Power of indigenous communities and their conflicts with MNEs : the moderating effect of MNE experience

(1)

Amsterdam Business School

Power of indigenous communities and their conflicts with MNEs: the

moderating effect of MNE experience

MSc Business Administration International Management

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

Robin Klaas Jan Potze 10972811 28-01-2016

(2)

2

Statement of originality

This document is written by Robin Potze 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.

(3)

3

Abstract

Despite an increase in the amount of research in conflict situations between indigenous communities and multinationals, there is limited research that studies the effect of internal and external power of the community on the length and type of violence in conflicts. Therefore, this thesis builds upon previous studies that predominantly have been exploring conflicts qualitatively by identifying which factors of the communities and multinationals (MNEs) affect conflicts. We argue that community governance systems, size and MNE experience influence these conflict situations. By analysing 350 cases of conflict situations spread over the world, we explore how the governance systems (i.e. community’s concentration of power and contestation of power) and size (i.e. the number of members in the community) of the communities affect the length and type of violence of their conflicts with MNEs. Furthermore, this study examines the moderating role of multinationals’ experience. Our results show that the amount of contestation of power within the community influences the duration of the conflict. In addition, the experience of the multinational moderates the relation between community size and the type of violence during the conflict. By showing that the size of the community is irrelevant in the duration of the conflict; however, the MNE’s experience with the community is, our results have important policy implications for MNEs operating within these indigenous territories and opens avenue for further research.

(4)

4

Table of figures

Figure1 – Conceptual framework 32

Table 1 – Summary of the logistic regression models 38

Table 2a – Multicollinearity statistics for dependent variable length of conflict 39 Table 2b – Multicollinearity statistics for dependent variable type of conflict 39 Table 3 – Descriptive statistics, means, standard deviations, and correlations 43

Table 4 – Logistic regression analysis results 44

Table 5 – Ordinal regression analysis results 45

(6)

6

1. Introduction

Even with the existence of clear rules and regulations, conflicts between Multinational enterprises (MNEs) and indigenous communities cannot always be prevented. As Calvano (2008) proves with multiple real life examples, the local communities are the new battlegrounds where the activities of MNEs are challenged. For example, members of the First Nations community in Canada rarely have access to their native land. In the 1950s, the Canadian government closed parts of their land and turned it into a bombing range, which is currently the basis for five big oil companies. These agreements between governments and multinationals come at the cost of the native inhabitants . The First Nations communities eat fish and other wildlife in the area, since they are allowed to visit on weekends, but they are afraid that leaks from the oil companies contaminate the water and affect the ecosystem (Lim, 2015). Such conflicts are not limited to the First Nations community in Canada. Whether it is in Australia or South America, several indigenous communities around the world face conflicts with a multinational enterprise because of exploited economic opportunities (Fontana, Sastre-Merino & Baca, 2015). Such cases are frequently based on the acts of multinational enterprises (MNEs) in extractive industries. These firms are under pressure from various forces and movements regarding the environment, human rights and indigenous communities (Mutti, Yakovleva, Vazquez-Brust & Di Marco, 2012). This trend of conflicts has important consequences, both financial and reputational. Battles like these come at a great cost to MNEs because of project delays, damaged infrastructure, shut-down time and legal costs (Kemp, Owen, Gotzmann & Bond, 2011). The MNEs are aware of this fact, since local communities are often in a position to raise the costs of the MNEs’ projects as long as they possess the resources and the abilities to organize themselves. MNEs used to focus on communities in developed countries because of the aforementioned need for

(7)

7 resources and the ability to organize. However, it has become clear that this ‘truth’ is more complicated than they thought at first glance. In their active globalisation to developing countries, they experience resistance from communities which don’t have these direct resources and capabilities (Kapelus, 2002). Operating in these developing countries is a challenge in multiple ways since it also includes the problem of how to ensure that the local community can raise their concerns without needing extreme and violent means to that end. Peru did not have an agreement before 2008 to give the right indigenous communities to prior consultation and to free, prior and informed consent before a relocation from their lands (Bebbington & Bury, 2009). Besides this business view, these conflicts are also important in relation to the ethical and CSR issues and within the company.

However, the question remains as to which factors play a role in this matter of conflict. Preview literature explores qualitatively a few dimensions of the conflict, for instance, Calvano (2008) introduces the dimension of power, but she does not provide a complete explanation of the relation between power and conflict. As a result, based on power literature, it is likely that the power of the indigenous communities, which is measured by different constructs, influences the probability of conflict with a multinational enterprise. Different indigenous community governance structures have a different likelihood of conflict. In addition, the larger the group, the more powerful they perceive themselves and this results in more conflicts. Therefore, to truly understand the phenomenon of conflict between MNEs and indigenous communities, we need more in-depth comparative research so that we can examine the relationship between power of the communities and the conflicts with the MNEs. This study fills that gap by systematically analysing several power related variables, while taking a neutral perspective. This thesis will focus on the concept of power of indigenous communities and the leng th and type of

(8)

8 conflicts with MNEs. Power is here based on the size of the community as well as the contestation and concentration of power within the community and this leads to the following research question:

How does the power of an indigenous community affect the length and type of conflict with a multinational enterprise?

We take a neutral perspective and therefore it is not comprehensive to only examine the relationship between community power and conflict with MNEs , but also to understand the role of the MNE in this relationship and how it may influence the type of conflict. According to the literature on co-management, such strategies can offer the local communities a participation in resource decision-making. Co-management is defined as “the situation in which two or more social actors negotiate, define, and guarantee amongst themselves an equitable sharing of the management functions, entitlements, and responsibilities for a given territory or set of natural resources” (Borrini-Feyerabend Farvar, Nguinguiri & Ndangang, 2000:79). Second, MNE experience influences this liberating and empowering process for communities, state agencies and other stakeholders (Castro & Nielsen, 2001). Conflict can be very costly for the MNE in financial terms as well as in reputational damage. Thus, the MNE needs experience and skills to minimize struggles with indigenous communities to lower the costs of conflicts. Such skills are especially needed when the other party in the conflict is powerful, for example because of their large number and resources. The literature lacks an understanding of this moderating effect to reduce the probability and costs of conflict. Therefore, it is important to take into account MNE experience with indigenous communities and its moderation effect on the relation between community power and conflict. Thus, the second research question is as follows:

(9)

9

What is the moderating effect of experience of MNEs with communities on the relationship between community size and type of conflict?

