Master Thesis
27-01-2017
Conflict resolution between Multinationals and Indigenous Communities
in Developed and Developing Countries
Romy Rohmer Student number: 10001176
MSc Business Administration: International Management Supervisor: Dr. Ilir Haxhi
Statement of Originality
This document is written by Romy Rohmer 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
In the last decades many conflicts have occurred between MNEs and indigenous communities all across the world. When MNEs expand their activities into areas inhabited by communities and do not take them into account, violent conflicts can occur when indigenous people try to make themselves heard. MNEs need to become more accountable for the negative spillover effects on indigenous communities and take action to resolve and prevent conflicts. In order to do this it is necessary to know more about the factors that influence conflict resolution and the context in which the conflicts occur. That is why in the current study, we examine whether the degree of development, measured by the Human Development Index, impacts the ability to resolve conflicts between MNEs and indigenous communities, and to what extent this relationship is moderated by the involvement of NGOs in conflict resolution. We explore these relationships for a broad sample of 705 cases across 66 different countries. The findings show that degree of development has a significant positive effect on conflict resolution and the moderating role of NGOs is partly supported by the data. Our results have theoretical as well as practical implications. It contributes to current literature by finding a trend across many cases that the degree of development of a country impacts the ability to resolve a conflict and it confirms the importance of NGOs as third party actors. Development was connected to conflict resolution between MNEs and indigenous communities before in the literature but this is the first study that finds a direct effect. This has practical implications for MNEs, indigenous communities, governments and NGOs when trying to resolve and prevent conflicts in developing countries.
Table of Content
Abstract 3 Table of Content 4 List of Tables and Figures 5 1. Introduction 6 2. Literature Review 11 2.1 Indigenous Communities 11 2.2 Conflict between MNEs and Indigenous Communities 13 2.3 National Context and Conflict Resolution 16 2.4 Third parties in Conflict Resolution 20 3. Theoretical Framework 23 3.1 Conflict Resolution in Developed and Developing Countries 23 3.2 Violence in conflict and the development of a country 26 3.3 Length of a conflict and the development of a country 28 3.4 Involvement of NGO 31 3.3 Conceptual Framework 36 4. Research Design 37 4.1 Sample and Data Collection 37 4.2 Variables 38 4.2.1 Dependent Variables 38 4.2.2 Independent Variable 39 4.2.3 Moderating Variable 40 4.2.4 Control Variables 40 4.3 Method of Analysis 42 5. Results and Analysis 44 5.1 Descriptive Statistics and Correlation Analysis 44 5.2 Regression Analysis 47 6. Discussion 53 6.1 Findings 53 6.2 Theoretical Implications 57 6.3 Practical Implications 59 6.4 Limitations 61 6.5 Future research 62 7. Conclusion 64 References 66List of Tables and Figures
Figure 1. Conceptual frameworkTable 1. Regression analysis with dependent variable degree of violence Table 2. Regression analysis with dependent variable length of conflict Table 3. Descriptive statistic and correlations between variables Table 4. Multicollinearity and VIF
Table 5. Regression analysis with dependent variable degree of violence
Table 6. Regression analysis with dependent variable length of conflict (short/medium/long) Table 7. Regression analysis with dependent variable length of conflict (number of months)
1. Introduction
In the last decades globalization created many new opportunities for Multinational Enterprises (MNEs) across the world. Especially in developing countries where there are new markets to be entered, more natural resources to be distracted and possibilities for outsourcing in order to exploit lower wages (Calvano, 2008; Kapelus, 2002; Mutti, Yakovleva, Vazquez-Brust, & Marco, 2012). Besides the market-based economic incentives, governments in developing countries also developed policies to attract MNEs to make Foreign Direct Investment (FDI) in their country, like tax breaks or low environmental regulations (Hilson, 2012; Spencer, 2008). Governments try to attract MNEs in the hope that the practices of these MNEs have positive spillovers to the local economy. However, these investments can also come with negative economic, social and environmental spillover effects for the local population (Boele, Fabig, & Wheeler, 2001; Oetzel & Doh, 2009; Spencer, 2008).
When expanding their operations abroad MNEs can encounter local indigenous communities that do not agree with MNEs moving into their territory. These encounters can lead to violent conflicts (Calvano, 2008; Laplante & Spears, 2008; Lertzman & Vredenburg, 2005). Conflicts have negative consequences for both the MNEs and the indigenous communities. For the communities it can be a threat to their livelihoods and the violence can severely physically hurt people, even up to death (Bruijn & Whiteman, 2010; Foster, 2012;
also have major financial consequences (Calvano, 2008; Ikelegbe, 2005; Kemp, Owen, Gotzmann, & Bond, 2011; O’Faircheallaigh, 2013). Making sure that these conflicts are resolved and preventing them from happening again is therefore in the interests of both the MNE and the indigenous community (Calvano, 2008).
Conflicts between MNEs and indigenous communities do not only occur in developing countries, but also in developed countries like Canada and Australia (Foster, 2012; Lertzman & Vredenburg, 2005). There seems to be a general consensus that national context matters in conflict resolution but whether conflict resolution is better in developed countries than in developing countries is still debated in current literature (Acuña, 2015; Bellier & Préaud, 2012; Hilson, 2012). Conflicts between MNEs and indigenous communities still arise and escalate across the world. MNEs need to become more accountable for the negative spillover effects on indigenous communities and take action to resolve and prevent conflicts. In order to do this it is necessary to know more about the factors that influence conflict resolution and the context across the many countries in which the conflicts occur (Calvano, 2008). Current literature discusses the relationship between the development of a country and conflict resolution but no direct effect is found. That is why this research addresses this gap. Therefore the first research question addressed in this study is: To what extent does the degree of development of a country affect conflict resolution
When looking at conflict resolution in developed and developing countries it is important to consider the role of NGOs. Current literature states that NGOs as third party actors can play a very important role in conflict resolution (Bellier & Préaud, 2012; Calvano, 2008; Getz & Oetzel, 2010; Kolk & Lenfant, 2013; Wiessner, 2011). NGOs can effectively mediate between MNEs and communities (Kolk & Enfant, 2013; Millar et al., 2004; Teegen et al., 2004), they can also pressure MNEs (Humphreys, 2005) or collaborate with MNEs to help minimize negative spillover effects and promote more sustainable development in developing countries (Dahan, Doh, Oetzel, & Yaziji, 2010; Oetzel & Doh, 2009). The importance of the role of NGOs is widely discussed in current literature but this is not researched in relation to the development of a country and conflict resolution. It is important to gain more understanding about the role that NGOs play in conflict resolution between MNEs and indigenous communities, especially in developing countries. Therefore the second research question addressed in this study is: To what extent does the involvement of NGOs
moderate the relationship between degree of development and conflict resolution?
