The outbreak of the Syrian Civil War; Ethnic division and spill-over effects from neighbouring conflicts

(1)

I

Leiden University

Department of Political Science

Master Thesis

The outbreak of the Syrian Civil War

–

Ethnic division and spill-over effects from

neighbouring conflicts

By Caroline Ohlendorf, s2421615

Submitted to Prof. Dr. Nicolas Blarel and Prof. Dr. Roos van der Haer

Word count: 10 280

(2)

II

Abstract

This master thesis examines the risk of civil war outbreak, by evaluating the interplay between ethnic division and spill-over effects of neighbouring conflicts. This is done through a case study of the Syrian Civil War and how its outbreak was influenced by the conflicts in the neighbouring countries Egypt and Tunisia. Power divisions along ethnic lines within Syria that had existed for decades have not led to a civil war outbreak before 2011. Consequently, the following research question arises: Was the outbreak of the Syrian Civil War influenced by spill-over effects of neighbouring countries? In this paper, I will explain how the ethnic divisions within Syria as potential factors contributing to a civil war outbreak were influenced by Egyptian and Tunisian political leaders. These spill-over effects are examined by analysing Twitter tweets of political actors in all three countries, focusing on the repeatedly used terms and hashtags chronologically, relating to events of the Arab Spring, first by Egyptian and Tunisian political actors and then by Syrian political actors. This study finds some evidence for the claim that spill-over effects from political actors in Egypt and Tunisia to Syrian political actors occurred in the context of the Arab Spring and the outbreak of the Syrian Civil War.

(3)

III

Table of Figures

Figure 1: Causal mechanisms of the outbreak of the civil war in Syria, Hypothesis 1, Hypothesis 2, Hypothesis 3. ... 11 Figure 3: Amount of usage of hashtag Egypt by political actors in Egypt, Syria and Tunisia. ... 22 Figure 4: Amount of usage of term Egypt by political actors in Egypt, Syria and Tunisia. ... 23 Figure 5: Amount of usage of hashtag Tunisia by political actors in Egypt, Syria and Tunisia. .. 24 Figure 6: Amount of usage of hashtag USA by political actors in Egypt, Syria and Tunisia. ... 25 Figure 7: Amount of usage of term violence by political actors in Egypt, Syria and Tunisia. ... 26 Figure 8: Amount of usage of term Human Rights by political actors in Egypt, Syria and Tunisia. ... 27 Figure 9: Amount of usage of term Mubarak by political actors in Egypt, Syria and Tunisia. ... 28 Figure 10: Amount of usage of term Revolt/Protest/Revolution by political actors in Egypt, Syria and Tunisia. ... 29 Figure 11: Amount of usage of term Freedom by political actors in Egypt, Syria and Tunisia. ... 30 Figure 12: Amount of usage of term Change by political actors in Egypt, Syria and Tunisia. ... 31 Figure 13: Amount of usage of term Cooperation by political actors in Egypt, Syria and Tunisia. ... 32 Figure 14: Amount of use of hashtags/terms by Abd-al-Mun’im Abu-al-Futuh (Egypt) and Moaz al-Khatib (Syria). ... 35 Figure 15: Amount of use of the terms Palestine, Martyr and Revolt/Protest/Revolution by Abd-al-Mun’im Abu-al-Futuh. ... 36

(5)

V Figure 16: Amount of use of terms Palestine, Martyr and Revolt/Protest/Revolution by Moaz al-Khatib. ... 37 Figure 17: Overview of all hashtags and terms and how often they were used in Egypt, Tunisia and Syria between January 2011 and December 2012. ... VII

Table of Tables

Table 1: Ethnic Power Relations in Syria from 1970 to 2016 by the ERP Dataset Family 2019. . 19 Table 2: List of all identified hashtags. ... IX Table 3: List of all Arabic terms with the English translation. ... XI Table 4: Overview of political actors in Egypt, Syria and Tunisia and their Twitter Handle. .... XII

(6)

1

1. Introduction

When the Syrian Civil War broke out in March 2011, to outside observers it appeared as if the opposing groups were motivated primarily by ethnic grievances and power divisions along ethnic lines. However, ethnic divisions and ethnic grievances within Syria had been present for decades due to Syrian society being composed of several ethnic groups – such as Sunni Arabs, Christians, Kurds, Druze, and Alawites – and the constant struggle for political power that has often led to political and economic discrimination of a whole ethnicity, especially since the Alawite minority began ruling the country in 1970 (EPR Dataset Family 2019). However, neither ethnic divisions nor ethnic grievances had led to an intrastate conflict before 2011.

On 17 December 2010, in Sidi Bouzid, street sales assistant Tarek al-Tayeb Mohammed Bouazizi set himself on fire to raise attention to the economic grievances in his home country Tunisia(Fahim 2011). This incident was credited to have started a transnational political domino effect in the region of the Middle East and North Africa (MENA), called the ‘Arab Spring’. This movement resulted in mass protests by citizens against their regimes, aspiring to greater democratic freedom in several countries of the MENA region (Fahim 2011). The political domino effect in the Arab Spring was (at least partly) caused by spill-over effects, that transported the wave of protest from one country to another. Spill-over effects are defined in this study as influences that took place between political actors of different countries which resulted in the influenced actors being motivated to protest, demanding more power and rights for their ethnic group. The spill-over effects will be investigated by analysing Twitter tweets by political actors in Syria and political actors of neighbouring conflicts as proxies for a spread of information. Accordingly, the research question

(7)

2

of this paper is: In which ways did neighbouring conflicts have an influence on the outbreak of the

Syrian Civil War in 2011?

