in one construct. Building on current literature, we argue that both the host country development level and community autonomy enhance conflict resolution. In addition, we state that sustainable focus positively affects this relationship.

Countries with a lower level of development are more prone to violent conflicts because they dependent more on their natural resources. Klok & Lenfant (2013), argue that the slow economic development of the host country can result in inequality and unequal distribution of wealth, which in turn can lead to conflict. Bebbington (2008) states that the abundance of natural resources may generate a series of economic and political distortions,

which could undermine the contributions to the development of the extractive sector. State fragility and a low development level can enhance conflict since there is a higher probability that local governments favours industrial development over the protection of indigenous communities, despite the violation of human rights and deprivation of the environment.

The second factor influencing conflict resolution is community autonomy. The topic of community autonomy is highly relevant because of the adoption of the UNDRIP by the General Assembly of the United Nations in 2007, which enhanced the political capacity of indigenous peoples. This declaration empowers indigenous peoples and provides opportunity for effective change to the dire social, economic and political conditions in national spaces. Indigenous rights include the right to self-determination and autonomy. However, the international enforcement of the UNDRIP is lacking, and there is a wide variety in the position of indigenous peoples across different states. The increased willingness of MNEs, states, and international institutions to recognize indigenous rights reflect the growing political capacity of the indigenous community; in particular the ability of indigenous landowners to delay disrupt and sometimes force the abandonment of planned projects or extracting activities (O’Faircheallaigh, 2012; Bellier & Préaud, 2012). Calvano (2008) argues that the recognition of the community is important to reduce the inequality between the MNE and the indigenous group.

Lastly, we propose that a sustainable focus affects the influence of host country HDI and community autonomy on conflict resolution. MNEs can drain the resources of a region, channelling them back to the home country without returning benefits to the local community or economy. MNEs seldom reinvest the profits of extracting natural resources locally. Only few communities receive revenues for these activities, while bearing the economic,

environmental, and social costs that are generated by these activities (O’Faircheallaigh 2013; Acuña, 2015; Hook & Ganguly, 2000). Kolk and Tulder (2006) note that certain firms in

certain industries may pay attention to a relatively high number of poverty related issues, because of the reputational risk caused by the exposure to public scrutiny.

3.2 Host country HDI

As previously mentioned, the host country is important factor for determining the context and the conditions of the conflict. In this paragraph, we elaborate on two ways in which the host country development level may enhance conflict.

The first host country factor that diminishes conflict resolution is the local

government when they stimulate economic development over the protection of indigenous peoples, by granting permission to MNEs to operate in areas that are inhabited by indigenous communities. Developing countries try to attract MNEs since the economic growth induced by the associated growth in the industrial sector could help with poverty alleviation (Kolk & Tulder, 2006). Furthermore, states dependent on loans of institutions such as the World Bank or IMF, which impose conditions that might force the government to focus on industrial expansion at the expense of social and environmental safeguards (Garvey & Newell, 2005).

The second host country factor that can spur conflict is state fragility. When governments fail to extern effective control over their own territories, they are known as fragile states. States who lack government effectiveness are more prone to conflicts. Oetzel, Getz and Ladek (2007) accuse MNEs operating in fragile states of creating or contributing to economic, political or social conditions that foster violence. This illustrates an interaction between conflict and poverty; experts agree that conflict can contribute or reinforce poverty, due to the destruction of productive capacities (Pehn, 2009).

Oetzel, Kathleen and Ladek (2007) argue that if the institutional conditions are not right, minerals should not be extracted. In many developing countries, a high correlation is found between an abundance of natural resources and violent conflict. Garvey and Newell

(2005) state that host country governments frequently offer a range of concessions to companies and organizations, in order to attract more FDI. Furthermore, empirical evidence shows that without governmental interfere, a conflict lasts longer (Calvano, 2008;

Humphreys, 2005). Based on these findings, we can predict:

H1a: A high host country HDI has a negative influence on the degree of violence (the conflicts will be less violent)

H1b: A high host country HDI has a negative influence on the length of conflict (the conflicts be shorter)

3.2 Community Autonomy

Indigenous identity can be used as powerful resource to buffer against commercial pressure or external threat. Eriksen (1993) highlights the process of ‘ethnic revitalization’ within a group that feels threatened. In that process, indigenous peoples emphasize their latent cultural similarities. In these occasions, people reflect upon and objectify their way of life as a

tradition, and cultural identity is used as a resource that can be revitalized, to achieve

collective aims to resist cultural change. Bellier and Préaud (2012) emphasize that indigenous peoples actively use the discourse of local indigenous identities within transnational political context, as an effective strategy to gain land rights and to resist unwanted development. Additionally they argue that autonomy is a necessary condition for sustainable development. This illustrates the importance of community autonomy for conflict resolution.