While most studies in this field take a qualitative approach, we take a quantitative approach in establishing which indigenous governance and MNE characteristics are important factors in the conflicts they face. Thus, the purpose of this study is to show empirically how power of indigenous community relates to the length and type of conflict with MNEs. Second, we test the moderating effect of MNE experience with indigenous communities on the relation between size of the community and type of conflict. We will analyse approximately 350 cases from 41 countries and 18 industries. All data is secondary and is gathered through third party reports from organizations that study indigenous communities and from non-governmental organizations and then we coded it according to a coding scheme. First, we test the relation between contestation of power and length of conflict. Then, we test the relation between concentration of power and length and type of conflict. Third, we test the relation between the size of the community and length and type of conflict and finally, we test the influence of MNE experience on the relation between community size and conflict.

Our study makes several contributions to the existing literature on indigenous communities and MNEs. First, while most of the previous studies are of qualitative nature, we take a quantitative approach. Previous studies have mainly focused on one particular case of conflict and therefore fail to address the relationships between community and MNE characteristics across industries and countries. Second, we contribute to the existing literature by going more in depth in the construct of power, herein related to the division and governance of power inside the community as well as the external power measured in

(10)

10 group size. Finally, we show the impact of MNE experience on these conflicts. The results provide managers with a better understanding of important characteristics of indigenous communities. These relate to the power balance within a community as well as the size of the community. Managers should keep these factors in mind when they are about to enter a region with indigenous communities where conflict is likely. Furthermore, the moderating effect of experience provides the multinational insights in which factors influence the likelihood of violence during conflicts.

This study starts with a literature review of the main concepts in this matter, namely communities, MNEs, conflicts, power and MNE experience. Second, the theoretical framework will be posed as well as the hypotheses related to the theory. Third, the research method is discussed with an overview of the variables that are taken into account, the sample that is used and the method of analysis. Hereafter, the actual analysis with the results on the logistic and ordinal regression will be presented. Chapter six focuses on the discussion and the interpretation of the results as well as the implications and possibilities for further research. Last, this study concludes with an overview of the work done.

2. Literature review

2.1 Communities

A community is a cross-discipline concept and has therefore multiple definitions without a uniform definition in social sciences. Nevertheless, according to Calvano (2008) it includes dimensions of identity, geography and social interaction. This resulted in the definition of Brint (2001:8) who defines communities as: “aggregates of people who share common activities and/or beliefs and who are bound together principally by relations of affect, loyalty, common values, and/or personal concern (i.e., interest in the personalities

(11)

11 and life events of one another).” Since this definition excludes communities that share an instrumental or rational interest orientation (O’Mahony & Lakhani, 2011), a categorical definition was given by Dunham, Freeman and Liedtka (2006). They distinguish several types of communities. First, communities of place are defined by the physical proximity of the members, which is a common conceptualization of communities. For businesses it means that they should think about the effects of their actions in their near surroundings. Communities of interest connect through a similar focus or purpose and they can vary from religion based communities to groups that share a common hobby. With the important developments in communication technologies, the barriers that created distance between the members of such groups are removed. Nowadays, these groups can form and grow much faster than they used to do (Dunham et al., 2006). The third group is the virtual advocacy group. They are not agenda-driven, but they rely more on an oppositional identity, like activists and protestors. Again, the growth in technological possibilities made this possible. Lastly, the communities of practice are characterised by a strong feeling of identity, mutual obligation and they are open to learn and change.

A special kind of a community are indigenous people and they are important to MNEs. Indigenous people are defined in several ways. The United Nations Declaration on the Rights of Indigenous Peoples (2007), describes indigenous people as those who have suffered from historic injustices as a result of their colonization and dispossession of their lands, territories and resources, thus preventing them from exercising, in particular, their right to development in accordance with their own needs and interests. Thus, colonialists ignore their identity which causes the indigenous people to respond by seeking to merge with or incorporate themselves into the new group. Crawley and Sinclair (2003) describe this process of acculturation by four stages. In the first stage, the newcomers are expected to

(12)

12 adapt their appearance and interacting to the dominant norms. Separation, the second stage, explains that newcomers that differ culturally can get isolated by the organization. In the third stage both sides remain ill-defined when culture is neither pronounced nor influential. The last stage has the highest integration, since both cultures participate in a two way learning process (Dunham, Freeman and Liedtka, 2006).

2.2 Governance

It is critical for the existence of indigenous communities to organise themselves to address the challenges they face. The way non-indigenous governmental organisations either facilitate or frustrate these efforts of the communities is an important component of reconciliation. A combination of challenges and potential opportunities results in a fundamental task for the leader or leaders of the indigenous community. They have to design and exercise a governance system in their communities and develop the capabilities to do so (Hunt & Smith, 2005). Furthermore, the foundations of governance are influenced by cultural aspects like values, institutions, behaviours and motivations and therefore often vary from governance of non-indigenous communities. The National Centre for First Nations Governance (2009) defines governance as the traditions and institutions that communities use to make decisions and to accomplish goals. Norms, values, culture and language are the base of traditions, and institutions are the formal structures, practices and organization of an indigenous community. Governance is mainly about the creation of effective, accountable and legitimate systems and processes that the people of the community use to show their interests and exercise their rights and responsibilities. The National Centre for First Nations Governance (2009) provide seventeen related principles of effective governance practices for indigenous communities that are equally important. These principles can be categorized

(13)

13 in five overarching components, namely the people, the land, laws and jurisdiction, institutions and resources.