There have been a number of significant contributions in the civil war literature regarding the role of the factors of ethnic division (Cederman et al. 2018; Horowitz 1985; Fearon & Laitin, 2003; Walter & Denny 2014) within a state and on the role of spill-over effects for the onset of civil wars (Weidmann 2015; Weidmann 2019; Tarrow 1996).However, only few scholars, such as Salehyan & Gleditsch (2006) and Weidmann (2019), have conducted research on how the dynamics of ethnic division within a country and spill-over effects from neighbouring countries interact with each other leading to a civil war outbreak. Therefore, this thesis examines this causal mechanism in the case of the Syrian Civil War outbreak.

Before delving into the analysis, it is important to keep in mind that at the time of writing this thesis – January 2020 – the Syrian Civil War is still ongoing.Therefore, it is difficult to fully identify all factors that influenced the outbreak of the Syrian Civil War, especially because new actors and political players with own interests and motivations are entering and leaving the conflict continually. An analysis of the roots of the conflicts and a reconstruction of the course of events leading to the current situation in Syria might help policy makers address the ongoing impacts of the crisis. Moreover, analysing the interplay of spill-over effects and ethnic grievances might help in identifying strategies and mechanisms to discourage spill-over effects from one country suffering from ethnic conflict to another in the future. In terms of academic relevance, this paper aims to add findings to the literature on civil war outbreaks by combining the influence and interactive effects of the two factors of ethnic division and spill-over effects on a civil war outbreak. Because this research paper is tackling the influential factors for only the outbreak of the civil war, the timeframe of the data analysis was between January 2011 and December 2012.

(8)

3

The results show, firstly, that the existence of ethnic grievances along ethnic lines in the Syrian society can be found. The small Alawite minority holds most of the political power, while larger ethnic groups like the Arab Sunni seem to have very little political power. Secondly, the results of the data analysis of the tweets seem to confirm the existence of spill-over effects from political actors in Egypt and Tunisia to Syrian political actors to some extent. However, the data unfortunately did not include clear information and hints about the ethnic identity of the political actors in the three countries. Therefore, this paper can only provide an inductive test on how a Sunni Arab political actor in Syria has been influenced by a Sunni Arab political actor in Egypt. The results do not provide sufficient indications to actually prove spill-over effect.

This paper will initially proceed with a review of civil war theory literature (Section 2), focusing on the importance of ethnic division and spill-over effects from neighbouring countries for the outbreak of a civil war. This will be followed by the presentation of my theoretical framework (Section 3), which presents the relevant theories integral to this paper’s argument, as well as the hypotheses I intend to test. My research design is then presented in the following section (Section 4). This section will discuss the research method used (Section 4.1), the case selection strategy of Syria (Section 4.2), data collection (Section 4.3), as well as explain the operationalisation of the variables for the testing of my hypotheses (Section 4.4). The next section then presents the data analysis for the hypotheses of this paper and discusses whether these could be verified or needed to be falsified (Section 5). Finally, the conclusion will show the main findings of this paper, evaluate the methods used and give an outlook on future research (Section 6).

(9)

4

2. Literature Review

This section gives an overview of the existing literature dealing with the drivers and factors that influence the outbreak of a civil war. Kalyvas classified civil wars as an “armed combat taking place within the boundaries of a recognized sovereign entity between parties that are subject to a common authority at the outset of the hostilities” (2005, p. 21). According to Gleditsch (2017), civil war is the most common form of armed conflict since the end of the Cold War. This increasing prevalence of civil wars has also led to the emergence of civil war related research (Cederman et al., 2018). Several scholars have examined different factors and causes for the outbreak of a civil war. Much of the early literature on civil wars has focused on domestic factors that can result in the outbreak of a civil war, such as the research conducted by Horowitz (1985) on ethnic divisions.

2.1. Ethnic division and ethnic grievances as factors accounting for civil

war outbreak

Several scholars have examined ethnic division as a causal factor for a civil war outbreak. Varshney defined ethnicity as “a sense of collective belonging, which could base on common descent, language, history, culture, race or religion” (2007, p. 277). Wolfsfeld et al. (2013) discussed the importance of political grievances in particular in political autocracies in the MENA region. They focused on inflexibility amongst rulers who are not capable of adjusting policies according to the demands of their citizens in an appropriate timeframe. Political grievance can arise for example if the division of power within a country happens along ethnic lines. Therefore, political grievances and ethnic grievances can go hand in hand, which might lead to instability within a country and increase the risk of violent conflict and civil war. Collier & Hoeffler (1998) claimed that ethnic grievances are a product of an unequal distribution of wealth. It appears that members of some

(10)

5

ethnic groups might experience economic disadvantages because they are not members of the ruling ethnic group and therefore are discriminated against and neglected. This shows how ethnic and economic grievances are intertwined and hence, how both can be considered a source of protest. Fearon & Laitin (2003) claimed that ethnic division only increases the risk of a civil war outbreak if an ethnic minority is ruling the country because otherwise, civil wars could potentially occur in many more countries around the world. Ethnic division is one of two influential factors for the outbreak of a civil war that are investigated in this paper because the Syrian society is composed of several different ethnicities and has been ruled by the Alawite minority since the 1970s (EPR Dataset Family 2019).

2.2. Political Opportunity Structure as a driver for civil war outbreak

A further aspect Collier & Hoeffler (1998) have identified as influential for the outbreak of a civil war is greed. Greed in this context is understood as the practical structural opportunity enabling citizens to protest against their regime (Collier & Hoeffler, 1998). The concept of practical structural opportunities was implemented by Tarrow in 1996. He argued that a ‘window of opportunity’ is essential for a civil war outbreak. He claimed that protesters need adequate structural and political conditions, which open up the window of opportunity for group leaders to mobilise individuals for a protest. These newly opening opportunity structures do not necessarily have to originate from formal elements such as governmental institutions but can also be initiated by non-governmental institutions, such as religious institutions, neighbourhood associations or even external support by groups of the same ethnicity from neighbouring countries. External actors can provide these rising protest groups with significant resources, such as financial support or arms, that are crucial for the formation and success of a protest group (Tarrow, 1996). Tarrow identified four striking signals suggesting new Political Opportunity Structures: “the opening up of access to

(11)

6

power, shifting alignments, the availability of influential allies, and cleavages within and among elites” (Tarrow 1996, p. 43). In 2011, Syria was surrounded by states that were experiencing anti-regime protests and political changes, which provided opportunities for finding new allies and possible spill-over effects. Building on Tarrow’s theory, Political Opportunity Structures might have been used in the case of Syria through transnational ethnical linkages that provided new allies outside and across national borders of the same ethnicity.