According to the stakeholder salience model, the influence that different actors or stakeholders have on corporate decision making depend on three factors, namely: power, legitimacy, and urgency. When stakeholders possess all three, they will receive the most managerial attention. However, historically indigenous communities possessed relative little

power in comparison to other stakeholders. As a result, the capability of communities to affect MNE decision-making in their favour has been slight to non-existent (Calvano, 2008).

When the community receives autonomy, their stakeholder power increases, which in turn improves the relationship between the community and the MNE. We argue that

recognition enhances the stakeholder legitimacy of the community, contributing to a better relationship with the MNE and, diminishes the violation of indigenous rights and thus positively influences conflict resolution. Based on these findings, we can predict:

H3a: A high degree of autonomy influences the degree of violence negatively (the conflicts will be less violent)

H3b: A high degree of autonomy influences the length of conflict negatively (the conflicts will be shorter)

3.3 MNEs sustainable focus

As previously mentioned, the sustainability focus of an MNE is an important moderator in the relationship between indigenous communities and the state, and thus influences conflict resolution. Worldwide, indigenous peoples live in places were resource extraction companies are operating. Historically these interactions are hostile, yet there is a growing public

movement, pushing for the improvement of the ethical performance of companies and the engagement with indigenous peoples. MNEs operating in resource extracting industries are directly involved with natural systems, and thus cause environmental impacts (Esan, 2005).

Developing countries try to attract MNEs for a number of reasons: their potential to help with poverty alleviation; their power to use trade to link the developing world with the more developed world; the transfer of capital and technology; transfer management and marketing skills; and to provide jobs and help raise incomes (Kolk and Tulder, 2006). From a

sustainability perspective it can be motivated that MNEs expand to host countries with low a HDI, to contribute to the development of the host country. Esan (2005) and Pehn (2009) state that even though the responsibility for conflict resolution and prevention lies with the

government, there is also a pertinent role for the private sector in promoting peace in society. Most private sector actors aim at profits and not at the reduction of poverty, conflict or

fragility. However, foreign businesses sometimes exacerbate conflict and fragility by doing business with oppressive governments or excluding the participation of local communities.

Nevertheless, empirical research shows that resource extracting industries are

associated with persistent poverty rather than rising incomes ( O’Faircheallaigh 2012). Kolk and Tulder (2006) raise the question why not all MNEs have active policies regarding poverty alleviation. The scholars describe the MNEs who do not participate in poverty alleviation as free riders, and state that there is little hope that these companies can contribute to development. Research shows a “sector coordinated morality’’, where MNEs are only willing to state active commitment if others in their sector do as well. This can be explained by the fact that MNEs are afraid to lose out on competition that is not involved in poverty alleviation. Crawley & Sinclair (2003) emphasize the positive influence of adopting a

sustainable strategy on conflict resolution. Logically, conflicts that involve MNEs with a high sustainability focus, would be less severe and last shorter. Based on these findings, we can predict that:

H2a: The MNEs sustainability focus will positively moderate the influence of host country development on the degree of violence (conflicts will be less violent)

H2b: The MNEs sustainability focus will positively moderate the influence of host country development on the length of conflict (conflicts will be shorter)

Multinationals seeking to exploit natural resources on indigenous territories are often caught in a conflict situations between indigenous people and the government over the title to land and the rights for exploitation. The necessary interaction leads to ethically challenging situations, often resulting in conflict and the violation of indigenous rights (Esan, 2005; Lertzman & Vredenburg, 2005). Esan (2005) states that the expectations that the local communities have of the MNEs are sometimes unrealistic and unattainable, and therefore result in conflict. The scholar argues that MNEs should adopt conflict-sensitive policies and measures in their operations, when they operate in conflict zones.

