Conflicts between Indigenous Communities and
MNEs: the Impacts of Host Country Development
Level and Indigenous Community Isolation, MNE
Experience as a Moderator
Master Thesis
MSc Business Administration
International Management Track
Name: Yunni Yang
Student number: 11652020
Date: 21-06-2018
Thesis supervisor: Dr. Ilir Haxhi
Second reader: Dr. Markus Paukku
1
Statement of Originality
This document is written by Yunni Yang 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.
2
Abstract
Although investments in extractive industries have increased economic growth in many developing countries, the interaction between the multinational enterprises (MNEs) exploiting important natural resources and the local indigenous communities have often resulted in conflicting relationships. Previous studies on conflicts between the MNEs and indigenous communities have analyzed the dynamics of conflict resolution only from a single-case perspective by taking a qualitative research approach missing out on the big picture of country-, community- and firm-level effect on the resolution of these conflicts. We argue that the level of development of host country institutions and community isolation will impact the conflict resolution between MNEs and indigenous communities, as well as the MNE experience will moderate these relationships. We test our predictions for a sample of 297 conflict cases from 50 countries worldwide. Our results show two main findings: first, a higher developed host country, and second, a less isolated community lead to shorter and less violent conflicts between MNEs and indigenous communities. The contribution of this study is twofold: first, methodologically by applying a quantitative method it offers a broader perspective of country-, community- and firm-level effect; and second theoretically, it extends the current conflict literature and studies by examining the direct effects of host country development level and community isolation on conflict resolution, and the moderating effect of MNE experience on these relationships. This study also offers valuable insights for MNE managers on how to deal with conflict dynamics more effectively.
Keywords: indigenous communities; MNEs; resource conflict; conflict resolution; HDI; isolation; prior experience
3
Table of Contents
1. Introduction ... 5
2. Literature review ... 12
2.1 Indigenous communities ... 12
2.2 Indigenous community isolation ... 14
2.3 Conflicts between MNEs and indigenous communities ... 15
2.4 National development ... 17
2.5 MNE experience ... 19
3. Theoretical framework ... 21
3.1 Mechanisms influencing conflict resolution ... 21
3.2 Host country development level ... 24
3.3 Indigenous community isolation ... 26
3.4 MNE experience ... 29
3.5 Conceptual model ... 33
4. Methodology ... 34
4.1 Sample and data collection ... 34
4.2 Dependent variables ... 35
4.3 Independent variables ... 36
4.4 Moderator ... 38
4.5 Control variables ... 39
4.6 Method of analysis ... 40
5. Results & Analysis... 43
5.1 Descriptive Statistics ... 43
5.2 Correlation Analysis ... 45
5.3 Multicollinearity Test ... 47
5.4Regression Analysis ... 49
4 6.1 Findings ... 57 6.2 Methodological Implications ... 61 6.3 Theoretical Implications ... 62 6.4 Practical Implications ... 63 6.5 Limitations ... 64 6.6 Future Research ... 65 7. Conclusion ... 67 8. References ... 70
5
1. Introduction
Since the 1990s, the economic globalization has changed the world a lot. Many
multinational enterprises (MNEs) started their international business out of their
home countries. Aiming at profit maximization, MNEs, especially those in extractive
industries, usually enter less developed countries for cheaper raw materials and labor
force, and abundant natural resources. Territories which indigenous communities use
or live in become a continuous target of the developing extractive industries at a
global scale (Savino, 2016). Although FDI in extractive industries have brought
dramatic economic growth for many developing countries, extractive MNEs exploit
important natural resources from nearby local communities, leaving direct and
negative impacts on people’s livelihoods (Damonte & Glave, 2012). Opposition and
resistance from indigenous communities followed the economic development
(Paredes, 2016). These conflicts between MNEs and indigenous communities do
harm to interests of all the parties involved. Therefore, it is crucial to find out the
factors that can affect the MNE-indigenous conflict resolution.
These years, a lot of scholars have studied the influential factors in the
worldwide conflicts between MNEs and indigenous communities. Previous studies
on conflicts between the MNEs and indigenous communities have Although these
studies present some important insights, they analyze the dynamics of conflict
resolution only from a single-case perspective by taking a qualitative research
approach, missing out on the big picture of country-, community- and firm-level
6
Olin, 2014). To complement the current literature, this study applies a quantitative
approach to analyze conflict resolution between MNEs and indigenous communities,
from a multilevel perspective.
The first important actor in this conflict resolution is the host country. Country is
where the indigenous communities are embedded, and some characteristics of these
communities are derived from their own country. Many researchers have proved that
host country plays a significant role in the dynamics of conflict resolution. Conde and
Le Billon (2017) examine the effects of several state-related factors – policies,
criminalization, inadequate planning, and corruption – on the degree of local
community resistance to mining projects. Hess and Orphanides (1995) explain the
potential reasons why economic growth can affect the occurrence of conflicts.
Helwege (2015) mentions that the unfair negotiations between poor governments and
extractive companies can lead to local conflicts. Although these studies have
considered certain country-level characteristics as potential factors in conflict
resolution, few of them discuss about this effect systematically. There is still much
space for us to discover. To fill up the gap, we present our first research question:
Research Question 1: What is the effect of host country development level on conflict resolution (i.e., length and violence) between the MNEs and Indigenous communities?
This question aims to study whether longer and more violent conflicts between
MNEs and indigenous communities will occur in a less developed host country. In
7
which cannot constrain the behaviors of traders effectively (North, 1991). Moreover,
poor governments have limited bargaining power with extractive companies,
resulting in an unbalanced benefit distribution among citizens and companies. As a
result, local conflicts emerge more frequently in poor countries (Acuña, 2015).
Therefore, it can be supposed that host country development level has a significant
effect on conflict resolution.
The second important actor in conflict resolution is indigenous community.
Indigenous community refers to “any community of people who currently live within
the borders of the Republic, or who historically lived in the geographic area currently
located within the borders of the Republic” (Intellectual Property Law Amendment
Bill, 2010, in Geyer, 2010: 132). Conde and Le Billon (2017) investigate the effects
of some community-related factors – marginalization, mine dependency, territory,
alliances, and distrust – on the possibility of resistance to mining projects by local
communities. Most times, conflicts occur when the interest of indigenous
communities and MNEs collide in a specific geographic area (Calvano, 2008).