2.3 Conflict

Indigenous communities survive with a strong identity and their claims are diverse, although it is common that they seek the preservation and flourishing of their culture tied to their ancestral land. Through both domestic and international laws, the global community has recognized such claims and developed an appropriate regime for them. Policies foster cultural diversity and the protection of their threatened heritage, their language, rituals and land (Wiessner, 2011). The countless cases of conflict between indigenous communities and MNEs show that it is likely for multinational enterprises to engage in conflict. Therefore, it is remarkable that communities are not given the same level of priority as other stakeholders. It is important to note that not all conflicts lead to violence, many conflicts have a socio-environmental nature and concern the control of space, the governance of territory, access to land and water resources and defence of human rights (Bebbington, Hinojosa, Bebbington, & Warnaars, 2008). Independently of the amount of violence, conflicts can arise for various reasons and authors frequently note three elements that increases the probability of conflict. One of these elements is the nature of the industry; especially industries that have an impact on the environment are vulnerable. The money the communities get in return is often not enough to compensate the environmental loss (Fontana et al., 2015). Second, human rights abuse, the mistreatment of the communities and their land rights are also important factors. As noted by Urkidi and Walter (2011), the land of the communities used by MNEs is sometimes a physical and spiritual treasure. Conflicts occur despite the commitments that MNEs appear to make in their social responsibility programs. For example, Argentine stakeholders discovered that these

(14)

14 programs are used to protect their image instead of making a true commitment (Mutti et al., 2012). The third element is the amount of corruption that exist in the countries with an unstable political regime.

So far, the literature on conflicts between multinational enterprises and communities is not really comprehensive. Therefore, it is necessary to use concepts from other fields of study like literature on group conflict within organizations (Calvano, 2008). There are two principal factors for conflict regarding groups within organizations, namely the cognitive processes of individuals and the more structural characteristic of the group as a whole (Richter, Scully & West, 2005). Besides this distinction, the literature pays attention to the challenges for MNE managers to understand the economic and diverse cultures in which they operate and to make decisions in a country with different institutions and ethical rules than those in their home country (Buller, Kohls & Anderson, 1997).

According to Hook and Ganguly (2000), there are three theories that can be used to explain the conditions for conflict and ethnic political mobilization: the primordialist, the reactive ethnicity approach and the ethnic competition approach. The primordialist approach explains the critical importance to the fundamental and original identity of groups and their awareness in the ethnic political mobilization. Individuals in ethnic groups are born into a relationship network and therefore share common beliefs and objectives , but also share linguistic attributes. Thus, the identity of the group is given by nature (Rex, 1995). Furthermore, this approach explains how ethnonationalist awareness will grow when someone challenges and threatens their identity, which is the case when they get alienated from their home land. This alienation can be caused by MNEs that damage the environment such groups live in by extracting resources for example, which decreases the ability of these

(15)

15 groups to hunt and fish. Because the groups have to adopt to the new circumstances and provide themselves with new resources, they create a feeling of dislocation. Another example of how MNEs can change the local society where they operate, is when they change the local economy with payments and royalties,thereby creating winners and losers (Hook & Ganguly, 2000).

The second theory, the constructivist approach, has some contrasting elements with the primordialist theory. The constructivist approach neglects the viewpoint that nature causes the identity of such groups. This approach holds the belief that identity is a social construct where human actions and choices determine the identity. Ethnic identity derives from a cultural construction of descent which can be separated into social descent and genetic descent (Hook & Ganguly, 2000). Where social descent focuses on solidarity based on the feeling of being the same people, genetic descent focuses on the transmission of biological characteristics by genetic inheritance.

The third perspective is based on ethnic competition and states that competition for resources between self-serving ethnic elites drives ethnic conflict. In the competitive environment of ethnic groups that have the same needs and aspirations, it is more likely that groups compete harder with other groups for the same resources. Resource dependent MNEs like mining and oil companies exacerbate this situation. These companies bring new wealth into a region which creates the incentive for elites in a certain ethnic group to compete for a larger piece of the pie (Hook & Ganguly, 2000).

Fontana, Sastre-Merino and Baca (2015) take another perspective of conflict, by focussing on a methodology that integrates a strategic, political, social and ethical dimension to prevent conflicts. In their opinion, a team of professionals in social science should manage

(16)

16 the relation with the indigenous people. They have to gain trust and must build a bridge between the needs of the communities and the needs of the MNE. In the period of time and communication that follows, the aforementioned dimensions are important. Social learning is defined as “the openness of sensing the intrinsic qualities of others and understand their views” (Fontana et al., 2015: 8). To do so, communicating with the indigenous people requires a certain set of skills. In addition, a negotiating attitude is needed. This relates to the importance of creating a mind-set in which both sides feel they have the same interest. From this point, people are more flexible in the negotiation and are better able to understand each other. With regard to the political dimension, the local governance must complement the sustainability of the effort that MNEs make. The normative dimension focuses on the prevention of power abuse by the MNE and therefore, they need ethical codes. Last, the innovation dimension shows the importance of being open minded and the ability to let normal routines go (Fontana et al., 2015). Despite this strategy, it is more efficient for the MNE to internalize this knowledge instead of creating it when a conflict arises unexpectedly. Furthermore, Fontana et al. and other scholars, do not test the actual influence of the experience and knowledge as a moderator, leading to hypothesis four in this research.

According to Deutsch (1994), type and size can characterize conflicts. Some types of conflict situations are less suited for resolution than others, because these are seen as a conflict with a loser and a winner. An example of such an issue is having more status than the other side, but also gaining power or control over the other and having the exclusive right to something which does not have clear available substitutes. The latter is frequently the basis of conflicts between MNEs and indigenous communities. The other characteristic, size, is important because conflicts that escalate seem to grow in size. Likewise, smaller

(17)

17 conflicts are easier to resolve (Rubin, Pruitt & Kim, 1994). Conflict size can be defined as “being equal to the expected difference in the value of the outcomes that a party will receive if it wins compared with the value it will receive if the other wins a conflict” (Deutsch, 1994: 204). In this case, one of both sides experiences itself as the winner if it obtains the outcomes that are most satisfying. This means that conflict size can be both small and large, depending on the perception of the conflict by the parties. Conflict size is small when a party believes that both can win and large if one of the sides think that one party will lose while the other wins. Thus, one party may have the feeling that the outcome is satisfactory even if the other party wins, while the second party may perceive this completely different. Therefore, it is very important to control the issue of conflict. If the actors are able to control the importance of the stakes in the conflict, there is little chance that the conflict will take a destructive shape (Deutsch, 1994).