When the local government is unable or unwilling to execute sufficient governance, the MNE can take over governance tasks, and thus become a support factor for the local communities (Esan, 2005). The sustainability focus of an MNE is responsible for a

harmonious relationship between the firm and the community. Fontana et al. (2015) argue that the large number of indigenous conflicts demonstrate that traditional strategies are not sufficient to conduct corporate interventions in new territories. In order to overcome conflict, ethical values emerged from mutual trust are necessary. Sustainable development is

important for the well- being of the community and is an important vehicle for the

recognition of indigenous peoples’ status and implementation of their rights, and therefore have a positive long-term impact (Bellier & Préaud 2012). Based on these findings, we can predict that:

H4a: The MNEs sustainability focus positively moderates the negative relation between the community autonomy and the degree of violence (conflicts will be less violent).

H4b: The MNEs sustainability focus positively moderate the negative relation between the community autonomy and the length of the conflict (conflicts will last shorter).

3.4 Conceptual Framework

Figure 1 shows the conceptual model demonstrating the relationship between the selected variables in this study. The figure presents a visualization of the hypothesized direct effects of the independent variables HDI and community autonomy on the dependent variables degree of violence and length of conflict. Additionally, the moderator effect of the sustainability focus on the latter relations is introduced. The control variables within this study are ownership structure, geographic isolation and political stability.

4. Methodology

This chapter describes the methods that are used in order to conduct the statistical tests necessary to find results for the proposed hypotheses. In the first paragraph, the sample and data collection are described. Second, the dependent, independent and moderator variable are presented respectively. The third section comprises the methodology of the statistical tests.

4.1 Sample & Data Collection

This study uses a cross-sectional research design to study the influence of community autonomy and the host country HDI on conflict resolution. For the purpose of this study, a sample is drawn from a number of conflict situations within the Commonwealth of Nations. A conflict between multinational firms and indigenous communities, is considered in case of any occurrence of conflicting interests resulting in friction or protest of any sort. In total, 237 cases filled the geographical requirements of this study and are introduced in the dataset. These cases are spread over the following 20 countries: Belize, Botswana, Canada, Guyana, Jamaica, Kenya, Lesotho, Malawi, Malaysia, Mozambique, Namibia, Nigeria, Papua New Guinea, Rwanda, South Africa, Tanzania, Uganda.

The case selection and reviewing is conducted in a group format, in which each individual reviewed 135 cases. For the purpose of this study, data is collected from the following databases: Environmental Justice Organization; Liabilities and Trade; American University of Washington Database; Business & Human Rights Resource Centre;

International Work Group for Indigenous Affairs; UN Special Rapporteur on Rights of Indigenous peoples; International Work Group for Indigenous Affairs; Indigenous

also found through media sources. Data collection is followed by coding, using a coding manual. In total, 74 variables are coded in order to conduct a statistical analysis.

4.2 Variables

4.2.1 Dependent Variables

A violent conflict is defined as: “organized physical force resulting from grievances between two or more parties and leading to injury or death to persons or damage or destruction to property” (Oetzel, Getz, Ladek, 2007). When the anger towards the MNE escalates, and the point of action is reached, communities usually have limited resources to devote to the fight. This leads to the frequent usage of resistance strategies, such as protest, sabotage and

blockades to draw attention while stagnating corporate activities. These tactics are known as weapons of the weak, and aim at causing maximum disruption and visibility at minimum costs. Another aim of these tactics is raising public awareness and support by attracting media attention (Calvano, 2008). The dependent variable of this study is conflict resolution. In order to give a comprehensive analysis of conflict resolution, the variable is split into two separate variables: the conflict length and the degree of violence.

Conflicts can differ in terms of type and intensity, and can be arranged from high to low violence (Getz & Oetzel, 2009; Calvano, 2008). The variable degree is violence is measured on a 6-point scale, and consists of the following values: (1) latent or no violence; (2) low level violence, i.e. peaceful protest and negotiations; (3) low level violence, i.e. court actions and occupation; (4) medium level of violence, i.e. intimidation and roadblocks,(5) high level violence i.e. physical damage, kidnapping and injury and (6) high level of violence, i.e. all above and death. In case the conflict went through multiple stages of

The second dependent variable is the length of the conflict. Conflicts are dynamic and evolve over time. Longer conflicts are associated with an increased potential for violence and costs for the firm and are less likely to be resolved (Mitchell, Angle & Wood, 1997; Pondy, 1967). The variable length of conflict is categorized into three subgroups:

(a) short, < 3 years; (b) medium, 3-8 years; (c) long, >8 years.

4.2.2. Independent Variable

The first independent variable is the Human Development Index (HDI). The HDI is introduced by the United Nations Development Programme to emphasize that economic growth alone is not sufficient for assessing the development of a country. The index includes both social and economic indicators to measure well-being. The annual HDI report ranks countries according to their index-score (McMahon & Moreira, 2014; Pehn, 2009).