The purpose of this research is to gain more understanding on the influence of the development of a country on conflict resolution between MNEs and indigenous communities. Also, to gain more understanding on the importance of the involvement of NGOs as third party actors by looking at the role of NGOs as moderator in this relationship. Most research is focused on economic development and institutional development, but in this research the
focus is on human development. This is a broader measurement of development that focuses on enhancing people’s capabilities to leave the life they want to live, instead of just looking at economic growth or the strength of institutions (United Nations Development Programme (UNDP), 1990). Most research on conflict resolution between MNEs and indigenous communities in current academic literature consists of qualitative case studies, looking at specific countries and conflicts. This research takes a quantitative approach; this is needed in order to find results that can be generalized. This is important for MNEs, indigenous communities and third parties involved because it gives them more insights into the factors that affect conflict resolution.
The research questions will be addressed by doing a quantitative research in which 706 cases are codified and analyzed. This study argues that degree of development has a positive effect on conflict resolution and that this is moderated by the involvement of NGOs. This relationship is analyzed by doing a regression analysis. The findings show that degree of development has a significant positive effect on conflict resolution and the moderating role of NGOs is partly supported by the data.
This study has theoretical as well as practical implications. It contributes to current literature by showing that it is possible to generalize across many unique cases and find common factors that influence conflict resolution. It also adds to the current knowledge on conflicts resolution between MNEs and indigenous communities by showing that the degree
of development of a country impacts the ability to resolve a conflict. Furthermore, it shows the importance of the role that NGOs can have in developing countries in the resolution of these conflicts. This has practical implications for MNEs, indigenous communities, governments and NGOs when resolving and preventing conflicts. Since higher development can increase the chances for conflict resolution and reduces the risk for severe conflicts this emphasizes the importance for MNEs to promote sustainable development and to take into account the local population in the areas where MNEs operate, especially in developing countries. Collaborating with NGOs can help to reach this and resolve conflicts with indigenous communities.
The outline of this research is as follows. The next chapter consists of the literature review, which introduces current literature on conflicts between MNEs and indigenous communities and shows the need for this research. This is followed by the theoretical framework, which argues the hypotheses and concludes with the conceptual framework. After that the research design is explained: the sample, data collection, variables and method. The fifth chapter contains the results of the regression analysis. The discussion chapter follows this and the final chapter of this study is the conclusion.
2. Literature Review
This chapter discusses current literature on conflicts between MNEs and indigenous communities. This chapter starts by explaining what indigenous communities are and elaborates on the content of indigenous rights. After that it discusses conflicts in the national context and the role of third parties in conflict resolution.
2.1 Indigenous Communities
There is no universally stated definition for the term ‘indigenous community’ but in general indigenous communities are long existing local populations living in a certain geographic area with their own traditional language, culture and beliefs (Adamson & Pelosi, 2014; Calvano, 2008). All indigenous communities are different and generalizing is therefore difficult. However, it can be said that environmental values are often a central aspect of their traditional culture and that indigenous peoples often have a different worldview than the dominant Western worldview (Acuña, 2015; Calvano, 2008). Due to the colonial history in which indigenous communities are suppressed and dispossessed of their lands, the communities often form marginalized groups in society (Laplante & Spears, 2008; Parlee, 2015). The difference in worldview and the feeling of marginalization is why indigenous communities often do not identify with states and why they want to have their own autonomy so that they are ableto protect their territory and conserve their traditional cultures (Bellier &
In 2007 the collective rights of indigenous communities were officially adopted in the United Nations Declaration on the Right of Indigenous Peoples (United Nations, 2008). This declaration is based on the autonomy for indigenous communities to collectively live according to their own institutions and gives them legitimacy as a political actor at the international level (Bellier & Préaud, 2012). It recognizes how important it is for indigenous peoples to have collective rights, and to be part of a community, instead of just individual human rights (Wiessner, 2011). This declaration also includes the rights to sovereignty of their lands: to control their lands, to not become displaced, and to ensure the preservation of their lands (United Nations, 2008).
The declaration for indigenous rights is adopted at the international level but the implementation and enforcement is at the national level. Unfortunately, the enforcement of indigenous rights is lacking and there are still many gaps between and within countries when it comes to addressing the rights (Bellier & Préaud, 2012). Indigenous peoples remain under great pressures trying to protect their territory. Because of the lack of enforcement of indigenous rights, they still remain minorities in political, economic and social processes (Bellier & Préaud, 2012; Wiessner, 2011). These processes do not only concern the relationship between the state and the indigenous communities; as economic globalization develops further, MNEs have also become prominent actors in the debate on indigenous rights (Bellier & Préaud, 2012). Especially with regard to indigenous land rights and the
control over natural resources, MNEs are increasingly involved and conflicts between indigenous communities and MNEs occur regularly (Bellier & Préaud, 2012; Calvano, 2008; Foster, 2012).
2.2 Conflict between MNEs and Indigenous Communities
Since the extractive industries started expanding in the 1980s and other companies also started to outsource production abroad, an increasing number of projects are located near marginalized communities. These communities are not only geographically marginalized, but also economically and politically marginalized within their country and often in countries that are marginalized within the global system (Laplante & Spears, 2008; Parlee, 2015). The members of the communities tend to not have a political voice, lack access to justice, lack access to social services, are not able to participate in the economy, have high poverty and do not have control over their lands (Laplante & Spears, 2008). When MNEs move their operations into a territory without taking into account the effects on the environment and on the local indigenous communities living there, this can lead to violent conflicts (Calvano, 2008; Lertzman & Vredenburg, 2005). When MNEs do not take into account the consent of the indigenous communities, the indigenous people can feel even more marginalized and excluded from participating in the decision-making process. This can lead to high discontent and violent conflict because they feel disempowered and want to make themselves heard
Not only can indigenous communities see it as an invasion of their territory, but also as a direct threat to their traditional values and their livelihood (Bruijn & Whiteman, 2010; Laplante & Spears, 2008). The livelihoods of indigenous peoples are mostly dependent on agriculture and fishing. When MNEs come in to extract resources, this can come with great environmental, economic and social costs for the indigenous communities (Boele et al., 2001; Oetzel & Doh, 2009; Spencer, 2008). Contamination can harm their drinking water and can decrease the fishing practices; the traditional culture and identity of the community can be under attack; the environment can be destructed and this can ruin the harvest; and they can become displaced from their ancestral lands (Foster, 2012). Besides concerns related to threats to their livelihoods, health, environmental impacts, security and indigenous rights, the major concern for a lot of communities is that the company does not consult them. This leaves them unable to have any control over the project and also leaves them unable to benefit from it (Laplante & Spears, 2008). The unequal distribution of costs and benefits for the communities, together with the feelings of marginalization and exploitation, are the main causes for violent conflicts between MNEs and indigenous communities (Ballard & Banks, 2003; Calvano, 2008; Lertzman & Vredenburg, 2005).