3. Theoretical framework

Although there are studies explaining how to deal with indigenous communities and conflicts, the question remains as to which factors play a role in conflict resolution and the type of violence during conflicts. Some of the conflict dimensions are qualitatively explored in the academic literature, but the focus mostly lies on single isolated conflicts as for example with Fontana, Sastre-Merino & Baca (2015) in Canada. Inequality of power between indigenous communities and multinationals is an important factor in conflicts (Calvano, 2008). This puts communities often in a disadvantage when they confront multinationals. However, we do not know its underlying forces and its exact relation to conflict. Therefore, we need more comparative in-depth research. A quantitative examination of characteristics of the indigenous communities as well as the MNEs could shed light on the aspects of power

(18)

18 that determine the length of conflict and the type of violence. Length of conflict is related to great cost for the multinationals because of project delays, damaged infrastructure, shut-down time and legal costs (Kemp, Owen, Gotzmann & Bond, 2011). Type of violence is important in examining conflicts because smaller conflicts are easier to resolve (Rubin, Pruitt & Kim, 1994). Furthermore, if the actors are able to control the stakes in the conflict, there is little chance that the conflict will take a destructive shape (Deutsch, 1994). This study argues that it is likely that the power of the indigenous communities, which is measured by different constructs, influences the probability of conflict with a multinational enterprise. Different indigenous community governance structures have a different likelihood of conflict. Powerful people and groups act more efficient and faster towards goals (Galinsky, Magee, Gruenfeld, Whitson & Liljenquist, 2008; Slabu & Guinote, 2010). Furthermore, group cohesion and unity are important in the resolution of conflicts and therefore we have to look at the power governance systems of the communities (Bonoma, 1976; Stein, 1976; Tajfel & Turner, 1979). Second, local communities possess governmental functions which are connected to the exercise of power and authority (Meggitt, 1964b). This can reduce the length of conflict and the type of violence. Third, the larger the group, the more powerful they perceive themselves and this results in more and more violent conflicts (Guinote, Brown & Fiske, 2006). Last, to truly understand the phenomenon of conflict between MNEs and indigenous communities we need to examine the variables that explain the relation between power of the communities and the probability of conflict with MNEs. This study fills that gap by systematically analysing several power related variables, while taking a neutral stance. Last, the prevention of conflicts with indigenous communities is of great importance to MNEs. Therefore, experience is needed to overcome such liabilities, which can be gained by preliminary visits, meetings and trainings (Jassawalla, Truglia & Garvey, 2004; Tung,

(19)

19 1987). More MNE experience should decrease the chance of violent conflicts. Aforementioned relationships will be explained in further detail below.

3.1 Power

Bonoma (1976) places power, conflict, cooperation and trust in a context irrelevant manner as a general theoretical explanation for multiple phenomena at for instance the personal level, group level, organizational level and societal level. The frequently used stakeholder salience model helps to understand how firms balance the dynamic competing interests of both internal groups and external groups (Mitchell, Agle & Wood, 1997). Power, legitimacy and urgency are the basis for corporate decision making and influence stakeholder groups. Stakeholders that are in the position of possessing all these variables and as much of them as possible will get the most attention. In the case of MNEs in conflict with indigenous communities, these variables are not easy to obtain for the communities and most times they just don’t possess them. Looking at power, communities usually have less power than other stakeholders. Furthermore, MNEs do not see the claims of indigenous communities as legitimate. Nevertheless, because some of the conflicts get media attention, urgency can be created in the conflict. Most of the time, however, this is not enough for communities to influence decision making in the process (Calvano, 2008).

Another theory that tries to explain the stakeholder prioritization of managers is the organizational life cycle approach of Jawahar and McLaughlin (2001). Three criteria have to be met to receive managerial attention. The stakeholders, in this case the indigenous communities, have the potential to satisfy critical organizational needs. In addition, they have to satisfy the rights needs at the right time in the life cycle of an organization. Third, the claims of the communities need to be more important in relation to the claims of other

(20)

20 stakeholders. Applying this model to the communities, again they don’t measure up to the criteria. Their claims are usually not as important as those of the other stakeholders in an organization. When we put this in the context of conflicts, the needs of the communities only receive attention if they threaten or support the survival of the particular firm, need an emergency response or influence the bottom line (Calvano, 2008).

Bonoma (1976) argues that the constructs of power, conflict, cooperation and trust need more investigation in relation to social occurrences. First, it seems that these occurrences do not fall into one grand conceptual framework, therefore Bonoma (1976) states three different concepts explaining their underlying relationship. Second, it becomes clear that power, conflict, cooperation and trust do not share a common theoretical linkage within the three interaction concepts. Last, there are some strong advantages to developing a good middle range conceptualization of power, conflict, cooperation and trust. The three concepts as stated by Bonoma (1976) are the unilateral power system, the mixed (bargaining) power system and the bilateral power system. The relative position of the actors separate these systems from each other. Looking at the relation between indigenous communities and MNEs, the former two are more relevant in this research.

3.1.1 Unilateral power systems

Fundamentally, the unilateral power systems requires just three elements for construction, a source or influence initiator, a communication or message system and a target that is influenced. The source or influence initiator is, by having access to the modes of influence, usually the strongest actor in the relationship and therefore it can direct the course of the interactions (Bonoma, 1976). Most research on influencing relationships is about actors as a weak target or as a strong and powerful actor. Therefore, the concept of

(21)

21 unilateral power must incorporate the fact that one of the involved actors is overpowering the other. This can result in tactical movements like intimidation, bribery and information control. Furthermore, conflicts subsist as long as the source and target have similar preferences regarding some commodities that are in relatively short supply. Or when both parties want to exist in a certain area, but there are not enough governmental, monetary or ecological resources available for both. The conflict that follows can lead to a situation where both sides are worse or better off, or where one receives more at the expense of the other (Bonoma, 1976).