The Human Development Index is a summary measure of average achievement in the key dimensions: development, health and standard of living. The development dimension measures whether people have a decent standard of living, a long and healthy life and the level of their knowledgeability. The health dimension is measured by life expectancy at birth. The education dimension is assessed by mean of years of schooling for adults aged 25 years and up, and the expected years of schooling for children. Lastly the standard of living dimension is evaluated by gross national income per capita. Critics of the index find it simplifying, since HDI only captures human development. The Index does not reflect poverty, empowerment, human security or inequalities. 1

The second independent variable is the community autonomy. This variable measures the relationship between the community and the government. Recognition by the state will

enhance development, according to O’Faircheallaigh (2012). This variable is coded as: (1) No autonomy at all; (2) Cultural recognition; (3) Limited recognition of rules, meaning that they can issue some minor rules, like marriages, recognized as valid by government; (4) Partial recognition of rules or right of consultation; (5) Almost full recognition of community rules by government with autonomy or right of consent.

4.2.3. Moderator Variable

The moderator variable of this research is the sustainability focus of the MNE. A sustainable focus comprises development that meets the need of the present, without diminishing the ability of future generations to meet their needs (Lertzman & Vredenburg, 2005). In this study, sustainability focus is measured by the FTSE4Good index membership. The index-series are designed to measure the performance of companies demonstrating the strength of Environmental, Social and Governance (ESG) practices. “Transparent management and clearly-defined ESG criteria make FTSE4Good indexes suitable tools to be used by a wide variety of market participants when creating or assessing sustainable investment products.’’2

This variable is coded as (0): no membership and (1): membership.

4.2.4 Control variables

Berkes (2009) emphasizes that any control analysis requires the inclusion of three major players within the sector: multinational firms, the state and the community. The control variable on firm level is the ownership structure. Second, on community level we use the geographical isolation violence. Lastly, on state level we introduce the political stability.

First, the ownership structure is, linked to the decision-making process and

distribution of power of the firm (Hart & Moore, 1990).This variable exists of the values: (1) state owned; (2) Publically traded (3) Privately owned.

Second, the geographic isolation indicates the ability of the community to gain access to viable recourses (Bebbington 1999). This variable is coded as: (1) Highly isolated, i.e., community members live within a community with little contact to outside world, (2) Most community members live within broader municipalities with mixed population.

Third, the political stability and absence of violence is a good predictor for conflict resolution, since fragile states are more prone to conflict (Oetzel, Getz and Ladek, 2007). The variable is measured by a percentage percentile, ranging from 0 to 100, where 0 represents no political stability and severe violence and 100 represents complete political stability. This variable is divided into six values: (1) 0-10th percentile, (2) 10-25th percentile, (3) 25-50th percentile, (4) 50-75th percentile, (5) 75-90th percentile, and (6) 90- 100th percentile. Appendix I presents an overview of all the variables.

4.3 Methods

Since this study includes two dependent variables, the analyses will be conducted in twofold. The proposed hypotheses are tested using a linear regression analysis, which is suitable when one or more independent variables are hypothesized to affect one dependent variable. The regression is described by the following equation:

𝑌 = 𝛽₀+ 𝛽₁𝑋₁+ 𝛽₂𝑋₂+ 𝛽₃𝑋₃𝜀

In this equation, Y represents the dependent variable conflict severity, measured by the degree of violence or length of conflict. The independent variables host country HDI and

community autonomy are respectively denoted as 𝑋1 and 𝑋2. The moderator variable of

sustainability focus is denoted as 𝑋3. 𝛽0 represents the intercept, 𝛽1,𝛽2𝛽3, respectively

represent the coefficient of the predictor variables of HDI, community autonomy and sustainability focus. 𝜀 is the error term, which corrects variance in de dependent variable, which are not explained by the model (Field, 2009).

Table 1 shows a summary of the combination of variables that are used in the two regression analyses. In order to test all the proposed hypothesizes, we made two blocks corresponding with each dependent variable. The first block which tests hypothesis H1a, H2a, H3a and H4a comprises the first six models. The second block corresponds include models 7-12 to test hypothesis H1b,H2b, H3b and H4b.