2.3. Transnational influence on civil war outbreak

In the early 2000s, researchers increasingly focussed on transnational factors that might influence a civil war outbreak (Cederman et al., 2018). Similar to Tarrow, Weidman (2015) focused on structural factors that influence the outbreak of a civil war. Weidmann noted that transnational mechanisms raise the possibility of political violence. He identified globalisation and the greater importance of information and communication technology as highly influential for this process to happen. Due to both factors, ethnic conflict can more easily spread across state borders. This means that an ethnic conflict group in country A can influence a group of the same ethnicity in country B to also rebel against its government. Weidmann (2015) distinguished between resource spread and informational spread between countries. In researching spill-over effects driven by the spread of resources, he followed the argumentation of Kaufman (1996), who spoke of a “mass-led” escalation pattern, which is the actual spread of fighters and weapons across borders. Weidmann’s recent research examined the role of the internet for the outbreak of a civil war, which showed that the spread of information through the internet is a useful catalyst for the expansion of an already ongoing protest throughout a whole country (Weidmann, 2019). Salehyan & Gleditsch (2006) also focused on transnational factors accounting for a civil war outbreak. They identified three possible effects: Firstly, there is the possible existence of ethnic linkages across national borders between

(12)

7

neighbouring countries. Secondly, there is the possibility of a demonstration effect, meaning that a neighbouring conflict can function as an exemplar for the outbreak of a civil war in another state. Thirdly, through a civil war in a neighbouring country, a state can experience negative externalities, such as trade disruption, which can increase grievances within the affected state. To assess which spill-over effects might have influenced the outbreak of the Syrian Civil War, this paper takes into account the statements made online by Syrian political actors as an example of how information spread that led to protest took place.

So far, only few studies (Salehyan & Gleditsch (2006) and Weidmann (2018)) have linked ethnic division and the influence of spill-over effects from neighbouring countries as influential factors for a civil war outbreak. This paper aims to contribute to the nascent scholarship explicitly linking these two factors by analysing the novel case of Syria and clarifying how the two factors interact with and influence each other.

3. Theoretical Framework

This section presents the theoretical framework this paper builds on as well as the hypotheses that will be tested. Ethnic division within a country and spill-over effects from neighbouring conflicts can be identified as influential factors that increase the probability for a civil war outbreak. Ethnic division can lead to more political exclusion, ethnic grievances and power division along ethnic lines. A conflict in a neighbouring country might spread through communication networks and open up Political Opportunity Structures for protest.

(13)

8

3.1. Hypothesis 1: Ethnic grievances and unequal distribution of power

along ethnic lines

As previously discussed, Walter & Denny (2014) argued that the division of power along ethnic lines, which gives the ruler the possibility to privilege the ethnicity he belongs to, increases ethnic grievances. Also, ethnic groups tend to live in geographical agglomerations with limited distance to each other, eventually resulting in them sharing the same language, traditions and socialisation (Walter & Denny, 2014). In addition, the bond of ethnic identity is less elastic than the sharing of other identities, such as cultural preferences (Walter & Denny, 2014). These factors are possible conditions that make a civil war outbreak motivated by ethnic grievances more likely. According to Walter & Denny (2014), these factors lead to three advantages, increasing the risk of a civil war outbreak because of ethnic division. Firstly, ethnic groups, especially minorities, are more likely to have stronger grievances against the state than majorities (Walter & Denny, 2014). Secondly, ethnic groups have more opportunities to mobilise and organise a protest movement than other group identities; and thirdly, they are more likely to be confronted with bargaining problems that cannot be solved (Walter & Denny, 2014). Considering the case of Syria, one can observe that the Syrian society is composed of several ethnicities. In addition, the ethnic minority of the Alawites is disproportionately represented in the government elite of the country. Based on this overrepresentation in decision circles, one can assume that they hold on to the most amount of political power. Considering the above presented theories on how ethnic division and power division along ethnic lines can influence the stability of a society, my first hypothesis is as follows:

H1: Ethnic grievances within Syria increase with a more unequal distribution of power along ethnic lines.

(14)

9

3.2. Hypothesis 2: The role of spill-over effects from neighbouring

countries for the providence of new Political Opportunity Structures

The outbreak of the Arab Spring resulted in a domino effect of intrastate conflicts and civil war outbreaks. Initially beginning in Tunisia, the wave of protests was carried to Egypt, Libya, Bahrain, Syria and Yemen (Ahmad, 2015). Recognizing Tarrow’s theory of new Political Opportunity Structures as well as the transnational factors Salehyan & Gleditsch (2006) described that might influence the outbreak of a civil war, it is likely that the neighbouring conflicts served as a comparable situation for the Syrian society, demonstrating that heavy protest against the regime can result in realignments of power distribution. Building on the spill-over theory advanced by Weidmann (2015), the internet and social media in particular have become more accessible through globalisation for a greater amount of people, which increased the accessibility of information on events happening in other countries. O’Callaghan et al. (2014) have analysed the behaviour of Syrian citizens during the beginning of the Arab Spring and the increasing protests within Syria on the internet on platforms such as Facebook, Twitter, YouTube and Instagram. They found that online activity had an influence on the outbreak of the protests in Syria. However, research specifically examining how spill-over effects have affected ethnic group leaders has not been conducted in the Syrian case. Therefore, the analysis of tweets is used as a proxy for information spread between political actors of different countries to examine how the wave of protest has spilled over from one country to another and hence, how the revolutions in each country influenced each other. Therefore, the second hypothesis for this paper is as follows:

H2: Spill-over effects from neighbouring conflicts provided new Political Opportunity Structures for group leaders within Syria to mobilise protest.