When the concerns of communities are not addressed and conflict arises, not only do these conflicts have negative consequences for the indigenous communities but also for the MNE. For the MNEs, conflict can be very bad for their reputation and it can also have major
financial consequences (Calvano, 2008; O’Faircheallaigh, 2013). The most frequent costs are the time spent by personnel on risk and conflict management and the costs that arise when the production process is disrupted (Davis & Franks, 2011). The financial consequences can also come from project delays, lawsuits, sabotage, destruction or theft of resources (Ikelegbe, 2005; Kemp et al., 2011). This can be a serious threat to the profitability of the MNE and can create complications in the day-to-day business activities as well as the long-term strategy of the MNE (Oetzel, Getz, & Ladek, 2007). Corporate policies of MNEs now try to address indigenous rights and seek to make agreements with communities to try to prevent conflict from occurring (O’Faircheallaigh, 2013; Onuoha & Barendrecht, 2012). MNEs try to address stakeholders, including indigenous communities, and development challenges in their Corporate Social Responsibility (CSR) practices (Calvano, 2008; Kolk & Lenfant, 2013; Mutti et al., 2012). However, companies are not always forthcoming in their CSR practices. The promises of more economic development, jobs and quality of life are broken and still many negative spillover effects occur (Hilson, 2012; Mutti et al., 2012). Conflicts still arise and escalate, making sure that these conflicts are resolved and preventing them from happening again is in the interests of both the MNE as the indigenous community. MNEs need to become more accountable for the negative spillover effects of their activities that violate indigenous rights. In order to do this it is necessary to know more about the factors that influence conflict resolution and the context across the many countries in which these
conflicts occur (Calvano, 2008).
2.3 National Context and Conflict Resolution
Many scholars link national context with conflict resolution, looking especially at the institutional context of developed and developing countries when considering MNE behavior across different countries. Instead of seeing communities as small, territorial, homogenous groups, it is important to take into account the political processes between other actors with their own interests and the institutional context that influences the outcomes of these political processes (Agrawal & Gibson, 1999). Local communities are often the actors with the least power in these political processes (Agrawal & Gibson, 1999). The institutional context can be seen as the “rules of the game”; for MNEs this varies from country to country. The national formal and informal institutions structure political, economic and social interactions and influence MNE behavior in that country (North, 1990). However, how to measure institutions and the effect they have on development is something that is under debate in current literature (Chang, 2010; Garrido, Gomez, Maicas, & Orcos, 2014).
Multiple views come forward when looking at institutions, development and conflict between MNEs and indigenous communities in current literature. The dominant view is that institutions are seen as weaker in developing countries than in developed countries. This decreases the position of indigenous communities due to a weak legal system and regulations,
Ikelegbe, 2005; Kolk & Lenfant, 2013). However, oppositely Bellier & Préaud (2012) state that sometimes indigenous rights are better protected in developing countries because the international community is watching more closely and tries to ensure enforcement of these rights. Instead, in developed countries, the state is the main actor to enforce indigenous rights (Bellier & Préaud, 2012). In countries with low quality institutions, the government is often seen as one of the factors that increases conflicts due to unaccountability and corruption in combination with weak regulations (Ikelegbe, 2005; Kolk & Lenfant, 2013). When government oversight is weak or corrupt, this can allow MNEs to gain unfair advantages over indigenous communities (Murphy & Arenas, 2010). Also, when the state is not able to provide local communities with social services like justice, security, education and health, these communities lose trust in the state and are more likely to take matters into their own hands and protect their own interests (Ballard & Banks, 2003).
Low development of a country is seen as one of the causes of conflict by some scholars, mostly because developing countries are seen as weaker in enforcing regulations for MNEs (Bebbington, 2008; Lertzman & Vredenburg, 2005). There is much focus on weak institutional governance, and on economic growth and neoliberal policies, when looking at the conflicts of MNEs with indigenous communities in developed and developing countries (Acuña, 2015). However, development is not just about institutional governance and economic growth; it is about human development. More important is how the economic
growth trickles down to the population and if national governance creates an environment that can enhance the wellbeing of society (UNDP, 1990). Human development focuses on providing people with the capabilities to enhance the quality of life and lead the life they want to live (Sen, 1990). National income numbers do not show this, and good governance, economic growth and more FDI does not necessarily lead to increased human development (Reiter & Steensma, 2010; UNDP, 1990). This research looks at development in the broader sense and not just at institutional and economic development, to see how this influences conflict resolution between MNEs and indigenous communities.
Acuña (2015) states that conflicts do not emerge simply due to a problem of governance but because of the colonial history where indigenous communities are suppressed and the dominant Western worldview that is very different from the indigenous culture and values and the way they see territories. Especially in countries that are rich of natural resources, indigenous communities are dispossessed of their lands to make place for resource extraction and economic development. This leaves the communities without the needed resources and land to create sustainable livelihoods and leaves them marginalized in their access to capital (Calvano, 2008; Murphy & Arenas, 2010; Parlee, 2015). This marginalization of indigenous peoples and the unequal distribution of costs and benefits are the main causes for violent conflicts between MNEs and indigenous communities (Ballard & Banks, 2003; Calvano, 2008; Lertzman & Vredenburg, 2005). This is happening in both
developed and developing countries. There seems to be a general consensus that national context matters in conflict resolution, but whether conflict resolution is better in developed countries than in developing countries is still debated in current literature (Acuña, 2015; Bellier & Préaud, 2012; Hilson, 2012).