In both blocks, the first model corresponds with the control variable for H1 and H2 and includes: ownership structure and geographic isolation; the second model introduced the first independent variable HDI; the third model includes the moderation effect and introduced the variable FTSE4GOOd membership. The fourth model relates to the control variables of hypothesis 3 and 4, and include: ownership structure, political stability and geographic isolation; the fifth model focuses on the second independent variable community autonomy; and the sixth model introduces the moderator variable FTSE4Good membership. The

predictor variables of block two are equal, but the dependent variable of degree of violence is replaced by length of conflict.

Table 1: Summary of models Regression Analysis

Control Variables Independent Variables Moderator Variables

Ownership Structure Geographi c Isolation Political Stability HDI Community Autonomy HDI * Sustainability focus Community Autonomy* Sustainability Focus D eg re e O f V io le n ce (B lo ck 1 ) Model 1   Model 2    Model 3     Model 4    Model 5     Model 6      Le n g th o f C o n fl ic t (B lo ck 2 ) Model 7   Model 8    Model 9     Model 10    Model 11     Model 12     

5. Results & Analysis

This chapter presents the results of the data analysis. Various tests are performed with SPSS in order to find any relevant statistics in this dataset. This chapter starts with an overview of the descriptive statistics, test of multicollinearity and the correlations between the variables. Followed by the regression analysis as described in the method section. The results of the regression analysis provide us with the information needed draw conclusions regarding the proposed hypothesis.

5.1 Descriptive Statistics

This study includes 237 cases, spread over 20 different countries worldwide. There are cases from Africa (44.7%), South- America (1.7%), North-America (48.9%), Asia (1.7%) and Oceania (3%). Canada is overrepresented in this study, with 116 cases (48.9 %).

Table 2 gives an overview of the descriptive statistics of the variables in the model, including their minimum, maximum, mean and the standard deviation. Some information can be inferred by analysing the descriptive statistics. The variable degree of violence

(M= 2.030, SD=1,9688) has a high standard deviation, which implies that there is a lot of variance between the observations. This is confirmed when we take the specific categories into account: latent or no violence (40,1%); low level violence, i.e. peaceful protest and negotiations (5,5%); low level violence; i.e. court actions and occupation (11%), medium level of violence, i.e. intimidation and roadblocks (3,8%) and high level of violence, i.e. all above and death (20,7%). The distribution of the variables is U-shaped, meaning that the lowest and the highest category have the most observations. Analyzing the second dependent variable length of conflict (M= 1,992, SD=,8130), we find an almost equal distribution between the categories short (33,3%), medium (34,2%) and long (32,5%).

The mean of the independent variable HDI- Index is rather high

(M= 0.74957384, SD=0.186471790). Given that the HDI is measured on a scale of 0-1, this mean shows that most countries are rather developed. As previously mentioned, almost 50% of the cases originate from Canada. With an HDI- index of 0.92, this probably has a strong influence of the average HDI of the dataset. The high standard deviation of community autonomy (M = 3.148, SD = 1,1158), implies a wide spread of observations. This variable comprises the categories: no autonomy at all (9,3%); cultural recognition (16%); limited recognition of rules (37,1 %); partial recognition of rules or right of consultation (25,7%) and almost full recognition of community rules by government with autonomy or right of consent (11,8%).

Finally, the moderator variable sustainability focus is measured by a FTSE4GOOD membership (M= 0.226, SD 0.4188). There are 182 (76.8%) MNEs within this sample with a membership, contrasting 53 (22.4%) MNEs without a membership. It becomes evident that the majority of the companies have a sustainability focus, which could bias the results. The descriptive statistics of the control variables can also be found in table 3.

Table 2: Descriptive Statistics

N Minimum Maximum Mean St. Deviation

Degree Of Violence 237 ,0 5,0 2,030 1,9688 Length of Conflict 237 1,0 3,0 1,992 ,8130 Community Autonomy 237 1,0 5,0 3,148 1,1158 HDI- Index 2016 237 ,417000 ,939000 ,74957384 ,186471790 Sustainability Focus 235 ,0 1,0 ,226 ,4188 Ownership Structure 237 1,0 3,0 2,122 ,5502 Geographic Isolation 237 ,0 2,0 1,717 ,4606 Political Stability 237 1,0 6,0 3,553 1,7007