When the indigenous communities are located in a geographically isolated
region, or they are embedded with a totally different language or culture from that of
dominant societies, situations could be worse. Their rights and identifications may be
rarely recognized by local governments or MNEs, leading to difficult relationships
between the communities and those two parties (Ballard & Banks, 2003). In addition,
there could be a big gap in knowledge and culture between isolated indigenous
8
likelihood of conflicts. Therefore, we believe that the isolation of indigenous
communities can have an impact on conflict resolution. Some current literature has
already studied the effects of the characteristics of indigenous communities on
conflict resolution (Arellano, 2011; Damonte & Glave, 2012), but few studied the
impacts of specific factors such as the community isolation. To fill up this research
gap, here comes our second research question:
Research Question 2: What is the effect of the isolation of indigenous communities on conflict resolution (i.e., length and violence) between the MNEs and Indigenous communities?
The purpose of the second question is to investigate whether more isolated
indigenous communities will be involved in longer and severer conflicts. Isolation
hampers communities’ access to the important resources which grant them authority
and identity (Bebbington, 1999). Lack of contact and knowledge between an MNE
and an indigenous community can possibly cause mutual misunderstandings and
conflicts. Therefore, it is important to find out the potential effect of community
isolation on conflicts. Moreover, this result is valuable for MNEs to make FDI
strategies. If it is proved that shorter and less violent conflicts occur in less isolated
indigenous communities, MNE managers will know it is more beneficial to operate in
these regions.
Finally, the third important actor in conflict resolution is MNE. We choose one
of the company characteristics, MNE experience, to analyze its effect on conflict
9
and the indigenous communities, and the previous conflict experience. It is
considered as a competitive advantage for MNEs in high-risk environments (Henisz
& Delios, 2004; Holburn, 2001; in Getz & Oetzel, 2009).
Studies on international issues often consider prior international experience as a
significant factor in MNEs’ success because it can facilitate new investment
(Andersson, Johanson, & Vahlne, 1997). Reade and McKenna (2013) argue that
foreign subsidiaries of MNEs can leverage indigenous knowledge to perfect their
conflict management system, and to create culturally relevant ways to deal with
conflicts between managers and local employees. Thus, it is important to take into
account the moderating effect of MNE experience on conflict resolution. To our
knowledge, however, existed literature mentions little about how MNE experience
acts in conflict resolution. This gap leads to our third research question:
Research Question 3: What is the moderating effect of MNE experience on the relationship between host country development level/community isolation and conflict resolution (i.e., length and violence) between the MNEs and Indigenous communities?
This question intends to explain whether experienced MNEs will encounter
shorter and less severe conflicts in more developed host countries and/or less isolated
indigenous communities. Experienced MNEs know better than the unexperienced
how to acquire legitimacy and survive in a less developed country that usually has
weak institutions. Also, they know more about indigenous communities so as to
10
reduce the related costs. If the results prove the moderating effect, it will inspire
MNE managers that they can benefit more if they learn before they do business,
especially in underdeveloped host countries and isolated communities.
To answer the research questions and test the hypotheses, this study will apply
297 cases from 50 countries across the whole world to run the statistical analysis via
the software SPSS. We hypothesize that longer and more violent conflicts will occur
in less developed host countries and/or more isolated indigenous communities, and
that MNE experience will act positively in moderating these relationships.
This study contributes to current conflict literature in three ways. First, this
study contributes methodologically. Most studies on conflict resolution use a
qualitative method while this study applies a quantitative analysis. Second, from a
theoretical perspective, it complements the current conflict studies by investigating
the conflict resolution from a multilevel perspective. Specifically, this study extends
the existed literature by discovering the rarely-discussed effects of both the host
country development level and indigenous community isolation on conflict resolution,
as well as the unexplored moderating effect of MNE experience on these
relationships. Many previous articles focus on isolated-level factors in conflict
resolution while this study manages to combine three-level factors into one analytical
model. Lastly, from a practical perspective, our results will bring valuable insights to
MNE managers. This study provides them with a better understanding of conflict
dynamics, and thus helps them make better strategies, especially in less developed
11
This study is structured in the following way: first, a literature review about
current research on related topics will be presented; second, based on our three
research questions, we will then develop a conceptual framework to show our
hypotheses and justify the relationships between these factors; third, the methodology
will be introduced, including the sample and data collection, dependent variables,
independent variables, the moderator, control variables, and the method; forth,
12
2. Literature review
2.1 Indigenous communities
There is no consensus on the definition of community in the social sciences
(Calvano, 2008), but it is agreed by scholars that community includes dimensions of
geography, social interaction, and identity (Hillery, 1955; Lee and Newby, 1983; in
Calvano, 2008). Studies can define community according to where a group of people
live, what common background or shared interests a group of people have, or whether
the group is the public or society in general (Geyer, 2010). Dunham et al. (2006)
classify community into four subcategories: community of place, community of
interest, virtual advocacy groups, and community of practice. Among these four
subcategories, community of place refers to those communities in which people live
in physical proximity, and it is the most commonly used concept of community by
stakeholder theorists (Dunham et al., 2006; Calvano, 2008). Dunham et al. (2006)
also argue that many different, and often competing, communities of place can live
together within one geographic region, and it will be a big challenge for MNEs to
develop a good understanding of all those communities, and thus to interact with
them appropriately.
There are different standards to define indigenous people in the world (Sanders,
1999). In the field of law, the word “indigenous” is often used to describe the origin
of a legal object. According to the Intellectual Property Law Amendment Bill (2010,
13
who currently live within the borders of the Republic, or who historically lived in the
geographic area currently located within the borders of the Republic.” Based on an
analysis of UN documents regarding indigenousness, Geyer (2010: 136) then
summarizes the characteristics of indigenous people: they have already lived in a
region before the region is colonized or conquered, or before the present national
boundaries are established; they have distinctiveness; they have their own customs
and/or traditions, laws, language, religion, etc; they are “determined to preserve their
own institutions, customs and so forth”; and they identify themselves as “being
distinct/tribal/indigenous”.
However, the meaning of indigenous community in Africa is not focused on the
originality and aboriginality, but includes some other features (Gilbert, 2011).