(15)

10

3.3. Hypothesis 3: The influence of ethnicity across national borders

As stated before, the Syrian society is composed of several ethnicities and has been ruled by the small Alawite minority for decades. If hypothesis 1, proposing that ethnic grievances increased through the division of power along ethnic lines, can be verified, it seems plausible that ethnic division might have played a supporting role in the Syrian Civil War outbreak. If hypothesis 2, assuming that spill-over effects provided a Political Opportunity Structure for Syrian group leaders to mobilise for protest, can be confirmed, one can argue that these Syrian group leaders were influenced and motivated by group leaders in neighbouring conflicts. Examining the theory of Salehyan & Gleditsch (2006), stating that cross border ties of ethnicities in neighbouring countries can exist, I argue that ethnic leaders in Syria were motivated by ethnic leaders in neighbouring countries to mobilise for protest. Therefore, the third hypothesis is as follows:

H3: Syrian ethnic group leaders were influenced by group leaders of the same ethnicity in neighbouring conflicts to mobilise for protests.

3.4. Main Argument

Previous research already established that there have been some spill-over effects of the Arab Spring to Syria (Lynch et al., 2014). The fact that the Syrian society consists of several ethnic groups and that the Alawite minority has been the ruling party for a long time allows for some speculation that ethnicity might have played a role in the outbreak of the civil war in Syria. Furthermore, the composition ofthe conflict parties in the current Syrian Civil War shows that they are mainly divided along ethnic lines (Rousseau, 2014).

Having presented the three hypotheses that will be tested, I argue that neither of the two factors, ethnic division within Syria and spill-over effect from neighbouring countries, could explain the

(16)

11

onset of the civil war in Syria by themselves. Instead, only the interplay of the two factors, previous existing ethnic grievances in Syria and spill-over effect from neighbouring conflicts, led to mobilization and initiation of protest by ethnic group leaders within Syria, influenced by the examples of ethnic group leaders in neighbouring conflicts which then resulted in the uprising of protest and finally in the outbreak of the Syrian Civil War. Therefore, hypotheses 1 and 2 need to be verified first before hypothesis 3 can be examined. To test hypothesis 1, I will rely on existing data, but the novel contributions of this paper to existing analyses by Weidmann and Salehyan & Gleditsch are related to hypotheses 2 and 3, which I will outline in the next section. The causality line of my argumentation is visualised in the following diagram:

(17)

12

4. Research Design

4.1. Method

The method of congruence analysis, as defined by Beach and Pedersen (2016) is used for this single case study of Syria. ‘Congruence analysis’ in this context means that the outcome, civil war, is examined by taking into consideration the variables ethnic division and spill-over effects from neighbouring countries. In doing this, it needs to be acknowledged that these variables are only two of a multi-variable system which includes several variables that might potentially explain civil war outbreak. This congruence analysis is only focusing on these two propositions and does not rule out any alternative approaches and theories in order to explain civil war outbreaks.

First of all, the theories of interest are chosen: the theory of ethnic division and the theory of spill-over effects from neighbouring countries. Both pose influential factors for a civil war outbreak. Secondly, a relevant case is chosen to which both of the theories can be applied: the Syrian Civil War outbreak. Thirdly, I check if the different variables that are necessary in order to examine a causal relation, the existence of ethnic division within the Syrian society and the existence of spill-over effects from neighbouring conflicts, are present or not.

In order to test if the different variables are present, for analysing hypothesis 1, I use the dataset Ethnic Power Relations (EPR) Dataset Family 2019 by the ETH Zürich to investigate whether ethnic division within Syria exists. For hypotheses 2 and 3, I use tweets from Egyptian, Tunisian and Syrian political actors as proxies for cross-national information spread to investigate if spill-over effects from neighbouring conflicts to Syrian political actors can be confirmed. Only if the existence of the two variables – ethnic division and spill-over effects – can be proven, the causal mechanism of the two factors can be examined. In a further step, the relationship of these two

(18)

13

propositions is analysed: How do the two factors of ethnic division and spill-over effects from neighbouring countries interact with each other regarding the outbreak of a civil war? This is mainly done when empirically evaluating hypothesis 3 in Section 5.3, because this hypothesis is the one investigating the relationship between both factors. The data that is collected for this paper – the tweets of the political actors in Syria, Egypt and Tunisia – is limited. Therefore, the level of theoretical proposition and the actual observable indications that can be retrieved from the collected data, can differ.

In a final step, it is examined to what extent the findings of this congruence analysis can be related to similar cases. It is discussed whether this analysis, which tackles the outbreak of the Syrian Civil War by taking into consideration the interplay of ethnic division and spill-over effects from neighbouring countries, provides helpful insights in order to explain other, similar and future civil war outbreaks.