3.1.2 Mixed power system

The second power system relevant for conflicts between indigenous communities and MNEs is called the bargaining interaction or mixed power system. As discussed, there is a low level of mutuality and dynamism during interaction in the unilateral power system. However, the opposite is true for the mixed power system since both sides are capable of making offers to the other party, threatening the other side but also proposing an acceptable solution for both sides. According to Bonoma (1976), the people who have the most power and are seen as legitimate and strong, must perceive some of the resources possessed by the protestors as valuable enough to bargain for. Thus, the meaning of power differs from its unilateral counterpart and so does the meaning of conflict. Where the reason for conflict in a unilateral power relation is a shortage of a resources, in the bargaining relationship it is about resource distribution. From a cooperation perspective, it means that there are more social elements in a mixed power system since there must be an agreement between both parties based on multiple bargaining rounds and suggestions for possibilities. Hereafter, the agreements need to be legitimized by a contract for example (Bonoma, 1976).

(22)

22 There is a clear agreement in the literature of various academic fields, stating that external conflict increases the internal cohesion and unity under a few conditions. These conditions find their origin in the nature of the conflict and the nature of the group. First, there must be a threat in the conflict that affects the complete group and the members of the group equally. Second, a solution must be possible to create a useful motivation for the group to fight the external threat. Last, it is important that there is cohesion and unity in the group. That means that a higher unity in the group will cause longer conflicts (Stein, 1976; Tajfel & Turner, 1979). Furthermore, research has shown that communities with high-power individuals think and act more towards an active goal than low-power people (Galinsky et al., 2008; Slabu & Guinote, 2010). In the case of conflict between indigenous communities and MNEs, the communities are low-powered, because they lack resources and coherence as a group in comparison to the MNE. Studies about goal pursuit show for example that the high-power group is more efficient and flexible in their pursuit and even faster to act on opportunities (Magee & Smith, 2013). Furthermore, the social distance theory predicts the high-power groups have greater self-control than the low-power group. According to Magee & Smith (2013:167), “Self-control is exercised when individuals prioritize more important, longer term motives over less important, shorter term motives when the two motives directly conflict”. The individuals need to have thoughts that focus on the long term, important outcome instead of the short term focus. In this line of reasoning, we expect the indigenous community to be less focused on the conflict. Therefore, the conflict will have a shorter duration. This relation between unity and length of conflict is also shown in the case of the Kankuamu, an indigenous community in Colombia. They organized themselves very well and this strengthened their cultural identity, unity and autonomy as a community. They

(23)

23 started a conflict to regain their land rights and this long lasting conflict eventually resulted in a satisfying agreement (Roy, Tauli-Corpuz & Romero-Medina, 2007).

Aforementioned literature and the example show that different power systems can cause conflict. Some of the indigenous communities have a high contestation of power and because of a lack of unity, we hypothesize that conflicts with such communities are shorter than conflicts with indigenous communities that show coherence and unity.

H1: Communities with a high contestation of power will have shorter conflicts with MNEs than communities with low contestation of power.

3.2 Community power

Local communities possess governmental functions which are connected to the exercise of power and authority (Meggitt, 1964b). According to a study on Australian indigenous communities, the Australian aboriginal societies display varieties of social grouping despite their demographically small scale. For example, tribes with up to 1.200 members can have governance structures that include both a patrilineal part and patrilineal clans, a matrilineal group and matrilineal clans, endogamous groups of grouped g eneration levels, sub-sections and age grades (Meggitt, 1964a). Thus, the indigenous communities can have several structures. Some communities have a medicine man, who derives his status from curing people from the community. Furthermore, in many tribes there is a system of age-grading, most of the time based on shared religious experience. The members of a respected age group receive their respect because they have more experience than the juniors. Usually, a different person in the group will be the central character in the problem that has arisen in the community. The likelihood of fluctuation in this matter in various social

(24)

24 situations would naturally mean the emergence of permanent leaders in the community (Meggitt, 1964b).

According to MacGill (2013), in dominance based hierarchies, those higher in hierarchy are able to control those that are lower in the hierarchy. We see this phenomenon of controlling leaders especially in more evolved species, like humans. These dominant leaders were in position during the time of great empires in the form of kings and pharaohs for example and the phenomenon is still relevant in our modern society. Such hierarchies see the world as parts that can be separated and exploited without thinking about the interconnectedness. Contrasting to this hierarchy based structure, there are so-called acephalous groups that have little or no structural leadership (Currie, Greenhill, Gray, Hasegawa & Mace, 2010). Characteristics of acephalous groups are fairer and egalitarian structures and a recognition of interdependence. Thus the members of acephalous societies have more to say in their organisation. Whereas the dominance hierarchy structures focus on setting goals, the acephalous groups favour group relationships (Brocklesby, 2007).

Autocracy and democracy are frequently viewed as opposites in a spectrum, because there are clear differences in the way power is acquired and exercised, how social order is defined and maintained, and how much non-executive peoples’ opinions influence the decision making process (Marshall & Cole, 2008). However, both are stable and useful in maintaining order, but are nevertheless seldom practiced perfectly. Therefore, they are opposites on a governance scale. Perfect democracy is characterized by institutionalized procedures to facilitate political participation and the power of the chief executive is balanced. Autocracy on the other hand, restricts political participation and executives exercise their power without obstruction from legal, judicial or societal institutions (Marshall

(25)

25 & Cole, 2008). Anocracy is in contrast to these stable environments, less capable of maintaining stability and continuity. Such regimes are more vulnerable to instability, outbreaks of armed conflict, changes in leadership or unexpected seizure of power. Thus, groups with a governance mechanism like this have more going on inside their own community and we predict that they are less focused on outward conflict with a multinational enterprise. Furthermore, a group that exists for a longer period of time, like the indigenous communities, needs cohesion and consensus to endure in a conflict. In addition, strong leadership that can authoritatively enforce the cohesion of a group is presumed to have longer lasting conflicts. Especially if not all the member of the community feel the external threat (Stein, 1976). The !Khomani San indigenous community in South Africa exemplifies the importance of a well-organized concentration of power during conflict. They lacked a clear leadership structure and there was no hierarchy at all. The stakeholders thought and hoped that the community would have the collective wisdom to make good decisions. However, this was not the case and many errors were made, whichreduced the morale of the community. Eventually, because the community lacked an appropriate concentration of power, the conflict was relatively short and the community got depressed and their farms are currently in a bad shape (Roy, Tauli-Corpuz & Romero-Medina, 2007). This example shows that it is important for a community to have a well-established high concentration of power to endure in conflict.