According to the report of Working Group1(n 33, chapter 4 89, in Gilbert, 2001: 251),
groups in Africa who identify themselves as indigenous people have different culture
(some is under extinction) and way of life from those in the dominant society; they
survive by access to lands and natural resources; they suffer from discrimination that
their places are less developed than other dominant sectors of society; they often live
inaccessibly and are geographically isolated; and they “are subject to domination and
exploitation within national political and economic structures”.
1
There is an organization called African Commission, on Human and People’s Right. Working Group is this Commission’s own Working Group of Experts on Indigenous Population/Communities, for short.
14 2.2 Indigenous community isolation
According to existed literature, scholars usually divide isolation into two forms:
geographic isolation and cultural isolation (Van Cott, 2000; Pringle, 2014; Correa &
Pavez, 2016). Indigenous communities often choose strong geographic isolation
voluntarily, because people are worried that modernization will hamper their own
customs, traditions, and cultures (Tang & Tang, 2001). For example, many
indigenous populations in the region of Amazon choose to live far in the remote
Amazon headwaters and avoid all contacts with the outside society (Pringle, 2014).
This phenomenon is contradictory to the argument of Dahl and Jensen (2002) who
emphasize the urban embeddedness and close interaction of indigenous communities.
Some indigenous tribes will become vulnerable if they try to move and contact the
outside world (Pringle, 2014). Therefore, these indigenous people “remain isolated
out of fear”, not because they want to “avoid all other human societies” (Pringle,
2014: 126).
Isolation hampers communities’ access to the important resources which grant
them authority and identity (Bebbington, 1999). Commonly, the habitats of
indigenous people are not only geographically isolated, but also in great poverty
(Sanders, 1999; Correa & Pavez, 2016). These people possess few economic and
educational resources, leading to a big gap in knowledge and culture between
indigenous people and dominant societies (Correa & Pavez, 2016). Furthermore, to
some extent, geographic isolation will intensify cultural isolation. Alston and Kent
15
hampers their social and psychological development and then worsens their social
exclusion. For example, the Mapuches-Huilliches community, one of the indigenous
communities, lacks technical assistance from the outside and lacks the ability to use
and manage the natural resources around them, because of their remote location
(Oltremari & Jackson, 2006). In Bolivia, indigenous communities lack continuous
governmental support and a judiciary which does well in legal pluralism or
multiculturalism, mostly because of their geographic and cultural isolation (Van Cott,
2000).
There is a significant cultural divide between urban, rural, and traditionally
remote-oriented indigenous communities (Brasche & Harrington, 2012). Different
indigenous communities in one region can bear different level of isolation. For
example, Shan-Mei residents in Taiwan Province have a tight social relationship with
outsiders and share common goals with each other, while Li-Chia people who also
live in Taiwan Province, strongly fight against any form of outside intrusion (Tang &
Tang, 2001).
2.3 Conflicts between MNEs and indigenous communities
The relationship between multinational enterprises (MNEs) and indigenous
communities has become tense these years (Calvano, 2008). Because of the economic
globalization, many MNEs transfer their resource and production centers to poorly
16
2008). At the same time, these host countries open their gates to MNEs for better
development and residents’ better standards of living (Calvano, 2008). Territories
where indigenous communities live become a continuous target of the growing
extractive industries at a global scale (Savino, 2016). Extractive industries such as
mining put indigenous people into asymmetric conditions, and raise global
competition over natural resources as well as pressures on the natural environment of
these local people (Paredes, 2016). Indigenous people are not treated equally by their
governments and extractive MNEs, therefore, the main opposition to MNEs’
activities and business has shifted from host countries’ governments to the indigenous
communities living in these countries (Calvano, 2008). As a result, indigenous people
raise conflicts out of resistance to the extractive operations, raise their opposition by
“framing their struggle as a quest for self-determination, territorial autonomy, and
decoloniality” (Savino, 2016: 404), some of whom make chaos also to gain benefit
from the extractive companies (Penman, 2016).
In terms of influential factors in the conflicts, based on previous conflict
literature, Conde (2017) summarizes four reasons for mining conflicts: environmental
burden, lack of participation in mining projects and lack of rights to determine own
development path, insufficient monetary compensation for land or resource losses,
and distrust to mining companies and government. In addition, Pinkey (2005)
indicates that MNEs themselves play a significant role in the conflict situation. MNEs
in the extraction industries are supposed to have stronger conflicts with local
17
(2008) examines the impacts of three components: stakeholder power inequality,
stakeholder perception gaps, and cultural context; he concludes that all of three
components intensify the conflicts between MNEs and indigenous communities.
Some researchers find that characteristics of indigenous communities can play a
significant role in the MNE-Indigenous conflict dynamics (Arellano, 2011; Damonte
& Glave, 2012), but they do not specify the factors. Acuña (2015) disagrees that the
conflict is simply a result of ill-designed policies, but he believes the conflict will
occur because of the historical colonial domination which denies the ontology of
indigenous people.
Since the conflict is a dynamic process and varies over time (Pondy, 1967),
Paredes (2016) indicates three factors that affect the dynamics of the contention of
indigenous communities to extractive industries: technological changes within the
industries, centrality of extractive industries in national economy, and new ways of
opposition of rural communities. A combination of these three factors leads to “the
glocalization of mining conflicts” (Paredes, 2016: 1047).
2.4 National development
It is commonly believed that most indigenous people live in rural and remote
areas, and they often suffer considerable social and economic disadvantage compared
to urban people (Fuller et al., 2007). For example, there is a huge gap in the living
18
(Psacharopoulos, Patrinos and Oaxaca, 1995). Indigenous poverty is “deep-rooted,
widespread and persistent” (Kumar, 2009: 247), which hinders the improvement of
standards of living as well as the recognition of their identities and rights.
Development is necessary to improve the economic, human and social situation of
these indigenous people (Fuller et al., 2007).
In underdeveloped countries, the social and legal institutional frameworks are
weak, which cannot constrain traders’ behaviors effectively (North, 1991).