4.2. Case Selection

The case selection strategy for this research paper is based on the logic of a plausibility probe. The case of Syria is chosen because the characteristics of the country are in congruence with the two factors that were identified as crucial for this research. Firstly, the Syrian society consists of several ethnic groups. Therefore, the chance that ethnic division within the country has been an influential factor for the country’s stability as presented by Cederman et al. (2018), Collier & Hoeffler (1998), Fearon & Laitin (2003) and Hegre et al. (2001) is high. Secondly, Syria is located in the MENA region. During the time of the civil war outbreak in Syria, immediate neighbouring countries and many countries in the whole MENA region experienced protest, revolutions and political changes in general. Therefore, the initial conditions for a spill-over effect from a neighbouring country to take place as argued by Weidmann (2015, 2018) and Salehyan & Gleditsch (2006) seem to be

(19)

14

present. The case of Syria therefore fits the conditions for a typical case of a civil war outbreak that was motivated by new Opportunity Structures and external shocks, even though the intrastate conditions, including ethnic division, had existed long before. Depending on the results of this study, the Syrian case can be connected to a larger sample of civil war outbreaks in countries that experience ethnic divisions and are located in regions that experience political upheaval. The single case study of Syria is an attempt to identify factors that can contribute to the existing research of the interplay of ethnic division and spill-over effects of neighbouring countries as influential factors for the outbreak of a civil war. So far, no scholar has conducted research on the interplay of ethnic division and spill-over effects from neighbouring countries as causal factors for the Syrian Civil War outbreak.

4.3. Data Collection

To test the first hypothesis, the EPR Dataset Family 2019 is used. To test the second and third hypothesis, data from Egypt, Tunisia and Syria is used to analyse whether a spill-over effect to Syria from Egypt and Tunisia took place, because these two countries were the first ones affected by the Arab Spring. The list of political actors for all three countries is taken from the Integrated Crisis Early Warning System (ICEWS). This program is a “comprehensive, integrated, automated, generalizable, and validated system to monitor, assess, and forecast national, sub-national, and internal crises” (Lockheed Martin Corporation, 2019). The system provides more than 100 data sources and 250 international and regional newsfeeds in real time to process all actors involved in conflict events and named in conflict-related news reports (Lockheed Martin Corporation, 2019). From this, I compile a list of all actors that were named and involved in conflict-related news reports per country. Based on the data from ICEWS, the list includes 374 political actors in Syria, 828 political actors in Egypt and 422 political actors in Tunisia. After retrieving this list of political

(20)

15

actors for Syria, Tunisia and Egypt, I identify actors owning a personal Twitter account: 53 in Syria, 106 in Egypt and 32 in Tunisia. With the help of an API Crawler, the last 3500 tweets per actor are collected from Twitter. Due to data privacy regulations imposed by Twitter, only a limited number of tweets (3500 per person) can be retrieved. As the focus of this research paper is the immediate outbreak of the Syrian Civil War, only data from 2011 and 2012 is included. Hence, tweets from 13 political actors in Syria, 31 political actors in Egypt and 10 political actors in Tunisia are gathered for the years 2011 and 2012. These numbers of actors per country result in 9378 tweets from Syrian actors, 21233 from Egyptian actors and 2216 from Tunisian Actors. Consequently, the 9378 tweets from the 13 Syrian political actors are scanned for hashtags and repeatedly used terms. To translate the Arabic tweets, I consulted with a private individual who is a Syrian Arab. Within this process, 63 hashtags and 40 repeatedly used terms are identified. The list of all hashtags as well as all Arabic terms including their translation to English, which was done by myself for the analysis of the terms and hashtags, can be found in the Appendix (Tables 2 and 3). In order to examine whether a spill-over effect from Egypt and Tunisia to Syria can be found, Python and STATA are used to aggregate the data, extracting how often, when and by whom these hashtags and terms are used in the three countries.

4.4. Variables

For my first hypothesis, the independent variable is ‘power division along ethnic lines’ while the dependent variable is ‘existence of ethnic grievances’. ‘Power division along ethnic lines’ is retrieved from the EPR Dataset Family. As stated before, based on the theories by Salehyan & Gleditsch (2006), Cederman et al. (2018) and Fearon & Laitin (2003), power division along ethnic lines typically result in ethnic grievances, especially if minorities of a country hold a greater part of political power than majorities. Therefore, the existence of ethnic grievances prior to the Syrian

(21)

16

Civil War outbreak in 2011 can be proven, if power division along ethnic lines led to elimination of political power for certain ethnic groups.

For my second hypothesis, the independent variable is the ‘spill-over effects from neighbouring conflicts’, while my dependent variable is the ‘provision of new Political Opportunity Structures for group leaders in order to mobilise for protest’. As stated before, Tunisia and Egypt were the first two countries where the wave of protest against the authoritarian regimes started. Therefore, data from these two countries as well as from Syria is used to examine whether a spill-over effect has taken place. Building on Weidmann’s (2018) research on the internet’s role for conflict spill-over effects, a social media analysis of tweets from Syria as well as Egypt and Tunisia is conducted in order to investigate if similar hashtags and terms in tweets referring to the rise of protest and the outbreaks of civil wars are used. The measurement strategy whether spill-over effects from political actors in Egypt and Tunisia to political actors in Syria take place needs to be defined. I define the threshold for confirming a spill-over effect as 10 or more mentions of a term or hashtag by the trendsetting countries (Egypt and/or Tunisia) and the receiving country (Syria). In addition, another condition is necessary. The 10 mentions or more should occur within a limited timeframe, i.e. less than two months should pass between a hashtag or term being used in Egypt and/or Tunisia and it being used in Syria. The data provides information about the months in which the terms and hashtags were used in the different countries. However, the data does not provide any information about on which exact date the hashtags and terms were used. Therefore, if a hashtag or term is used in the same month in Egypt/Tunisia and Syria, it cannot be determined in which country the hashtag or term was used first, meaning which country was the trendsetting and which one the receiving one. For these obvious practical constraints, a mentioning of a hashtag or term in the same month by political actors in Egypt/Tunisia and in Syria is counted as confirming a spill-over effect in the data analysing part of this paper.

(22)

17

Only if these two conditions are met – a minimum number of 10 mentions of each term and hashtag by political actors of Egypt and Tunisia and by Syrian political actors, as well as the limited period of time of maximum two months between the first mentioning – a possible spill-over effect of the specific term or hashtag can be confirmed.