H2a: Communities with a high concentration of power will have longer lasting conflicts with MNEs than communities with low concentration of power.

H2b: Communities with a high concentration of power will be more likely to have violent conflicts with MNEs than communities with low concentration of power.

(26)

26

3.3 Group size

Explaining the relation between group size and conflict is still important in multiple academic fields, for example in studies about intra-national conflict and civil wars (Humphreys, 2005; Brunnschweiler & Bulte, 2009). Many researchers use group size to operationalize power and these studies show that an increasing group size influences the chance of conflict (Guinote, Brown & Fiske, 2006). Such an operationalization of power and the relation between group size and the idea of control is also confirmed by literature about the strength and safety of numbers (Latané & Wolf, 1981; San Martin, Swaab, Sinaceur & Vasiljevic, 2015; Zdaniuk & Levine, 1996). A feeling of power as described can be embedded in the institutional power of a group and a larger group size can help to achieve the goals of the group. A higher population of a group is an important resource in conflict between groups and it increases the likelihood of success for any collective effort.

As Kabanoff (1991) points out, a party which overpowers the other, makes more attempts to influence the other party and the other way around. In addition, these attempts tend to be more successful for the more powerful party. The powerful side also communicates more and receives more communications in return (Shaw, 1981). Another important finding is that powerful groups pursue their goals more actively and they perceive gaining a larger share of the outcomes of conflict as justice. This is the application of the equity rule, which contends the beliefs of weak and strong groups’ of their worth and capacity to satisfy their desires. Therefore, distributions that favour a stronger group are believed to be fair (Kabanoff, 1991). Studying the relation between group size and both intra- and inter-nation conflict, Stavig and Barnett (1977) found significant evidence of group size having influence on several forms of conflict. For example, it relates strongly to verbal activity, including protests, threats and accusations. However, there is also evidence of

(27)

27 increased violent activity. With an increasing community size, the probability of troop movements and the total number of people killed increase as well. Although this research focused on conflicts between nations, the same line of reasoning can be used for conflicts between indigenous communities and MNEs.

An example of a large indigenous community involved in a long and violent conflict to fight for their rights is the case of the Jumma people in Bangladesh. They fought for their rights under the leadership of their chieftains and armed themselves. The conflict started in 1989 and ended in 2001 (Roy, Tauli-Corpuz & Romero-Medina, 2007). Thus, examples and academic studies about power show that groups which feel stronger and more powerful are more likely to face conflict situations. Therefore:

H3a: Larger communities will be more likely to have longer lasting conflicts than smaller communities.

H3b: Larger communities will be more likely to have violent conflicts than smaller communities.

3.4 MNE Experience

To understand the moderating effect of MNE experience on the relation between power of indigenous communities and conflict, a closer look at the outcomes of collaboration is needed. One of the possible solutions to defuse the conflict is co-management, which is defined as “the situation in which two or more social actors negotiate, define, and guarantee amongst themselves an equitable sharing of the management functions, entitlements, and responsibilities for a given territory or set of natural resources” (Borrini-Feyerabend Farvar, Nguinguiri & Ndangang, 2000:79). Despite the fact that many arrangements focus on restrictions on natural resources, the main

(28)

28 purpose might not be the protection of the resources. The goals of such an arrangement may be different from those that characterise modern and scientific resource management (Holm, Hersoug & Rånes, 2000). Co-management can offer the local communities a participation in resource decision making and it can be a liberating and empowering process for communities, state agencies and other stakeholders which is influenced by MNE experience (Castro & Nielsen, 2001). The co-management regimes are applicable in a worldwide sense and in several fields of resources like forests, wildlife, fisheries and other protected areas. Many situations involve state agencies that share resource allocation or management responsibilities with indigenous communities. The fundament of co-managing is that sharing authority and decision making will make the process of resource management easier (McCay and Jentoft, 1998).

Conflicts are the basis of many agreements and in some cases these conflicts are settled in court or government offices, but it is also likely that the battle is fought on land and sea or in the media. Either way, the source of co-management agreements lies in conflicts and these include the fight of indigenous communities to resist state and private resource appropriation, to defend their local livelihood and to maintain their identity. However, these conflicts cause social, personal and financial costs for both parties and it is still an important reason to negotiate arrangements (Castro & Nielsen, 2001).

Kozan (2002) states five other conflict management styles based on assertiveness and cooperativeness. First, the community or the MNE can opt for competing, which is high on assertiveness and it involves using power to win at the other party’s expense. This shows the importance of power, in this case the power of the indigenous community, in relation to conflicts. The second style is called accommodation, based on high cooperativeness, which

(29)

29 seeks to satisfy the needs of the other side at the expense of your own needs. Third, the collaboration or problem solving style is distinguished, which is high on both dimensions. It brings all the issues together and then tries to find a solution to satisfy both parties. Fourth, the avoiding style is low in both dimensions and it means that both parties run away from negotiating with each other. Finally, the compromise is based on a middle solution in both dimensions and it is about giving up some issues to receive something in return.

Some early models of conflict resolution strategies, show that these strategy styles can be explained on a single dimension scale that ranges from selfishness to cooperativeness. However, such dimensions fail to include the possibilities that involve high concern for both, or low concern for both dimensions. Therefore, several authors use a conflict management framework with five styles, as described before (Thomas, 1992). These styles are placed on a grid with two axes; cooperativeness, which is about attempting to satisfy the other’s concerns. The other axis is assertiveness, which is about attempting to satisfy one’s own concerns. Two styles are especially important in this research, namely avoiding and competing. Avoiding is a management style which is low in both own and other concern. Morris et al. (1998) show that the competing style shows high own concern and low other concern. These styles are likely to cause friction in a relationship and this explains why these styles are the focus in cross cultural conflict management. This is also an important factor in the early work of Rubin, Pruitt and Kim (1994). They provide an overview of the nature of conflict, a range of strategies which are based on a psychological background and the fertile ground of conflict escalation. One of the reasons for escalation of conflict is the inexperience and lack of knowledge of people in a crisis situation. Substantive knowledge could help to see solutions that might not be thought of by other conflicting parties and it can help to assess other proposed solutions more realistically. So, the

(30)

30 participants in a conflict need skills to establish a cooperative relationship with each other that can focus on problem solving. This will help them to develop the ability to look at the conflict from another perspective so that they avoid getting stuck in the unproductive and even destructive traps of conflicts (Deutsch, 1994). This experience and knowledge is based on what did and didn’t work and is influenced partly by their native culture. For example, Japanese disputants are expected to avoid direct confrontation before addressing a conflict because that is part of their culture (Wall & Callister, 1995).