5.2 Correlation Analysis

After observing the descriptive statistics, we conduct a correlation analysis to uncover preliminary relations between the degree of violence, the length of conflict, the community autonomy, the HDI, sustainability focus, ownership structure, geographic isolation and political stability. The correlation coefficients in table 3 shows whether the correlation is positive or negative and if it is significant. The analysis of correlations shows a 10 significant correlations p<.01; 5 significant correlations p<.05; and 2 significant correlations p<.1. The correlation analysis is presented in table 6. The first notable result is the rather low correlation between degree of violence and length of conflict (r-.008, p.<.901), since both of the variables measure the dependent variable conflict resolution. For hypothesis 1a and 1b, concerning the direct effect of the independent variable HDI, we only find a

significant correlation with the dependent variable length of the conflict. The negative correlation coefficient (r-,395, p.<0.01) indicates that if the host-county has a lower HDI

score, the conflicts tend to last longer. The correlation coefficient between HDI and the degree of violence is -.100, however this correlation is not significant. The second independent variable community autonomy negatively correlates with degree of violence ( r -,178 p.<0.01) and length of conflict (r -,242 p.<0.01). This indicates that in cases where community autonomy is lower, the conflicts are more violent and last longer. The latter correlations confirm the direction of our proposed hypothesis.

Table 3: Correlations 1 2 3 4 5 6 7 1 Degree of violence 2 Length of conflict ,008 3 HDI -,100 -,395** 4 Community Autonomy -,178** _-,242** _,528** 5 Sustainability Focus ,125+ _,238** _-132* _,033 6 Ownership Structure -,101 ,069 -,151* _-,188** _-,099 7 Geographic Isolation -,089 -,052 ,202** _,164* _{-,020 -,047} 8 Political Stability -,088 -,367** _,933** _,477** _-,135* _-,113+ _,200**

** Correlation is significant at 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)

With respect to the moderating variable, the sustainability focus positively correlates with both the degree of violence (r, 125 p.<0.1) and the length of conflict (r, 238 p.<0.01). This indicates that when an MNE has a sustainability focus, the conflict tend to have a higher degree of violence and will last longer. This result is surprising, since it indicates that MNEs with a sustainability focus are involved in longer conflicts with indigenous communities. This result contrasts previously existing literature. In addition, we find a significant negative correlation between the independent variable HDI and the moderator variable sustainability focus (r, -132, p.<0.05). This implies that in countries where the HDI is rather low, more MNEs with a sustainability focus can be found.

Finally, when we analyze the control variables, we only find a significant negative correlation (r, -,367 p.<0.01) between political stability and the dependent variable length of conflict. This result indicates that the lower the political stability, the longer the conflict will endure. This in is line with our previous result concerning the HDI correlation.

5.3 Multicollinearity Test

Following the analysis of the descriptive statistics, a multicollinearity test is necessary. The multicollinearity test measures whether or not there are two or more predictor variables that correlate too highly. With perfect collinearity between predictors, it becomes impossible to obtain unique estimates of the regression coefficients because there are an infinite number of combinations of coefficients that would work equally well. The multicollinearity test is conducted to increase the reliability of this study. To test the multicollinearity between variables, we used the variance inflation factor (VIF) and the tolerance level. The VIF statistic indicates whether a predictor has a strong linear relationship with other predictors. The VIF score should be below 10 and the tolerance statistic should be higher than 0.1. Additionally, correlations above .80 also indicate critical values (Field, 2013).

The outcomes of the multicollinearity test can be found in table 4 and 5. There are no VIF values higher than 10, and no Tolerance below 0.1. However, we find concerning high correlation (r, -.933 p.<0.01) between HDI and political stability . This can be due to the fact that they both measure country level perceptions. Therefore, we exclude this control variable from the models including HDI (models 2,3,7,8).

Table 4: Multicollinearity test

Degree of Violence Tolerance VIF Ownership Structure ,946 1,057 Geographic Isolation ,951 1,051 Political Stability ,735 1,361 HDI score 2016 ,947 1,056 Community Autonomy ,743 1,345 Sustainability Focus ,956 1,046

TaTable 5: Multicollinearity test

Length of Conflict Tolerance VIF Ownership Structure ,946 1,057 Geographic Isolation ,951 1,051 Political Stability ,735 1,61 HDI score 2016 ,928 1,182 Community Autonomy ,743 1,361 Sustainability Focus ,960 1,042

5.4 Regression Analysis

For the final phase of the data-analysis, SPSS is used to perform a linear regression. The results, according to the proposed hypotheses, can be found respectively in table 6 and 7. Table 6 presents the regression results of the dependent variable degree of violence and table 7 presents the results of the dependent variable length of conflict. In total, 12 models are tested, for the 8 proposed hypotheses, and 5 out of 8 hypotheses are statistically confirmed.