For my last hypothesis, the independent variable is ‘Syrian ethnic group leaders are influenced by transnational ethnic influence of neighbouring conflicts’ while my dependent variable is ‘the mobilization for protest’. In order to find evidence that Syrian ethnic leaders are influenced by ethnic group leaders in neighbouring countries, I first try to identify the ethnic identity of all political actors of the three countries. In a second step, I want to examine whether political actors of the same ethnicity within the three different countries use the same hashtags and terms when tweeting about the incidents happening in their countries in the context of the Arab Spring to analyse, if spill-over effects from one ethnic leader to another ethnic leader of the same ethnicity in another country take place.

However, due to the limitation of the data, it is not possible to identify the ethnic identity for all political actors. Therefore, the empirical section for hypothesis 3 includes an inductive example, testing if the data provides evidence for a spill-over effect from a Sunni Arab in Egypt to a Sunni Arab in Syria. Examining this limited sample shows whether the expectation holds that a transnational ethnic influence has taken place to mobilise for protest in Syria. A possible spill-over effect can be detected if both actors repeatedly use the same hashtags and terms and especially, as stated before, if the use of the terms and hashtags is chronological in time. The actors for this inductive example, Abd-al-Mun'im Abu-al-Futuh from Egypt and Moaz al-Khatib from Syria are chosen because they are confirmed members of the Sunni Arab ethnicity. Their ethnicity is identified with the help of newspaper articles that claim that both of them are Sunni Arabs. Moaz al-Khatib is a member of the National Coalition for Syrian Revolutionary and Opposition Forces

(23)

18

and used to be its president (El Gamal & Doherty, 2012). Before the outbreak of the Syrian Civil War, was an imam at the Umayyad Mosque in Damascus (El Gamal & Doherty, 2012). Abd-al-Mun'im Abu-al-Futuh is a former leader of the Muslim Brotherhood who ran for President of Egypt as an independent candidate in 2012 (El-Sisi, 2018). He has been a student activist and Islamist politician throughout his whole life (Hamid, 2012).

5. Empirical Section

5.1. Hypothesis 1: Ethnic grievances and unequal distribution of power

along ethnic lines

In the following section, the data for hypothesis 1 is analysed in order to examine if proof for hypothesis 1 can be found. Before the outbreak of the civil war in 2011, the Syrian society had 21 million citizens and consisted of more than 15 ethnic and religious groups (Said, 2013). The Arab affiliated groups were in the majority before the outbreak of the civil war in 2011. Besides the Arabs, there were also Kurds (10-12%), Turkomans (9%), Assyrians and Arameans (together 4.5%), Armenians (less than 1%) and Circassian (1000) in Syria (Said, 2013). Besides this ethnic diversity, Syria was also divided into different religious confessions (Said, 2013.). The Muslim majority (around 85-90%) was divided between Sunnis (73%), Alawites (around 10%), Druses (3%) and a small number of Ismailis and Shiites (Said, 2013). The Christian minority accounted for around 10-12% of the population and furthermore consisted of several different Christian confessions (Said, 2013). The formation of the conflict parties in the Syrian Civil War mainly along ethnic lines shows that ethnic divisions might have been an influential factor for the Syrian Civil War outbreak. Considering the theories about the relevance of ethnic divisions for a civil war outbreak by Walter & Denny (2014) as well as by Collier & Hoeffler (1998), the case of Syria

(24)

19

shows that these conditions were present. Moreover, Fearon & Laitin (2003) claimed that if a country is ruled by a small ethnic minority, this also increases the risk of a civil war outbreak. The Alawite minority, to which president Bashar al Assad belongs, has been the ruling party in Syria since 1963. By contrast, parts of the society, such as the Sunnis and Kurds, were not well represented in the regime, and therefore only held a small share of the power (Rousseau, 2014). Examining the data of the EPR Dataset Family 2019, the results for the power relations along ethnic lines in Syria are the following:

From To Group Size Status

1970 2011 Sunni Arabs 0.65 Powerless

1970 2011 Alawi 0.13 Dominant

1970 2011 Christians 0.1 Powerless

1970 2011 Kurds 0.08 Discriminated

1970 2011 Druze 0.03 Powerless

2012 2016 Sunni Arabs 0.65 Powerless

2012 2016 Alawi 0.13 Dominant

2012 2016 Christians 0.1 Powerless

2012 2016 Kurds 0.08 Powerless

2012 2016 Druze 0.03 Powerless

Table 1: Ethnic Power Relations in Syria from 1970 to 2016 by the ERP Dataset Family 2019.

The data shows how the political power within Syria is divided along ethnic lines, which ethnicity is rather dominant and which ones are rather powerless. According to the EPR Dataset Family 2019 “dominant” means that the ethnicity holds a lot of the political power within the country, while if an ethnicity is “powerless”, their share of political power is rather small. The data shows that the only group holding on to the power in Syria are the Alawi, even though they only represented 13% of the Syrian Society.

In contrast, the Sunni Arabs seem to be powerless even though they are the largest ethnic group in Syria with 65%. The other small ethnicities in Syria listed by the EPR Dataset Family 2019, the Christians and the Druze, are either also powerless or even discriminated against like the Kurds. As stated in the research design, the independent variable is ‘Power division along ethnic lines’.

(25)

20

As the EPR Dataset Family 2019 shows, the power division in Syria is strongly divided/distributed along the ethnic lines, meaning that the existence of ethnical grievances, the dependent variable, can be proven. As mentioned in the literature review and the theoretical framework, ethnic division resulting in ethnic grievances can be an influential factor for the outbreak of a civil war, as shown by Cederman et al. (2018), Horowitz (1985), Fearon & Laitin (2003) and Walter & Denny (2014). Therefore, my first hypothesis Ethnic grievances within Syria were increased by the unequal

distribution of power along ethnic lines can be confirmed.