Developing countries often have poor ex ante negotiations with extractive companies
(Helwege, 2015). In the negotiations, these national governments give away resource
rights on behalf of their local communities, and often charge much less than they will
lose (Helwege, 2015), leading to an unfair benefit distribution between companies
and local people. As a result, local communities will raise conflicts with both
extractive companies and their national governments due to their poorly defined
rights and unfair internal distribution (Helwege, 2015).
In addition, when people cannot express their needs in institutionalized ways,
which often happens in poor countries, the only thing they can do is conflict and
protest (Conde, 2017). Polachek and Sevastianova (2012) conclude that poor
countries suffer from more wars than rich countries. Results of Pinstrup-Andersen
and Shimoawa (2008) also show that in countries whose people suffer income
poverty and poor health status, armed conflicts are more likely to happen. Collier et al.
(2003, in Braithwaite, Dasandi & Hudson, 2016: 46) present the concept of “conflict
19
development, which in turn increase the occurrence of future conflicts. Therefore,
conflicts often interrelate with national growth (Braithwaite et al., 2016).
2.5 MNE experience
MNE experience plays a significant role in MNEs’ foreign operations as well as
the resolutions in difficulties and barriers. For MNEs operating worldwide, one big
challenge is to understand economic conditions and diversified cultures of various
countries (Buller & Kohls, 1997). They need to make specific strategies based on
characteristics of different nations. Studies on international issues often consider prior
international experience as a significant factor in MNEs’ success because it can
facilitate new investment (Andersson, Johanson & Vahlne, 1997).
Some researchers believe that experience enables companies to learn and
perform better in similar operations in the future (Cyert & March, 1963; Levitt &
March, 1988; in Zeng et al., 2013). They consider experience as a competitive
advantage for MNEs in high-risk environments (Henisz & Delios, 2004; Holburn,
2001; in Getz & Oetzel, 2009). The capability of managing risks is a “valuable, rare
and costly-to-imitate resource” which an MNE can leverage in its operations and
competition with other companies (Getz & Oetzel, 2009: 378).
In addition, experience allows companies to better adapt to local environment, to
overcome liabilities of foreignness and newness, and thus to reduce the cultural and
20
Kaufmann, 2015). Reade and McKenna (2013) indicate that foreign subsidiaries of
MNEs can leverage indigenous knowledge to perfect their conflict management
system, and to create culturally relevant ways to deal with conflicts between
managers and local employees. Even if MNEs do not have previous experience about
the local context, they can cooperate with NGOs that are knowledgeable and
experienced to understand the local communities (Kolk & Lenfant, 2013). This
enables MNEs to innovate and improve their conflict resolutions in foreign regions
(Lertzman & Vredenburg, 2005).
Some other researchers, however, argue that previous experience may not
always be helpful since companies may make a wrong decision based on a wrong
inference from experience (Feldman, 1986; Huber, 1991; in Zeng et al., 2013). Up to
now, the actual effect of prior experience on future investments has not been clearly
21
3. Theoretical framework
The first section of this part introduces the mechanisms that influence conflict
resolution between MNEs and indigenous communities. The following three sections
focus on the influential factors in MNE-Indigenous conflict resolution and put
forward related hypotheses. The second section explains how host country
development level can affect conflict resolution between MNEs and indigenous
communities while the third section explains the effect of community isolation on this
type of conflict resolution. The forth section discusses the moderating effect of MNE
experience on the relationship between host country development level and
MNE-Indigenous conflict resolution, and the relationship between community
isolation and MNE-Indigenous conflict resolution. The final section presents the
conceptual model of this study.
3.1 Mechanisms influencing conflict resolution
Resource extraction industries always have a direct negative impact on the
natural environment. Industrial development does harm to the lands and lives of
indigenous peoples (Lertzman & Vredenburg, 2005). To protect their territories and
rights, indigenous people often raise conflicts against those extractive companies,
which has become a great challenge for many companies. This phenomenon leads to
an increasing growth in literature about conflict resolution between MNEs and
22
these conflicts (Calvano, 2008; Arellano, 2011; Damonte & Glave, 2012; Acuña,
2015; Savino, 2016; Paredes, 2016). However, most of these researches study with a
qualitative approach and focus on specific cases in certain geographical areas.
Literature on the dynamics of conflict resolution from a multi-level perspective is
limited. To fill up these gaps, therefore, this study tries to quantitatively investigate
influential factors in conflict resolution at multiple levels.
MNEs’ overseas operation can be influenced by both external and internal
environment. This study intends to analyze the factors of both sides in one construct.
Country characteristics and community characteristics constitute the external
environment of MNEs. Different external environment forces MNEs to make
different strategies and operations, which thus lead to different outcomes, sometimes
different degrees of conflicts. Therefore, this paper aims to find out the direct effects
of one of the characteristics of host countries, national development level, as well as
one of the characteristics of indigenous communities, community isolation, on the
MNE-indigenous conflict resolution.
On one hand, underdeveloped countries often have weak social and legal
institutions, which cannot constrain people’s behaviors effectively. When extractive
MNEs enter these countries, governments often have poor benefit-cost negotiations
with MNEs, ending up with an unfair outcome (Helwege, 2015). This will lead to an
imbalance in the benefit distribution between companies and local people. Local
communities, as a result, will raise conflicts with both extractive companies and their
23
distribution (Helwege, 2015). Therefore, it is meaningful to examine the exact effect
of host country development level on conflict resolution.
On the other hand, it is proved in several current conflict studies that certain
aspects of indigenous communities will contribute to the conflicts between
indigenous communities and MNEs (Damonte & Glave, 2012; Acuña, 2015).
However, few of them study about the specific effects of community isolation.
Isolation hampers communities’ access to the important resources which grant them
authority and identity (Bebbington, 1999). Commonly, the habitats of indigenous
people are not only geographically isolated, but also in great poverty (Sanders, 1999;
Correa & Pavez, 2016). They possess few economic and educational resources,
leading to a big gap in knowledge and culture between indigenous people and
dominant societies (Correa & Pavez, 2016). MNEs from outside societies may have
difficulty in communication with culturally isolated communities. Therefore, it is
easy for foreign MNEs to encounter conflicts with isolated indigenous people who
are embedded with different cultures. Community isolation could be an effective
factor in MNE-indigenous conflict resolution.