MNEs need to become more accountable for the negative spillover effects on indigenous communities and take action to resolve and prevent conflicts. In order to do this it is necessary to know more about the factors that influence conflict resolution and the context across the many countries in which the conflicts occur (Calvano, 2008). Current literature discusses the relationship between the development of a country and conflict resolution but no direct effect is found. It can be rewarding to explore the differences between developed and developing countries because it can help to understand the factors that obstruct conflict resolution and how this influences the strategies and performance of MNEs when going abroad. That is why this research addresses this gap. This research tries to find out how important national context, in particular the degree of development of a country, is in conflict resolution across the world. Therefore the first research question of this study is: To what
extent does the degree of development of a country affect conflict resolution between multinationals and indigenous communities?
2.4 Third parties in Conflict Resolution
When looking at conflict resolution between MNEs and indigenous communities it is also important to take into account the involvement of third party actors. Involvement of third parties can create pressure for conflict resolution and therefore shorten the length of conflicts (Humphreys, 2005). Not only can they pressure for resolution but they can also create the possibility to move from confrontation to collaboration (Arenas, Sanchez, & Murphy, 2013). The third parties prominent in current literature are governments and NGOs.
When the government is involved they can often play a positive role in the resolution of the conflict between MNEs and indigenous communities (Calvano, 2008). However, governments in developing countries can lack the capacity and resources to interfere in conflicts within the national borders (Getz & Oetzel, 2010; Murphy & Arenas, 2010), or they might not have the willingness to do it because they want to favor the MNE over the indigenous community for economic reasons, as stated in the previous section. Oppositely, in developed countries the recognition of indigenous communities is usually more advanced and the legal framework is stronger, which can provide the communities with land rights and a stronger position against MNEs (Foster, 2012). Unfortunately, government involvement in conflicts does not always have a negative effect on conflict resolution; weak regulations in combination with a unaccountable and corrupt government can increase conflicts (Ikelegbe, 2005; Kolk & Lenfant, 2013; Owoye & Bissessar, 2012).
Besides governments, NGOs also come forward in current literature as increasingly important third parties in resolving conflicts between MNEs and indigenous communities (Arenas et al., 2013; Bellier & Préaud, 2012; Calvano, 2008; Getz & Oetzel, 2010; Kolk & Lenfant, 2013; Wiessner, 2011). NGOs come into the playing field as informal institutions that are changing the context in which government and the private sector operate (Teegen, 2003). NGOs are organizational entities that arise from social movements that have become more structured and organized. Since NGOs arise from social movements they have a form of social purpose and serve a particular societal interest (Millar, Choi, & Chen, 2004; Teegen, Doh, & Vachani, 2004). NGOs step in when the government cannot effectively protect the common good. Sometimes governments do what is best for the national interest, like economic growth, but instead forget about the global collective good, like the environment or human rights and protecting ethnic minorities (Schepers, 2006; Teegen, 2003). International NGOs can then have an active role in pressuring MNEs or they can also work together with MNEs to resolve conflicts. NGOs can have a good relationship with the local communities and can therefore play an important role in mediating for conflict resolution between the MNEs and indigenous communities (Kolk & Lenfant, 2013; Millar et al., 2004; Teegen, 2003). By collaborating with NGOs, MNEs can try to minimize the risk of conflicts with local communities, and promote more sustainable development (Dahan et al., 2010; Oetzel & Doh, 2009).
From the literature it comes forward that NGOs are important third party actors in conflict resolution. However, no direct effect of their role in conflicts resolution across developed and developing countries is found in current literature. It is important to gain more understanding about the role that NGOs play in conflict resolution between MNEs and indigenous communities, especially in developing countries. This leads to the second research question: To what extent does the involvement of NGOs moderate the relationship
between the degree of development of a country and conflict resolution between multinationals and indigenous communities?
3. Theoretical Framework
This chapter gives theoretical background on conflict resolution between MNEs and indigenous communities in developed and developing countries and on the moderating role of NGOs. Furthermore, the hypotheses are proposed based on current literature and this chapter concludes with the conceptual framework of this study.
3.1 Conflict Resolution in Developed and Developing Countries
There is a general consensus in current literature that the national context is important to consider when looking at conflict resolution between MNEs and indigenous communities. However, whether conflict resolution is better in developed countries than in developing countries is still debated in current literature (Acuña, 2015; Bellier & Préaud, 2012; Hilson, 2012). Most research on conflict resolution between MNEs and indigenous communities in current academic literature consists of qualitative case studies, looking at specific countries and conflicts. This research takes a quantitative approach; this is needed in order to find results that can be generalized. This is important for MNEs, indigenous communities and third parties involved because it gives them more insights into the factors that affect conflict resolution.
From the literature review it comes forward that the main causes for violent conflicts between MNEs and indigenous communities are: exploitation of indigenous territories,
(Ballard & Banks, 2003; Calvano, 2008; Lertzman & Vredenburg, 2005). Even though this is happening both in developed and developing countries, this research argues that conflicts are more severe in developing countries and that this decreases the chances for conflict resolution. MNEs often expand their business to developing countries because they see possibilities for exploitation of cheap labor and resources. The economic benefits that the MNEs can bring in developing countries are often chosen over the rights of indigenous communities and therefore the MNEs are less restricted in their practices, which can lead to violent conflicts when indigenous communities try to stand up for their rights (Foster, 2012; Ikelegbe, 2005; Kolk & Lenfant, 2013). Less regulation for the MNEs can lead to severe exploitation of indigenous communities and this is also a major reason for conflict (Calvano, 2008; Hilson, 2012; Lertzman & Vredenburg, 2005). Especially in developing countries with large extractive industries that dominate the national economy, there are many cases where government takes the money from the MNEs to benefit themselves and suppress the local communities and opposition (Boele et al., 2001; Hilson, 2012).
When looking at conflict resolution in developed and developing countries it is important to consider the role of NGOs. As comes forward from the literature review, current literature states that NGOs as third party actors can play a very important role in conflict resolution (Bellier & Préaud, 2012; Calvano, 2008; Getz & Oetzel, 2010; Kolk & Lenfant, 2013; Wiessner, 2011). NGOs can effectively mediate between MNEs and communities
when conflicts occur (Arenas et al., 2013; Kolk & Lenfant, 2013; Millar et al., 2004; Teegen, 2003). They can also pressure MNEs (Humphreys, 2005) or collaborate with MNEs to help minimize negative spillover effects and promote more sustainable development in developing countries (Dahan et al., 2010; Oetzel & Doh, 2009). NGOs are increasingly important actors in the global institutional context (Millar et al., 2004; Teegen et al., 2004). NGOs trying to protect the rights of indigenous peoples are present in developed and developing countries, but especially in developing countries the role of NGOs is increasingly important. This is because there are more NGOs active in developing countries and they are seen as important actors (Brown & Kalegaonkar, 2002). Also, due to weaker governments that do not have capacity to interfere, NGOs more often target MNEs in developing countries (Murphy & Arenas, 2010; Schepers, 2006).