5.2. Hypothesis 2: The role of spill-over effects from neighbouring

countries for the providence of new Political Opportunity Structures

5.2.1. General explanation of the Data

In total, 63 hashtags and 40 repeatedly used terms were found when analysing the tweets of the selected Syrian political actors. Due to obvious practical constraints, not all tweets could be examined. Also, not all of the terms and hashtags could be directly linked to political events in the context of to the Arab Spring and the possible spill-over effects of conflicts in Tunisia and Egypt to Syria. Therefore, firstly those hashtags and terms were identified that showed a contextual connection to the political events happening during the time of the Arab Spring in the countries themselves. Secondly, all hashtags and terms that did not provide enough data in order to interpret their content in context of the hypothesis were dismissed. Consequently, hashtags or terms that were only used a couple of times were not further investigated because they were not significant enough. An overview of all hashtags and terms and how often they were used in the tweets of the three countries in total can be found in the Appendix (Figure 16).

The data shows that no tweets could be retrieved for January and February 2011 for Syria. Therefore, the data for these two months is missing in the presentation of the data. This gap in the

(26)

21

data is probably due to an initial shutdown of the internet in most parts of Syria by the Syrian government in early 2011 (Chulov, 2012).

In the process of identifying which hashtags and terms show the strongest connection to political events in the context of protest and the Arab Spring in the three countries, 11 hashtags and terms were chosen for further examination. The following section goes through these 11 hashtags and terms one by one and investigates the data patterns. Then, a determination whether the data demonstrates the presence of spill-over effects from Tunisia and Egypt to Syria takes place. Because a possible spill-over effect can best be shown chronologically, meaning whether a hashtag or term was first used in Egypt and/or Tunisia and then later on in Syria, the data is presented along a timeline.

As stated before, the number of political actors that could be identified as well as the number of the released tweets is very limited for Tunisia. Therefore, the analysis mainly concentrates/focuses on the interplay of hashtags and terms used by political actors in Egypt and Syria.

5.2.2. Analysis of the 11 hashtags and terms one by one

Hashtag Egypt and term Egypt

The data shows that the hashtag Egypt was mainly used in Egypt. However, it is striking that the hashtag Egypt was increasingly used in Syria in September 2011. This might show that political actors in Syria referred to something that happened in Egypt in September 2011. This incident might have been the event when several thousand protesters violently broke into the Israeli embassy, which resulted in the return of several staff members to Israel and the declaration of a state of emergency by the Egyptian army (CNN, 2011). The increasing numbers of Syrian tweets including the hashtag Egypt in September 2011 might provide hints for a spill-over effect from Egypt to Syria, because the hashtag was mentioned 88 times – which is significantly higher than

(27)

22

the threshold of 10. Since the hashtag was repeatedly used in the same months in Egypt and Tunisia, the condition of a limited period of time between the mentioning in the trendsetting and the receiving countries is fulfilled.

Figure 2: Amount of usage of hashtag Egypt by political actors in Egypt, Syria and Tunisia.

The observations regarding the use of the term Egypt is nearly the same as for the hashtag Egypt. The data also shows a similar rise in the use of the term Egypt within Syria and Egypt in August and September 2011. However, an increasing use of the term Egypt in Egypt in July 2011 and April 2012 did not lead to any significant rise of use in Syria in the same period of time. Still, the term Egypt has been used 63 times in Syria in September 2011 and 203 times in the same month in Egypt. Therefore, the conditions for a possible spill-over effect are fulfilled.

0 100 200 300 400 500 600 Ja n-11 Fe b-11 M ar -11 A pr -11 M ay -11 Jun-11 Jul -1 1 A ug-11 Se p-11 O ct -11 N ov-11 D ec -11 Ja n-12 Fe b-12 M ar -12 A pr -12 M ay -12 Jun-12 Jul -1 2 A ug-12 Se p-12 O ct -12 N ov-12 D ec -12 Jan -11Feb-11 Ma r-11 Apr -11 Ma y-11 Jun -11Jul-11 Au g-11 Sep -11Oct-11 No v-11 De c-11 Jan -12Feb-12 Ma r-12 Apr -12 Ma y-12 Jun -12Jul-12 Au g-12 Sep -12Oct-12 No v-12 De c-12 Egypt 2 126 36 16 37 7 41 65 155 85 52 55 39 27 22 36 44 557 13 26 20 7 11 27 Syria 4 4 0 0 3 1 88 2 10 2 2 6 0 0 0 0 0 0 1 0 0 0 Tunisia 0 0 0 0 0 0 0 0 0 0 0 0 2 0 3 0 0 0 0 0 0 0 0 0

(28)

23

Figure 3: Amount of usage of term Egypt by political actors in Egypt, Syria and Tunisia.

Hashtag Tunisia

The Hashtag Tunisia has been used in all three countries intensively at different times. Overall, it was not used to the same extent as the hashtags Syria and Egypt. This might be due to the problem of limited availability of data. It is possible that the hashtag would normally be used frequently by political actors in Tunisia, but only a small number of tweets from a reduced number of political actors in Tunisia could be retrieved. Therefore, the data for this hashtag does not provide much information on a possible spill-over effect to Syria.