Besides the effects of the external environment, it is significant to consider
together with the impact of the internal environment of MNEs, which is the
characteristics of MNEs. We will choose one of the company characteristics, MNE
experience, as the factor of the internal environment as well as the moderator.
Conflicts between MNEs and indigenous communities can vary in duration and
24
leverage indigenous knowledge to perfect conflict resolution through culturally
relevant ways (Reade and McKenna, 2013). These subsidiaries will have a better
understanding of the host country environment as well as the features of indigenous
communities than other subsidiaries. Therefore, when conflicts occur, experienced
companies are able to handle these conflicts better than inexperienced companies. We
can thus believe that experience can moderate the effects from the external
environment and strengthen companies’ abilities to survive.
In short, this study states that host country development level and isolation of
indigenous communities both have a significant effect on the conflict resolution. To
be more exact, we believe that longer and more violent conflicts will occur in less
developed host countries and/or more isolated indigenous communities. In addition,
we argue that MNE experience will act positively in moderating these relationships.
In other words, experienced MNEs can handle the conflicts with indigenous
communities better than inexperienced MNEs in less developed host countries and/or
more isolated indigenous community regions.
3.2 Host country development level
Scholars find two ways in which poverty affects conflicts, economically and
politically. Do and Iyer (2010: 736) find that poverty is one of the determinants of the
intensity of conflict, because the “opportunity cost of conflict is lower in poorer
25
Laitin (2003), however, present that poverty increases the possibility of conflicts
because poverty is usually associated with a weaker state. This kind of state has weak
financial and military capabilities, increasing the likelihood of success of rebel people
(Braithwaite, Dasandi, & Hudson, 2016). Collier et al. (2003, in Braithwaite, Dasandi
& Hudson, 2016: 46) present the concept of “conflict trap”, indicating that a country
in conflicts will suffer a reversal of economic development, which in turn increase the
occurrence of future conflicts. Existed conflicts cannot be resolved properly, making
conflicts longer and severer.
Country is where the indigenous communities are embedded, and some
characteristics of these communities are derived from their own country. There are
two ways in which the host country development level can negatively affect the
conflict between MNEs and indigenous communities. First, less-developed countries
often depend more on foreign investments than on self-production, so that
governments in these countries provide MNEs with more privileges and benefits than
their own indigenous citizens. These local people cannot acquire their deserved rights
and identifications, so they make conflicts to express their resistance to their
governments and MNEs. Second, less-developed countries usually have weaker
institutions and governance. Laws in these countries cannot be implemented
effectively (North, 1991), which increases the possibility of collisions and conflicts.
When a conflict has occurred, even the government is not able to solve it well, not to
say the foreign MNEs, which will make the conflict longer and more violent. Based
26
country has a significant effect on the length and violence of the conflicts between
MNEs and indigenous communities in this host country.
In underdeveloped countries, poverty constrains people’s access to education
and health care (Pehn, 2009). Insufficiently educated communities are likely to
misunderstand MNEs’ behaviors and owe their sufferings to MNEs. They struggle
and persist, but cannot organize and advocate against governments’ poor policies
effectively (Olson, 1971). As a result, the conflicts continue and become severer. In
relatively richer countries, indigenous communities can benefit from the sound and
forceful institutions and better education, which facilitate their interaction with MNEs.
Efficient communication can reduce conflicts’ occurrence, duration and severity.
Thereby the hypotheses are following:
H1a: The more the host country is developed, the shorter the conflict between the MNE and an indigenous community will persist.
H1b: The more the host country is developed, the less violent the conflict between the MNE and an indigenous community will be.
3.3 Indigenous community isolation
Conflict occurs when there is difference in values and interests, vagueness in
responsibility, poor mutual communication, and unwillingness to new changes
(Upreti, 2004). Calvano (2008) also states that conflict between MNEs and
27
particular geographical area. There are two types of isolation, geographical isolation
and cultural isolation (Van Cott, 2000; Pringle, 2014; Correa & Pavez, 2016).
Geographically isolated communities live far from the broader societies and have no
contact with the outside world. They are usually poor and have few basic services in
their residence, including the conflict resolution services from governments (Upreti,
2004). Remote location hampers their access to the important living resources
(Bebbington, 1999). In addition, geographically isolated communities usually have a
low literacy rate (Upreti, 2004), which can lead to cultural isolation.
Geographically isolated indigenous communities rely heavily on their natural
environment for food and housing. As a result, extractive actions will do more harm
to the living environment of these isolated people than that of dominant communities.
When the extractive MNEs break in, companies exploit the abundant natural
resources for their own business, depriving and polluting the important living
resources of the local people without any compensation. To protect their rights, those
local people protest against MNEs. The more isolated communities will have more
opposition to the damage MNEs have done, thus leading to bigger and severer
conflicts. It is also harder to repair the loss for more isolated areas because of poor
infrastructure, so the conflicts between more isolated communities and MNEs will
last longer.
Culturally isolated communities have different customs, traditions, languages
and cultures from the broader mainstream societies. People from the two parties
28
indigenous people cannot get sufficient education and training because of their
isolated geographical location (Alston and Kent, 2003). When MNEs operate with
this type of isolated indigenous communities, it is hard for companies to adapt to the
local social environment due to cultural distance. For example, it is hard to find
enough well-educated and skilled employees. Also, when a conflict occurs, it is hard
for MNEs to communicate with indigenous people, which thus worsens the conflict.
MNEs need more time to adapt to the indigenous environment and deal with an
indigenous community when the indigenous community is more isolated, either
geographically or culturally. When more time passes, longer conflicts will get more
violent and cause more damage.
There is an example about the community isolation. According to Tang and Tang
(2001), Shan-Mei and Li-Chia are two villages in Taiwan Province. Residents in
these two villages have different attitudes towards the outside world. Shan-Mei
people have a tight connect with the outsider society, and they implement a successful
fishing ban to their residents just as the dominant societies do. With the help of this
distinct ban, Shan-Mei people have few conflicts with other communities on fishing.
In contrast, Li-Chia villagers are more indigenous and they avoid building
relationship with the outside world. They don’t know what rules they should follow.
Once they break the rules, they refuse to pay the fine, which can lead to a bigger
conflict. It takes them a long time to resolve the conflict. This example shows that
community isolation does matter in influencing the conflict resolution. Therefore, we
29
H2a: The more isolated an indigenous community is, the longer the conflict between an MNE and the indigenous community will persist.