Because of the important role that NGOs can have as a mediator or facilitator in conflict resolution, especially in developing countries, it is proposed that the involvement of NGOs influences the relationship between the degree of development and conflict resolution between MNEs and indigenous communities. To gain better understanding about conflict resolution in developed and developing countries, two important aspects that influence the severity of a conflict are taken into account. First the amount of violence in a conflict is linked to the degree of development of a country and then the length of a conflict is linked to
the degree of development of a country. After that the role of NGOs in conflict resolution in developed and developing countries is discussed.
3.2 Violence in conflict and the development of a country
One aspect of conflicts to consider when looking at conflict resolution in developed and developing countries is the amount of violence. The degree of violence is an important factor in determining the severity of a conflict and the chances of resolving a conflict (Getz & Oetzel, 2010; Humphreys, 2005). When MNEs operate in fragile states with weak governments this can add to the already existing political, social and economic conditions that support violent conflicts (Oetzel et al., 2007). Various reasons come forward from current literature regarding why the conditions in developing countries have a higher risk of violence when conflicts arise.
The first reason why conflicts can become more violent in weaker states is because the state is more dependent on natural resources. In developing countries, abundant natural resources tend to have the opposite effect on the wellbeing of society than in developed countries, this paradox is called the resource curse (Bebbington, 2008). This resource curse in developing countries can also lead to more violent conflicts (Oetzel et al., 2007). This is due to the fact that if a countries wealth is largely based on natural resources, for example in Nigeria, governments are more likely to favor extractive industries MNEs when a conflict
and the community does not have a political voice, the conflicts are more prone to violence (Foster, 2012; Ikelegbe, 2005; Kolk & Lenfant, 2013).
This is also supported by the argument of Collier and Rohner (2008) that as the income of countries rises, they become safer democracies. The reason that democracies are not always safe in poor countries is due to bad governance and corruption. Owoye & Bissessar (2012) find that weak institutions and corruption in African countries lead an increasing level of violence. Furthermore, in developing countries, sometimes the government has a very centralized power that can become oppressive when responding to conflict instead of resolving the conflict (Sawyer, 2004). For example, in Nigeria, state intervention in conflict was extremely suppressing and instead of resolving conflict it created even more violence (Ikelegbe, 2005). Having a decentralized, more regional governance that provides public goods and services and that can protect common pool resources can decrease violent conflicts (Sawyer, 2004).
The last argument for more violent conflict in developing countries comes from the research of Sambanis (2004), that shows that there is a strong negative relationship between income per capita of a country and political violence. Poverty does not necessarily cause the violence, but high inequality in a country does significantly impact the amount of violence. The highest risk of violence is in countries where there is poverty as well as ethnoreligous separation in society (Sambanis, 2004). Thus, where indigenous communities form
marginalized groups in low-income countries, there is the highest risk of violence. As mentioned before, the already marginalized indigenous communities feel even more marginalized and excluded when MNEs do not consult them. This can lead to high discontent and violent conflicts because they want to make themselves heard (Ballard & Banks, 2003; Laplante & Spears, 2008).
To sum up, there are various arguments coming from literature regarding why there is a higher risk of violence in conflicts that occur in developing countries. First, governments are more likely to favor MNEs when the country has a high economic dependency on natural resources. Furthermore, higher income countries are generally safer and higher inequality in a country also significantly increases risk of violence. Following these arguments, it is proposed that in developing countries the conflicts between MNEs and indigenous communities are more violent.
H1: The lower the development of a country is, the higher the degree of violence in a conflict is (negative relationship).
3.3 Length of a conflict and the development of a country
Violence is not the only characteristic that negatively contributes to the resolution of conflicts between MNEs and indigenous communities; the length of a conflict can also decrease the chances of resolving the conflict. The longer a conflict lasts, the less likely it is that the
institutional context in developing countries can be seen as weaker (Hilson, 2012; Kemp et al., 2011; Kolk & Lenfant, 2013), whereas the institutional context in developed countries is more stable and involves more regulations and codes (Young & Thyil, 2013). This weaker institutional context that often favors MNEs and the poor relationship between state and society can explain why conflicts that last for a long time are usually in weaker states (Hironaka, 2009; Humphreys, 2005). In many developing countries, MNEs are less regulated and communities are less protected by the state, which creates a very unequal battlefield for the communities and can result in long-lasting conflicts.
A weak institutional context with a corrupt and unaccountable government also makes the chances less likely that the government will interfere in the conflict. When the government interferes it can mediate for conflict resolution between the MNE and the indigenous community and therefore shorten the length of conflict (Calvano, 2008; Humphreys, 2005). If the government interferes in the conflict, this is most often done at the beginning of the conflict because, if the government does not interfere, it is usually because the government does not want to interfere or does not have the capacity to do so (Getz & Oetzel, 2010). Since governments in developing countries can lack the willingness or the capacity and resources to interfere in a conflict, this is another reason that conflicts can last longer in developing countries.
What also influences the length of a conflict is the position of the indigenous communities. One of the differences between opposition from indigenous communities in developed and developing countries against MNEs is that communities from developed countries are better in making their demands heard. This is due to better political representation and the availability of better organized and resourced social movements and NGOs. In developing countries there is increased activism by indigenous communities, which can lead to conflicts with MNEs (Fabig & Boele, 1999). Kapelus (2002) investigated the case of Rio Tinto, a British-Australian mining company, which faced violent conflict with local communities in South Africa since 1980. The global CSR policy of Rio Tinto makes general assumptions about the needs of communities and the legitimacy of local governments, leading to failure in effectively implementing the CSR in subsidiaries (Kapelus, 2002). The political position that an indigenous community has in a country can make it harder to resolve the conflicts and thus can result in long lasting conflicts between MNEs and indigenous communities.