According to Morris et al. (1998), cultures have different signalling languages and negotiators from MNEs that face a counterpart from another culture easily misread signals or send unintentional messages. This happens when they face indigenous groups as well. Some authors note that one of the factors that leads to possible conflicts, is the context in which the MNE operates. Context is here explained as, among other things, the history of the region or country, the existing groups of interest and the representativeness of the communities (Lotila, 2010). Lacking this analysis often relates to the cultural distance between the MNE and the community. Furthermore, this distance is also reflected in the way the MNE communicates with the community. It is hard for the indigenous people to understand the technical and industry specific argument that the MNEs bring forward.

Literature on expatriates from multinational companies also shows the importance of culture in conflicts with indigenous communities and their resolution. Scholars agree that expatriates must be comfortable with the values, attitudes and behaviours that are the norm in the other country (Caligiuri, 2000). If they aren’t comfortable, the chance is higher that they will reject the other culture as well as its people and create a feeling of being opponents (Sanchez, Spector & Cooper, 2000). Therefore, experience is needed to overcome

(31)

31 such liabilities, which can be gained by preliminary visits, meetings and trainings (Jassawalla, Truglia & Garvey, 2004; Tung, 1987). Cultural literacy, as discussed by Lertzman and Vredenburg (2005), is the skill of adaption and two-way learning. To share power and to overcome conflict with indigenous people, MNEs have to build relationship that are guided by traditional authority, experience and learning. Social learning is one of the ways to build the aforementioned relationships and it is defined as “learning that occurs when people engage one another, sharing diverse perspectives and experiences to develop a common framework of understanding and basis for joint action” (Schusler, Decker & Pfeffer, 2003: 311). It is important because managing conflicts requires knowledge that reflects the values of the community. Thereby, in the process of social learning, both individuals and groups improve their understanding of the issues that are relevant in conflict, the areas where an agreement can be reached and their own and other’s values. As Schusler et al. (2003) explain, this is indispensible in conflicts with indigenous communities, especially in the extraction industries. These conflicts can lead to high losses in all kinds of resources. Indigenous conflicts in Canada show this theory in practice. After multiple occasions of conflicts, the actors involved get around the table to establish an agreement so that the MNE and the indigenous communities can operate close to each other. With this experience they can build on their relationship and prevent future clashes. Thus, the choice for a management style is based on experience, thoughts and knowledge from either side. Therefore, we predict that the experience of a MNE influences the relation between group size and conflict because they are more capable of managing the tensions between MNE and the indigenous community.

H4: MNE experience will influence the relation between community size and conflict with MNE.

(32)

32

3.5 Conceptual framework

Figure 1: Conceptual framework

4. Methodology

4.1 Data collection

The data that was collected for this study consists of several cases from countries all over the world, covers multiple continents and has been codified by certain characteristics. All data is secondary and was gathered through third party reports from organizations that study indigenous communities and from non-governmental organizations. The data was selected on availability and the reliability of the source. Furthermore, additional information was needed to identify certain MNE characteristics like country base, ownership structure and firm industry. This information was retrieved from the Orbis database. The Orbis database provides information about companies worldwide and is updated monthly. One of the variables that had to be coded is the GINI index, which is the extent to which the distribution of income within a country’s economy deviates from an equal distribution (World Bank, 2015). Lastly, the Human Development Index score, which is needed as a control variable, was retrieved from the United Nations Development Programme. The HDI

(33)

33 is a summary measure of average achievement in three dimension of human development (UNDP, 2015).

4.2 Sample

The data was collected together with other students who use the same coding scheme of conflict situations and together we constructed the dataset. In total, the sample consists of 348 cases about conflicts between indigenous communities and MNEs and the number of observations as well as the variables are appropriate for this research. The cases are from Australia, Papua New Guinea, Chile, Guatemala, Colombia, Nicaragua, Panama, Mexico, Honduras, Ecuador, Peru, Nigeria, Bolivia, Argentina, Venezuela, Paraguay, Costa Rica, the Dominican Republic, Madagascar, Tanzania, El Salvador, Sweden, Finland, South Africa, Lesotho, Namibia, Botswana, Zimbabwe, Mozambique, Malawi, Indonesia, the Philippines, New Zealand, the United States, Canada, Vietnam, Cambodia, Lao PDR, Thailand, China and Malaysia. The industries under consideration are agriculture, forestry, fishing and hunting; mining; utilities; transportation and warehousing; professional, scientific, and technical services; manufacturing; wholesale trade; arts, entertainment, and recreation; accommodation, and food services. The relationship between indigenous people and their power with the length and type of conflict is analyzed and therefore the analysis is on the firm level. A conflict is classified as interactions between opposing forces that range from minor disagreements to escalated or violent conflict (Kemp, Owen, Gotzmann & Bond, 2011). The research design is cross-sectional, because the data was collected for a specific period of time.

(34)

34

4.3 Dependent variables

The first dependent variable used in this research is the length of conflict between a MNE and an indigenous community. Length of conflict relates to great cost for the multinationals because of project delays, damaged infrastructure, shut-down time and legal costs (Kemp, Owen, Gotzmann & Bond, 2011). Quick conflict resolution is beneficial for the multinational and therefore length of conflict is a good variable to measure successful resolution of conflicts. Based on the agreed coding scheme, it was measured by noting the month and year of the beginning and the end of the conflict. Afterwards, these numbers were used to compute the total length of the conflict measured in months. For ongoing conflicts, the number of months was calculated by taking November 2015 as end date, since this was the last month of data collection. Then, a new variable was created to identify cases that could be considered either long or short. By using the median length of conflict, a cut-off point was identified. This point was 77 months or almos t 6.5 years. Therefore, the chosen cut-off point was 78 months. Ongoing conflicts that had a duration between 71 months and 78 months were deleted from the dataset because it couldn’t be sure if they would end within half a year or not.