H2b: The more isolated an indigenous community is, the more violent the conflict between an MNE and the indigenous community will be.
3.4 MNE experience
Unfamiliarity with host country environment makes MNEs’ foreign operations
difficult in the host country (Alcantara & Hoshino, 2014). Foreign subsidiaries have
to face more difficulties and barriers than head office at home. And things get more
challenging when foreign subsidiaries are operating in a less developed country.
However, experience can mitigate the risks. Through host country experience,
companies can acquire the knowledge about business system, institutional structure
and also the culture of the host country (Alcantara & Hoshino, 2014). MNEs can
improve their knowledge base, capabilities and competitiveness through experiential
learning in the host country (Chen, 2017). Experience allows foreign companies to
better adapt to local environment, to overcome liabilities of foreignness and newness,
and thus to reduce the cultural and administrative distances (Henisz & Williamson,
1999, in Reimann, Rauer & Kaufmann, 2015). Thus these companies are able to
organize and manage themselves better in the host country.
As mentioned before, a less-developed host country usually has a weak and
30
behaviors effectively (North, 1991). If an MNE enters a less-developed host country
without either experience dealing with the indigenous communities in that country or
knowledge about the laws and rules of the host country, it will be easy for the MNE
to encounter conflicts with the local communities, even at a disadvantage, thus
causing huge economic loss.
In contrast, if an MNE has had business in a less-developed host country before,
and has dealt with one of the indigenous communities in that country, the MNE will
better survive and gain legitimacy in this imperfectly instituted country and
understand the reason and purpose of the conflicts created by the indigenous
community. Equipped with previous experience, MNEs can handle future conflicts
with this indigenous community more smoothly and effectively, and even avoid
potential conflicts through some positive and legal firm behaviors such as CSR.
Experienced MNEs will need less time to adapt to national and indigenous
environment and to come up with good conflict resolutions, and will minimize the
loss and damage that conflicts could cause. Therefore, MNE experience can moderate
the effect of the host country development level on the length and violence of the
conflicts between MNEs and indigenous communities. Thus, we address our
hypotheses:
H3a: The more an MNE has prior experience dealing with an indigenous community, the more positively it will moderate the effect of host country development level on the length of conflict between the MNE and the indigenous community (i.e. the conflict will be shorter).
31
H3b: The more an MNE has prior experience dealing with an indigenous community, the more positively it will moderate the effect of host country development level on the violence of conflict between the MNE and the indigenous community (i.e. the conflict will be less violent).
Conflict is a result of human behavior, so it is important to take into account
effective interaction, mutual learning and understanding, and cooperation with
humans when dealing with a conflict (Upreti, 2004). Collective learning is significant
in conflict resolution (Walker & Daniels, 1997). To build a good relationship with
local communities, companies need to investigate local needs, interests and social
conditions, and integrate their knowledge and skills into the conflict resolution
(Upreti, 2004). Hence they can have harmonious relationships with local people,
which reduces the likelihood as well as the duration and violence of conflicts.
Isolated indigenous communities have little contact with the outside world.
Their languages, traditions and cultures may be very different from those in the
mainstream societies. If an MNE begins to operate in an isolated region where the
MNE has never been before, the MNE will not have any experience and knowledge
about the indigenous communities in this region. Without experience and knowledge,
the MNE knows nothing about the culture, traditions, demand, preferences,
consumption habits and social condition of the indigenous community, and thus it
will be very likely for the MNE to make wrong strategies and sell wrong products,
32
handle these conflicts without knowing anything, which will lead to longer and more
violent conflicts.
In contrast, if an MNE has equipped with sufficient knowledge and previous
experience about an indigenous community before it enters the community, and has
transferred these knowledge and experience among its staff, the whole company will
learn from the previous experience how to better deal with indigenous community
people as well as possible conflicts. Reade and McKenna (2013) indicate that MNEs’
foreign subsidiaries can leverage indigenous knowledge to improve their conflict
management system, and to create culturally relevant ways to deal with conflicts. For
example, by hiring bi-culturally trained employees, and learning about the culture,
demand and preferences of the indigenous community, MNEs can serve and
communicate with the community more efficiently, making conflicts shorter and less
severe. Collective learning also helps MNEs to know local communities better and
look at conflicts from a local-person perspective (Upreti, 2004). In this way, these
MNEs are able to handle the conflicts with the indigenous community with more skill
and ease, and to avoid useless action for conflict resolution, thus moderating the
negative effect of indigenous community isolation on the length and violence of the
MNE-indigenous conflicts. Therefore, our hypotheses are:
H4a: The more an MNE has prior experience dealing with an indigenous community, the more positively it will moderate the effect of indigenous community isolation on the length of conflict between the MNE and the indigenous community (i.e. the conflict will be shorter).
33
H4b: The more an MNE has prior experience dealing with an indigenous community, the more positively it will moderate the effect of indigenous community isolation on the violence of conflict between the MNE and the indigenous community (i.e. the conflict will be less violent).
3.5 Conceptual model
Figure 1 shows the conceptual model of this paper, including the relationships
between all the chosen variables in our study. It illustrates clearly the hypothesized
direct effects of host country development level and indigenous community isolation
on the length and violence of an MNE-Indigenous conflict, and the moderating effect
of MNE experience on these relationships.
34
4. Methodology
4.1 Sample and data collection
This study examines the potential effects of host country development and
indigenous community isolation on the MNE-Indigenous conflicts, as well as the
moderating effect of MNEs’ experience on these relationships. The sample consists of
297 cases, covering 50 countries across the world: Argentina, Australia, Bolivia,
Botswana, Brazil, Cambodia, Cameroon, Canada, Chile, China, Colombia, Congo,
Costa Rica, Dominican Republic, DRC, Ecuador, El Salvador, Ethiopia, Finland,
Guatemala, Honduras, Indonesia, Kenya, Laos, Lesotho, Liberia, Madagascar,
Malaysia, Mali, Mexico, Myanmar, New Zealand, Nigeria, Panama, Papua New
Guinea, Paraguay, Peru, Philippines, Senegal, Sierra Leone, Somalia, South Africa,
South Sudan, Sudan, Suriname, Sweden, Tanzania, Thailand, Uganda, USA,
Venezuela, Vietnam, and Zimbabwe. These countries are from North America
(26.3%), Asia (26.3%), South America (23.6%), Africa (21.9%), Oceania (1.3%), and
Europe (0.7%).