Thus, since there are more regulations and stronger institutions in developed countries it is therefore more likely that government will interfere to enforce these regulations. Not only are governments less likely to interfere in developing countries, but the indigenous communities also tend to have a weaker position when trying to make their demands heard.
Following these arguments, it is proposed that in developing countries the conflicts between MNEs and indigenous communities are not only more violent but also last longer;
H2: The lower the development of a country is, the longer the length of a conflict is (negative relationship).
3.4 Involvement of NGO
NGOs are a powerful force in the field of business, both in developed and developing countries. NGOs can have the capacity to disrupt business practices when trying to protect the environment and the rights of indigenous peoples (Schepers, 2006). In developed countries MNEs are often pressured to take into account the environmental and social dimension. However, MNEs are more often the target of NGOs in developing countries (Schepers, 2006). This is not only because there are a more NGOs active in developing countries and NGOs are seen as very important actors in developing countries (Brown & Kalegaonkar, 2002), but also because in countries where the government is weaker the involvement of NGOs as facilitators and mediators is increasingly important (Murphy & Arenas, 2010). When there is exploitation by MNEs and a lack of enforcement from the government, they can step in to protect the public goods and pressure the MNE (Schepers, 2006; Teegen, 2003). By pressuring the MNE to take into account the local community, the involvement of NGOs can prevent the indigenous communities from using violence to make
Not only can NGOs pressure MNEs but MNEs can also collaborate with NGOs to overcome foreign liabilities. When entering another country MNEs can experience liabilities of foreignness (LOF) due to challenges that occur when entering a foreign market and moving their operations into a developing country. By engaging with NGOs, MNEs might be able to overcome the LOF, minimize negative spillover effects, like conflicts with local communities, and promote more sustainable development (Dahan et al., 2010; Oetzel & Doh, 2009). NGOs can have the necessary knowledge, credibility and tangible resources that MNEs lack because the NGOs know the local context. Also, NGOs can help with environmental and social problems or provide access to certain areas and resources (Dahan et al., 2010). It can help MNEs to reach markets and access resources that would otherwise not be available to them. NGOs and MNEs can have shared goals and complementary resources and this can make beneficial collaborations: economic returns for the MNE and social benefits for the community (Oetzel & Doh, 2009). For example, Nestle has corporate-NGO collaborations in West Africa in which they promote more sustainable farming for local communities, providing better ingredients for Nestle and higher yields for the farmers (Dahan et al., 2010). This collaboration between MNEs and NGOs can therefore minimize the discontent of unequal benefits by indigenous communities and thus prevent escalation of violent conflicts.
However, despite the benefits that collaboration between MNEs and NGOs can bring, unfortunately the collaboration does not always become a success. Multiple factors can influence the success of the collaboration: the country, industry, company and NGO (Oetzel & Doh, 2009). Furthermore, the fundamentally different values, objectives and structures of MNEs and NGOs can also lead to conflicts in collaborations due a lack of trust, common experience and communication (Dahan et al., 2010; Oetzel & Doh, 2009). Especially in developing countries, there is often mistrust and hostility between NGOs and MNEs (Oetzel & Doh, 2009). When NGOs are trying to pressure MNEs and attempting to establish a partnership while there are no common values, the activities of the NGO might even lead to even greater conflict and make the communities that they are trying to help even worse off (Millar et al., 2004). There are also cases where the involvement of NGOs does not have the desired effect because the NGOs are funded by corporations and are focused on gaining resources instead of protecting the rights of the indigenous community (Murphy & Arenas, 2010).
To sum up, since NGOs can step in to protect the common good and pressure the MNEs to take into account the environmental and social dimensions, this can prevent indigenous communities from using violence to make themselves heard. Moreover, NGOs can also collaborate with the MNE to minimize the negative spillover effects on the local communities and promote more sustainable development; this can also prevent violent
conflicts from occurring. Therefore, it is proposed that NGOs moderate the influence of development of a country on the degree of violence in a conflict. However, since the involvement of NGOs in a conflict does not always positively influence conflict resolution, but can also have negative effects, it is not clear whether this moderation effect will be positive or negative.
H3: The involvement of NGOs will moderate the influence of development of a country on the degree of violence.
When NGOs do successfully interfere and pressure MNEs for conflict resolution, this can also help shorten the length of a conflict (Humphreys, 2005). As mentioned, interference by NGOs occurs more often in developing countries because governments are weaker and, when there is a lack of enforcement from the government, NGOs can step in to protect the public goods and pressure the MNE for conflict resolution (Brown & Kalegaonkar, 2002; Murphy & Arenas, 2010; Schepers, 2006; Teegen, 2003). For example, the activities of the Spanish oil and gas company Repsol in Latin America encountered various conflicts with indigenous communities. The NGO Intermon Oxfam published a series of report in which it accuses Repsol of human rights violations of indigenous people in the Chaco region in Bolivia. After several years of accusations and media attention, Repsol acknowledged the need to change their policies towards indigenous peoples. In collaboration with Intermon Oxfam and several other international NGOs, Repsol adopted a policy that acknowledges the
rights of indigenous people in accordance with United Nations Declaration on the Right of Indigenous Peoples (Arenas et al., 2013).
NGOs are seen as important actors in the global context and especially in developing countries (Brown & Kalegaonkar, 2002). This gives them the legitimacy to work as mediator or facilitator in conflict resolution. When the mediation or facilitation with NGOs is successful this can shorten the length of conflict. For example, the international NGO Oxfam Novib facilitated a dialogue between BHP Billiton, the largest mining company in the world, and seven indigenous communities in Peru. After three years of negotiating they came to an agreement in which BHP agreed to compensate the communities for the loss of their livelihoods due to pollution and land lost. The parties agreed that this agreements would not have been made without the independent and respected third party actor Oxfam Novib that facilitated the discussion (Murphy & Arenas, 2010).
However, as said before, not all involvement of NGOs leads to conflict resolution. The same reasoning as the previous hypothesis can be used for the length of a conflict. Since the role that NGOs can have as a mediator or facilitator in conflict resolution is important, especially in developing countries, it is proposed that the involvement of NGOs can influence the relationship between the degree of development and the length of a conflict. However, since the involvement of NGOs in a conflict does not always positively influence conflict
resolution, but can also have negative effects, it is not clear whether this moderation effect will be positive or negative.
H4: The involvement of NGOs will moderate the influence of development of a country on the length of a conflict.