The second independent variable in this research is type of conflict. This was coded as a categorical variable which ranges from peaceful negotiations between the community and MNE to a high level of violence including deaths. It was created as a seven point scale that increases in the level of violence. If the conflict went through multiple stages of violence, the highest level was used in the coding scheme. This variable works because smaller conflicts are easier to resolve (Rubin, Pruitt & Kim, 1994). Conflict size can be defined as “being equal to the expected difference in the value of the outcomes that a party will receive if it wins compared with the value it will receive if the other wins a conflict”

(35)

35 (Deutsch, 1994: 204). Thus, it is important to control the issue of conflict. If the importance of the stakes in the conflict are controlled, there is little chance that the conflict will take a destructive shape (Deutsch, 1994).

4.4 Independent variables

The governance mechanism of indigenous communities is hypothesized to affect the length and type of conflict with MNEs. Two separate dimensions were used to construct this independent variable.

The first dimension is concentration of power, which indicates the governance type of a community. Cohesion and unity are important in a group. This means that a higher unity in the group will cause longer conflicts (Stein, 1976; Tajfel & Turner, 1979). Furthermore, research shows that communities with high-power individuals think and act more towards an active goal than low-power people (Galinsky, Magee, Gruenfeld, Whitson & Liljenquist, 2008; Slabu & Guinote, 2010). Based on the agreed coding scheme, the data for this dimension was coded from for example ethnographic studies of the community as a stage in continuum. The first stage is acephalous communities that have little or no structural leadership (Currie, Greenhill, Gray, Hasegawa & Mace, 2010). This stage is followed by democratic communities, age-group based, elder councils and monarchical communities. Thus, five types of power concentration communities were identified and this coding has worked in previous studies.

The second dimension, contestation, describes the type of dissent within the community and this dimension was also coded as a stage in a continuum that increases in level of dissent. The stages are: no contestation, some dissent within community, strong

(36)

36 communities that lacked an appropriate concentration of power, conflicts were relatively short (Roy, Tauli-Corpuz & Romero-Medina, 2007). This example shows that it is important for a community to have a well-established high concentration of power to endure in conflict. The coding schema was tested in previous studies regarding this topic as well and was proven to work.

Furthermore, the influence of a third independent variable on length and type of conflict was tested. This variable is community size, which was coded in the number of individuals that are part of the indigenous community. Studying the relation between group size and both intra- and internation conflict, Stavig and Barnett (1977) found significant evidence of the group size on several forms of conflict. It is , for example, strongly related to verbal activity, including protests, threats and accusations. However, there is also evidence of violent activity. With an increasing community size, the probability of troop movements and the total number of people killed increase as well.

4.5 Moderator

The moderator in this research is MNE experience which influences the relation between community size and conflict. With experience, multinationals can build on their relationship and prevent future clashes. The choice for a management style is based on thoughts, knowledge and thus experience from either side (Schusler, Decker & Pfeffer, 2003 Jassawalla, Truglia & Garvey, 2004; Tung, 1987). Therefore, the experience of a MNE is predicted to influence the relation between group size and conflict because they are more capable of managing the tensions between MNE and the indigenous community.

This construct was separated in two dimensions, namely experience in number of conflicts and number of years. The number of conflicts is a count number of projects around

(37)

37 the world in territory of indigenous communities since its start of the operations. The number of years reflects the amount of experience the MNE has in dealing with indigenous communities, starting from their first encounter until 2015. These moderations were turned into interaction variables to measure its influence on the relation between the size of the community and the chance of a high level of violence during the conflict.

4.6 Control variables

The control variables are industry and development of country. To code the firms industry, the 8 digit code of the NAICS was used. To make the codes useful for analysis, the NAICS codes were restructured into five categories: mining, utility, manufacturing, agricultural/forestry and a category with the remaining industries. Furthermore, to control for the development of the particular country, the Human Development Index scores was used. These are defined as follows for country j:

𝐻𝐷𝐼_𝑗 = 1 − (∑ 𝐼_𝑖𝑗/𝑘

𝑘 𝑖=1

)

Where 𝐼_𝑖𝑗 is the ith indicator of human deprivation in country j. The deprivation of country j was assessed based on the following variables, life expectancy, adult literacy and the

logarithm of purchasing power adjusted GDP per capita (McGillivray, 1991).

4.7 Method

A linear regression analysis was used in this thesis to show the effects of the independent variables on the dependent variables. To analyse the effect of the independent variables on length of conflict, a logistic regression was used and the models can be found in table 1. The effects of the independent variables on degree of conflict was tested by an ordinal regression. The essence of a regression analysis is to predict the outcome of the

(38)

38 dependent variable from one or multiple independent variables. The model that was tested is linear, which means that the data set is summarized by a straight line (Field, 2009).

To examine the variables that could influence conflicts between indigenous communities and MNEs, several variables and characteristics of the communities and firms were codified. All the selected cases were analysed and the variables needed were coded in an appropriate way and suitable for data analysis tools. The database was created by several students and cases were selected from all over the world. In total, 348 cases were identified and codified. The total dataset was constructed together with other project members.

Table 1 – Summary of the logistic regression models

Control variables Independent variables

HDI score Industry

Concentration of power

Contestation of

power Size of community

Model 1 x x

Model 2 x x x

Model 3 x x x x

Model 4 x x x x x

5. Results and analysis

Table 3 shows both the descriptive statistics as well as the correlation between the variables that are examined in this research. The table shows a mean value for deg ree of violence of 4.03, therefore it can be concluded that in general there is a high degree of violence in conflict situations over the world. Furthermore, the descriptive statistics provide a mean of 2.58 for concentration of power, implying that power tends to be more concentrated on one person in indigenous communities. Last, the mean for contestation of power is 1.48. In this variable, 1 means that the distribution of power is not contested in an

Power of indigenous communities and their conflicts with MNEs : the moderating effect of MNE experience

Amsterdam Business School