The data collection is conducted by a group of fellow students, each responsible
for collecting 117 cases. The data is collected from the databases and reports on
indigenous peoples and conflicts, including Environmental Justice Organizations,
Liabilities, and Trade (EJOLT); Intercontinental Cry; Minewatch; Business & Human
Rights resource Center; International Work Group for Indigenous Affairs; Indigenous
35
coded based on a coding manual, completing the information on each variable.
4.2 Dependent variables
In order to assess the impacts of above factors on the resolution of conflicts
between MNEs and Indigenous communities, we operationalize conflict resolution
through two different variables, the length and the violence of the conflict.
The first dependent variable, the length of conflict, refers to the duration of the
conflict between MNEs and an indigenous community. “Timing” of a conflict is the
point when the conflict between an MNE and a community occurs (Calvano, 2008).
To measure the length of conflict, we will assign each casea binary value of 0 or 1
based on whether the conflict is short or long. The criterion is determined by the
median of the conflict duration in the sample. According to the frequency analysis,
the median of this sample is 89. Therefore, when the duration of a conflict is less than
89 months, the conflict is short and coded as “0”; when the duration is more than 89
months, the conflict is long and coded as “1”. For the ongoing conflict whose
duration is shorter than 89 months, we remove it from our sample, because we have
no idea about its end date and hence whether it will be short or long. For the ongoing
conflict whose duration is longer than 89 months, we keep it in the sample and code it
as long, or say “1”.
The second dependent variable, the violence of conflict, refers to the degree of
disputes and damages that conflict causes. According to Davis and Franks (2014), the
36
are peaceful and involve no violence, while others can result in damage in property,
injuries and even deaths (Calvano, 2008). To measure the violence of conflict, we
will rate the degree of violence by one of five points. The codes and standards are as
follows: 1 refers to no violence; 2 refers to low level of violence (i.e. peaceful protest
and court negotiations); 3 refers to medium level of violence (i.e. threat and
barricade); 4 means high level of violence (i.e. physical damage and injury); 5 means
extreme violence (i.e. all above and death). For example, the Ok Tedi Mine in Papua
New Guinea is the valuable natural resource of this oceanian country. However, in the
1990s, the mine was extracted from international mining companies indiscriminately,
which has done a great damage on the living environment of local people. Finally, the
indigenous peoples living in the Ok Tedi area accused the major extractive mining
company BHP, and received an out-of-court compensation settlement at 28.6 million
dollars (NRI and World Bank, 2002; in Ghazi, 2003).
4.3 Independent variables
To assess the potential impacts of development level of host country and the
degree of indigenous community isolation on MNE-Indigenous conflict, we will use
these two factors as our independent variables. The first independent variable, the
level of development in host country, will be measured by the Human Development
Index (HDI), which is introduced by the United Nations Development Program
37
achievement in key dimensions of human development, including life expectancy,
knowledge and education, and standard of living. The three dimensions are calculated
respectively by life expectancy index, education index, and GNI index. A country
with a higher HDI means that citizens have higher lifespan, higher education level
and higher income level.
The second independent variable is the degree of isolation of indigenous
communities. Isolation hampers communities’ access to the important resources
which grant them authority and identity (Bebbington, 1999). As mentioned before,
the isolation can be divided into geographic isolation and cultural isolation, both of
which have a significant impact on the interactions among indigenous people’s
organizations, the judiciary department, and national institutions (Van Cott, 2000).
In this study, the two types of isolation will be analyzed separately. Geographic
isolation is measured by the assignment of one of two digits (i.e. 1 or 2). The criterion
is: 1 is low-isolated, which means that indigenous people in this type are willing to
contact the broader societies and live with bigger population and more mixed
environment; 2 is high-isolated that means this type of indigenous people prefer to
live in communities which have almost no contact with the outside world. Similarly,
cultural isolation is measured by the assignment of one of three digits (i.e. 1 or 2 or 3).
The criterion is: 1 refers to weak isolation, meaning that the community is not only
connected to but even embedded within a broader mainstream ethnic group; 2 refers
to moderate isolation, meaning that the community is connected to a broader ethnic
38
community has an isolated culture with a different language and is not connected to
or influenced by the outside society.
4.4 Moderator
To assess the moderating effect of MNEs’ prior experience dealing with conflict
or indigenous communities, we make MNE experience as our moderator. Experience
enables companies to learn and perform better in similar operations in the future
(Cyert & March, 1963; Levitt & March, 1988; in Zeng et al., 2013). Moreover,
experience allows companies to better adapt to local environment, to overcome
liabilities of foreignness and newness, and thus to reduce the cultural and
administrative distances (Henisz & Williamson, 1999, in Reimann, Rauer &
Kaufmann, 2015). Experienced MNEs are able to build good relationship with
indigenous communities so as to avoid potential conflict in the future. It is supposed
to positively moderate both the relationship between host country development level
and MNE-Indigenous conflict, and the relationship between indigenous community
isolation and MNE-indigenous conflict.
This study will measure MNE experience in two indicators. The first indicator is
the number of previous projects which an MNE has had with indigenous
communities since the beginning of its operations. The second indicator is the
number of years that have passed since the first encounter of an MNE with an
39 4.5 Control variables
Control analysis should include all the major players within the relationship
(Berkes, 2009). In this study, we discuss the effects of characteristics of host
countries, indigenous communities, and MNEs on the conflict resolution. Therefore,
the control variables in this study will be respectively at country-, community- and
firm-level. The country-level control variable is the political stability of the host
country, which is measured by political stability and lack of violence, one of the six
dimensions of the Worldwide Governance Indicators (WGI). The WGI is based on a
long-standing research program of the World Bank and covers six dimensions of
governance starting in 1996 (Kaufmann, Kraay & Mastruzzi, 2011). The dimension
of political stability and absence of violence/terrorism (PV) indicates the possibility
that a government will be destabilized or overturned by violence including
politically-motivated violence and terrorism (Kaufmann, Kraay & Mastruzzi, 2011).