3.5 Conceptual Framework
Figure 1 shows the relationships between the variables in the conceptual framework. This study proposes that degree of development has a positive effect on conflict resolution, i.e. shorter and less violent conflicts. It is also proposed that this relationship is moderated by the involvement of NGOs.
Figure 1. Conceptual Framework
Degree of Development Degree of Violence Length of Conflict Involvement of NGO H1 H2 H3 H4
4. Research Design
This study uses an explanatory cross-national research design with quantitative methods to study the relationship between the degree of development of a country and conflict resolution between MNEs and indigenous communities. This chapter describes how this research is conducted. First, the sample and the method for data collection are explained. After that the dependent, independent, moderating and control variables are described. The chapter concludes with the method of data analysis.
4.1 Sample and Data Collection
The sample for this research consists of cases where there was a direct conflict between a MNE and an indigenous community. An existing dataset with 706 cases that other students computed in previous years was used in this study. The cases were selected by non-random sampling; based on the availability of enough information on the cases and if this information is suitable to be coded and added to the dataset. The dataset includes cases from 1980 onwards in countries spread across 67 countries in six different continents and involving multinationals from different industries. For Somalia the HDI score was not present. Therefore this case was removed from the dataset, leaving a sample of 705 cases in a total of 66 countries.
which then were codified by the students themselves. Other variables at the country level were collected from the World Bank and UNDP database. The 67 variables of the 706 cases from the existing dataset were double checked and missing variables were imputed by this year’s group of students that are studying this topic.
4.2 Variables
4.2.1 Dependent Variables
The length of a conflict and the severity of violence are important characteristics of conflicts and positively contribute to the resolution of a conflict (Getz & Oetzel, 2010; Humphreys, 2005). That is why in this study the dependent variable conflict resolution is measured by two different variables: degree of violence and length of conflict.
The dependent variable degree of violence is divided into a six point scale with different types of violence from either side. The variable is coded in the following types of violence: (1) no violence, (2) low level of violence with peaceful protests and negotiations, (3) low level of violence with court actions and occupation, (4) medium level of violence with intimidation and roadblocks, (5) high level of violence with physical damage, kidnapping and injury, and (6) the highest degree of violence with deaths during the conflict. The dependent variable length of the conflict is measured by the number of months that a conflict lasts. The variable was checked for outliers by using the z-values, all the cases
dataset. All together 17 cases were removed, leaving a sample of 688 cases. After deleting the outliers the mean of the length of conflict was 99 months, which is slightly more than 8 years. That is why the conflicts that lasted longer than 8 years were coded as (3) long. Since the mean was higher than the median of 79 months it was decided to split the rest of the sample into two groups, in order to create groups that are more equal in the amount of cases. It was coded in (2) medium, 3 to 8 years and (1) short, up to 3 years.
4.2.2 Independent Variable
The independent variable of this study is the degree of development of a country. The right measure for development is something that is under debate. GDP is most commonly used as an indicator for development based on economic performance. However the GDP of a country is not an adequate measurement because it does not say anything about the wellbeing of people and about how the economic benefits are divided in society (Fitoussi, Sen, & Stiglitz, 2009). That is why in this study the degree of development of a country is measured by using the Human Development Index (HDI), which is collected from the UNDP database. This measurement of development does not only take into account economic activity (GDP per capita) but also measures the quality of life by looking at life expectancy, education and purchasing power. It is a broader measurement of development because it tries to take into account people’s capabilities to lead a long life in which they can acquire knowledge and have access to the resources they need for a decent standard of living (McGillivray, 1991;
UNDP, 1990). The index is between 0 and 1, where 0 is least developed and 1 is most developed.
4.2.3 Moderating Variable
The moderating variable is involvement of NGO and considers whether an NGO was involved in the conflict between the MNE and the indigenous community. As discussed, there can be different types of involvement by third party actors. NGOs can work as mediator or facilitator and can also be actively involved in the solution by designing and/or implementing the solution (Arenas et al., 2013).The type of involvement is not taken into account in this study thus the variable involvement of NGO includes all types of involvements; pre or post entry discussions with an NGO, or an agreement was made with involvement of an NGO, or there was co-management of the project with an NGO. A dummy variable was computed to show whether an NGO was involved; (0) when there is no NGO involved in the conflict and (1) when there is an NGO involved.
4.2.4 Control Variables
Several control variables are also taken into account in this study. Control variables at four different levels are selected; country level, industry level, firm level and community level.
At country level the control variable level of corruption is selected because corruption comes forward as an important factor in violent conflicts (Oetzel et al., 2007). The variable is measured as one of the World Governance Indicators (WGI) computed by the World Bank. It
measures the control of corruption in a country and is expressed in percentile, where 0 is the lowest control of corruption and 100 the highest control of corruption. This variable is divided into 6 categories based on the percentile; 0-10th, 10-25th, 25-50th, 50-75th, 75-90th, 90-100th.
Since access to natural resources is often a source for conflict (Humphreys, 2005), the type of industry might be a variable that can influence the model. That is why industry is also used as a control variable. Industry is indicated by the North American Industry Classification System (NAICS) codes of the companies. The first two digits of this code indicate the industry that the company is active in. In this sample 62% of the cases are companies that are in the mining, quarrying and oil and gas extraction industry. The projects of companies from the extractive industries are extra vulnerable to community conflicts because they tend to have complex long-term projects that are very capital intensive (Laplante & Spears, 2008).
At the firm level, the amount of experience that the MNE has in a country is selected as the control variable because this can also influence conflict resolution between the MNE and indigenous community (Oetzel et al., 2007). The control variable experience is measured by the number of years that the MNE is active in the country of conflict.
The last control variable that is selected is at the community level; the geographic
outside world than others (Montenegro & Stephens, 2006). This can make it even harder for MNEs to negotiate with the communities and thus could possibly effect conflict resolution. This variable is measured by either (1) highly isolated, meaning that the members in the community live with almost no contact to the outside world, or (2) community members live within mixed broader populations.
4.3 Method of Analysis
The data is analyzed by using linear multiple regression analysis in SPSS. Regression analysis is a tool for predicting an outcome variable from one or more predictor variables (Field, 2009, p. 198). The multiple regression analysis can be described by the following formula:
𝑌! = 𝑏! + 𝑏!𝑋!!+ 𝑏!𝑋!!+ … + 𝑏!𝑋! + 𝜀!