The community-level control variable is the size of the community, which is
measured by the population of community. According to Black (1974), community
size can affect the characters of the conflicts a lot.
The firm-level control variable is MNEs’ ownership structure, which is measured
in three digits: 1, 2, and 3 refer to state-owned, publicly-owned, and privately-owned
respectively. Firms’ ownership structure is proved to play a significant role in their
40 4.6 Method of analysis
This study will carry out a quantitative research, and use cross-sectional data
which has been collected for a period of time. In addition, to test the hypotheses, the
software SPSS will be used to do the statistic calculation and regression analysis.
This study has two dependent variables, the length and violence of conflict, so
the analysis will be done in twofold. First, we will apply a logistic regression analysis
on the dependent variable length of conflict, because this kind of regression is used
for dummy variables. Second, we will use a linear regression analysis on the
dependent variable degree of conflict, which is often used when at least one
independent variable is proposed to affect one dependent variable. The regression
equation is described as follows:
Y = β0 + β1X1 + β2X2 + β3X3 + β4X1X3 + β5X2X3 + 𝜀,
where Y refers to the dependent variable conflict resolution, measured by the length or the violence degree of conflict. X1 refers to the first independent variable, host
country development level, measured by HDI. X2 refers to the second independent
variable, community isolation, measured in digits according to degrees. The moderator, MNE experience is denoted as X3. β0 is the intercept. β1, β2 and β3 are
the coefficients corresponding to each independent variable and moderator. β4 and β
5 are, respectively, the coefficients of the interaction terms of HDI scores and MNE
experience, and community isolation and MNE experience. The two values indicate
the moderating effect of our moderator variable. 𝜀 is the error term, used to explain
41
To test our hypotheses, we created 12 models to analyze. These models include
different combinations of variables. Table 1 shows the models we will use in the
regression analysis.
Table 1: Models for Regression Analysis
Control Variables Dependent Variables
Independent Variables Moderator Political Stability Community Population Ownership Structure Length of Conflict Degree of Violence HDI scores Community Isolation MNE experience Model 1 X X X Model 2 X X X X Model 3 X X X X X Model 4 X X X X Model 5 X X X X X Model 6 X X X X X X Model 7 X X X Model 8 X X X X Model 9 X X X X X Model 10 X X X X Model 11 X X X X X Model 12 X X X X X X
42
We will run regression analysis as indicated by model 1-6 which aim to test hypotheses H1a, H2a, H3a and H4a, and by model 7-12 which aim at hypotheses H1b, H2b, H3b and H4b.
43
5. Results & Analysis
5.1 Descriptive Statistics
Table 2 shows an overall descriptive statistics of the variables of this study,
including the minimum, maximum, mean, and standard deviation. According to the
analysis of the descriptive statistics, we can find some information. The length of
conflict (M-0.64, SD=.480) has a mean higher than 0.5, which implies that there are
more long-conflict cases than short-conflict cases in our sample. Specifically, there
are 106 (35.7%) cases that have short conflict while 191 (64.3%) cases have long
conflicts. The second dependent variable, violence of conflicts (M=3.60, SD=1.332),
has a high standard deviation, which means that the variance between our cases is
large. Based on the frequency analysis, this conclusion is confirmed: there are 19
no-violence cases (6.4%), 56 low-level-violence cases (18.9%), 65
medium-level-violence cases (21.9%), 43 high-level-violence cases (14.5%), and 114
extreme-violence cases (38.4%). The distribution of this variable is irregular.
Our first independent variable, HDI scores (M=0.7034, SD=0.14354), has a quite
high mean. The HDI is ranged from 0 to 1, so this mean implies that most countries
in our sample are quite developed. The descriptive statistics of the second
independent variable, geographical isolation (M=1.39, SD=0.488), shows that there
are more cases where communities live low-isolated from the outside society than
those consisting of high-isolated communities. Specifically, there are 115 cases
44
consist of low-geographical-isolated communities. Analyzing the third independent
variable cultural isolation (M=2.07, SD=0.677), we find that there is a normal
distribution, meaning that the medium category has the most observations. To be
specific, there are 79 (26.6%) cases that consist of strong-cultural-isolated
communities, 160 cases (53.9%) that consist of moderate-cultural-isolated
communities, and 58 cases (19.5%) consisting of weak-cultural-isolated
communities.
Finally, the moderator MNE experience is measured by two variables: number of
Table 2: Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Length of Conflict 297 0 1 .64 .480 Violence of Conflict 297 1 5 3.60 1.332 HDI Scores 297 .42 .94 .7034 .14354 Geographical Isolation 297 1 2 1.39 .488 Cultural Isolation 297 1 3 2.07 .677 Number of Previous Projects 297 1 150 30.96 38.994 Experience Years 297 1 170 27.07 26.095 Political Stability 297 1 6 2.92 1.275 Community Population 297 19 18000000 288563.80 1285909.447 Ownership Structure 297 1 3 2.07 .559
45
previous projects (M=30.96, SD=38.994), and number of years (M=27.07,
SD=26.095). The standard deviations of these two variables are rather high, implying
that the experience dealing with conflicts spread widely among our cases. Based on
the frequency analysis, the majority of the companies are experienced in conflict
resolution, which can influence the results. Table 2 also shows the descriptive
statistics of our three control variables. The considerably large standard deviation of
community population indicates there is a great difference in this item between cases.
5.2 Correlation Analysis
After the above descriptive statistics, we run a correlation analysis to test the
preliminary relations between all our variables, i.e. the length of conflict, the degree
of violence, HDI scores, geographical isolation, cultural isolation, number of previous
projects, experience years, political stability, community population, and ownership
structure. Table 3 shows whether these variables are significantly correlated, and
whether these correlations are positive or negative.
First of all, there is a low correlation between the two dependent variables, the
length of conflict and the degree of violence (r=0.07, p=0.232), implying that these
two variables are different and not correlated. For the first independent variable, HDI
scores, we only find it significantly correlated with one of the dependent variables,
the length of conflict (r=-0.36, p<0.01). The negative correlation coefficient implies
that if a host country has a lower HDI score, MNEs will be likely to encounter longer