In this formula the outcome variable Y is predicted by a combination of the different predictors (𝑋!!, 𝑋!!, … ) multiplied by the coefficients of the predictors (𝑏!, 𝑏!, … ) plus a residual term (𝜀! ) (Field, 2009, p. 210).
Since there are two dependent variables, the linear regression analysis to test the influence of development on a country and the moderating role of NGO involvement is conducted twice. All together six regressions models are created to analyze the relationship between the dependent, independent and moderation variables and to test all hypotheses. The
1, where the first model only includes the control variables as predictor. The second model will also include the independent variable and in the third model the moderation effect will be analyzed. The same is done in another three models with length of conflict as the dependent variable, as can be seen in Table 2.
Table 1. Regression analysis with dependent variable degree of violence
Level of corruption Industry MNE experience Geographic isolation
HDI score Involvement of NGO Degree of Violence Model 1 X X X X X Model 2 X X X X X X Model 3 X X X X X X X
Table 2. Regression analysis with dependent variable length of conflict
Level of corruption Industry MNE experience Geographic isolation
HDI score Involvement of NGO Length of conflict Model 4 X X X X X Model 5 X X X X X X Model 6 X X X X X X X
5. Results and Analysis
This chapter presents the results from the data analysis. It starts with an overview of the descriptive statistics, the correlation test and the multicollinearity of the different variables. This is followed by an explanation of the results of the regression analysis, which provides the information needed to conclude whether the proposed hypotheses in this study are supported or not.
5.1 Descriptive Statistics and Correlation Analysis
The results of the descriptive statistics and correlation test of all the variables are shown in Table 3. As can be seen in the table the mean for the dependent variable degree of violence is 2.87 with a standard deviation of 1.46. This means that on average the degree of violence falls into the range of a low level of violence, including peaceful protests and negotiations, and a high level of level violence, including physical damage but no kidnapping or deaths. Furthermore, the mean for number of months that a conflict lasts is 99.31 with a standard deviation of 79.53. This means that the average time that a conflict lasts is 99.31 months so that is 8.28 years. The dependent variable length of conflict, which is measured in three groups: (1) short, (2) medium and (3) long, has a mean of 2.20 with a standard deviation of 0.77, which means that on average the cases are in the medium scale (3 to 8 years) but there are more cases that fall into the long category (over 8 years) than in the short category (up to
of 0.15 and, for the moderator variable involvement of NGO, the mean is 0.43 with a standard deviation of 0.50. This means that in less than half the cases, 43.30% to be exact, there was involvement of an NGO in the conflict. The descriptive statistics of the control variables can also be found in Table 3.
Table 3. Descriptive statistic and correlations between variables
M SD Degree of violence Length of a conflict Number of
months HDI score
Involveme nt of NGO Level of corruption Industry type Years of experience Geographic isolation Degree of violence 2.87 1.46 Length of a conflict 2.20 0.77 0.214** Number of months 99.29 79.59 0.162 ** 0.786** HDI score 0.72 0.15 -0.262** -0.310** -0.194** Involvement of NGO 0.43 0.50 0.134** 0.102** 0.066 -0.099** Level of corruption 3.41 1.64 -0.247** -0.311** -0.183** 0.829** -0.064 Industry type 23.84 10.20 -0.002 0.003 0.090* -0.041 -0.048 0.004 Years of experience 32.64 80.11 0.011 -0.017 0.006 0.030 -0.031 0.013 0.215** Geographic isolation 1.67 0.48 -0.093* -0.073 -0.054 0.119** -0.085* 0.080* 0.073 -0.045
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
In order to do regression analysis, there can be no multicollinearity between the predictor variables (Field, 2009, p. 220). To test the variables for multicollinearity two different methods are used. First is to look at the correlations between the variables as are shown in Table 3. When correlations are above 0.8 this is considered as extremely high (Field, 2009, p. 224). Therefore, the significant correlation between HDI score and Corruption level of 0.829 can be an indication of multicollinearity between these two
variables. In order to see whether this is the case, the second method is used; the Variance Inflation Factor (VIF) shows whether a predictor has a strong linear relationship with the others predictors. When the VIF is higher than 10, this is considered to be severe multicollinearity (Field, 2009, p. 224). As can be seen in Table 4, there is no strong multicollinearity between the predictor variables because all the VIF scores are below 10. Even though there is a high correlation between HDI score and Corruption level there is no multicollinearity, thus it is not a threat to the reliability of the analysis. Therefore the variables HDI score and Corruption level are not excluded from the analysis.
Table 4. Multicollinearity and VIF
Dependent variable: Degree of Violence Length of conflict
Tolerance VIF Tolerance VIF HDI score 0.305 3.284 0.304 3.287 Involvement of NGO 0.981 1.019 0.979 1.022 Level of corruption 0.310 3.228 0.301 3.230 Industry type 0.936 1.068 0.935 1.070 Years of experience 0.946 1.057 0.945 1.058 Geographic isolation 0.969 1.032 0.964 1.037
As shown in Table 3 there are several significant correlations between the different variables. First of all there is a significant positive correlation between the independent variables length of conflict and degree of violence, which indicates that when conflicts last longer there tends to be a higher degree of violence as well. Second, the independent variable HDI score has a significant negative correlation with the dependent variables length of conflict and degree of violence. This indicates that cases in countries with a higher HDI score tend to have a lower degree of violence and a shorter length of conflict. Third, the moderator
involvement of NGO has a significant positive correlation with degree of violence and length of conflict, which means that cases that are longer and have a higher degree of violence tend to have more NGOs involved. Involvement of NGO also has a significant negative correlation with HDI score, which indicates that more NGOs are involved in countries with a lower HDI score. The control variable level of corruption has a significant negative correlation with both degree of violence and length of conflict. This indicates that countries that have a higher control of corruption tend to have shorter and less violent conflict. As mentioned before, level of corruption also has a significant high correlation with HDI, which indicates that countries with a higher HDI score also tend to score higher on the amount of control of corruption. Lastly, the control variable geographic isolation has significant negative correlations with degree of violence and NGO involvement, and significant positive correlations with HDI score and level of corruption. This indicates that more geographically isolated communities tend to have conflicts with a lower degree of violence and tend to have less NGOs involved. Also, communities tend to be more geographically isolated in countries that have a higher HDI score and higher level of control of corruption.
5.2 Regression Analysis
Results of the regression models with degree of violence as the dependent variable can be found in Table 5. This is done in a stepwise multiple regression analysis. First, in Model 1