0 50 100 150 200 250 Jan-11Fe b-11 M ar-11 Apr-11_May-11 Jun-11 Jul-1 1 Aug-11 Se p-11 Oct-11 Nov-11 Dec-11Ja n-12 Fe b-12 M ar-12 Apr-12_May-12 Jun-12 Jul-1 2 Aug-12 Se p-12 Oct-12 Nov-12 Dec-12

Jan-11 Feb-11Mar-11Apr-11 Ma y-11 Jun -11Jul-11 Au g-11 Sep -11Oct-11 No v-11 Dec

-11Jan-12 Feb-12Mar-12Apr-12 Ma y-12 Jun -12Jul-12 Au g-12 Sep -12Oct-12 No v-12 Dec -12 Egypt 7 40 26 47 104 104 150 118 203 137 47 107 59 83 113 145 117 145 54 86 96 77 104 145 Syria 0 0 1 0 1 3 63 8 1 3 4 2 2 0 0 3 0 0 1 0 1 1 Tunisia 9 5 0 4 1 0 1 4 0 1 1 0 10 0 2 2 3 3 4 4 1 3 5 1

(29)

24

Figure 4: Amount of usage of hashtag Tunisia by political actors in Egypt, Syria and Tunisia.

Hashtag USA

The hashtag USA was used in all three countries a limited amount of times but only in Syria it has been mentioned more often than the minimum of 10. One explanation could be that the hashtag USA was mainly used by members of the opposition within all three countries to call for support in the protest against the regime and to raise attention to Human Rights violations by the regimes. The data for Egypt shows an increase of use in March and July/August 2011, followed by a rise of the use of the hashtag USA in Syria in April and August/September 2011. This might be a hint that political actors in Syria were influenced by political actors in Egypt in using the hashtag and bringing attention to the role of the USA in the Arab Spring. By using the hashtag repeatedly, often throughout several countries, the pressure on the US government to interfere and support the protesters might have risen. However, because the hashtag has not been mentioned/used in Egypt and Tunisia at least 10 times, the data for the hashtag USA does not provide strong enough indications for a possible spill-over effect.

0 5 10 15 20 Ja n-11 F eb-11 M ar -11 A pr -11 M ay-11 Jun-11 Jul -11 A ug-11 S ep-11 O ct -11 N ov-11 D ec -11 Ja n-12 F eb-12 M ar -12 A pr -12 M ay-12 Jun-12 Jul -12 A ug-12 S ep-12 O ct -12 N ov-12 D ec -12 Jan -11 Fe b-11 Ma r-11 Ap r-11 Ma y-11 Jun -11-11Jul Au g-11 Se p-11 Oct -11 No v-11 De c-11 Jan -12 Fe b-12 Ma r-12 Ap r-12 Ma y-12 Jun -12-12Jul Au g-12 Se p-12 Oct -12 No v-12 De c-12 Egypt 3 10 0 0 0 0 0 2 1 0 0 2 0 0 4 1 0 0 0 0 0 0 0 0 Syria 0 0 0 0 0 0 12 0 0 0 0 3 0 0 1 0 0 0 0 0 0 0 Tunisia 0 0 0 0 1 0 0 0 0 0 1 0 15 8 14 11 10 9 0 3 2 0 0 0 Use of hashtag Tunisia

(30)

25

Figure 5: Amount of usage of hashtag USA by political actors in Egypt, Syria and Tunisia.

Term Violence

The term violence has been used by political actors of all three countries, though again (probably due to the limited data for Tunisia), the data for Tunisia does not give much insight. In this case, it rather seems like the increase of the use of the term violence first took place in Syria and was followed by Egypt. It was mentioned 15 times in Syrian tweets (September 2011) and two month later 10 times in Egypt (December 2011). Even though the data shows that the term was first used in Syria and then in Egypt, it can be determined that violence was a topic in several tweets of political actors in all three countries. It seems that the data confirms a spill-over effect between all three countries to some extent but because it was not chronologically used first in Egypt and Tunisia and then in Syria, the data does not provide any information in the direction useful for this paper.

0 5 10 15 20 25 Ja n-11 F eb-11 M ar -11 A pr -11 M ay-11 Jun-11 Jul -11 A ug-11 S ep-11 O ct -11 N ov-11 D ec -11 Ja n-12 F eb-12 M ar -12 A pr -12 M ay-12 Jun-12 Jul -12 A ug-12 S ep-12 O ct -12 N ov-12 D ec -12 Ja n-11 Fe b-11 M ar-11 Ap r-11 M ay-11 Ju n-11 Jul -11 Au g-11 Se p-11 Oc t-11 No v-11 De c-11 Ja n-12 Fe b-12 M ar-12 Ap r-12 M ay-12 Ju n-12 Jul -12 Au g-12 Se p-12 Oc t-12 No v-12 De c-12 Egypt 0 3 4 2 1 0 6 5 4 4 0 3 1 2 6 2 1 1 1 0 2 5 1 1 Syria 4 5 0 1 0 0 22 3 1 1 0 0 0 0 0 0 1 0 0 1 1 0 Tunisia 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 4 0 0 0 Use of hashtag USA

(31)

26

Figure 6: Amount of usage of term violence by political actors in Egypt, Syria and Tunisia.

Term Human Rights

The term Human Rights has not been used as often by the investigated political actors of the three countries as I expected. In contrast, the data shows that its use was pretty low. Still, even though the data does not show a spill-over timewise from Tunisia and Egypt to Syria, it shows that the topic Human Rights was present in tweets of political actors in all three countries. However, it has not been mentioned in any of the countries at least 10 times. This indicates that there might have been an influence between the countries concerning the Human Rights conditions, but it does not provide any indications for a spill-over effect according to the measurement strategy of this paper.

0 2 4 6 8 10 12 14 16 Ja n-11 Fe b-11 M ar -11 A pr -11 M ay -11 Jun-11 Jul -1 1 A ug-11 Se p-11 O ct -11 N ov-11 D ec -11 Ja n-12 Fe b-12 M ar -12 A pr -12 M ay -12 Jun-12 Jul -1 2 A ug-12 Se p-12 O ct -12 N ov-12 D ec -12 Jan -11Feb-11 Ma r-11 Apr -11 Ma y-11 Jun -11Jul-11 Au g-11 Sep -11Oct-11 No v-11 De c-11 Jan -12Feb-12 Ma r-12 Apr -12 Ma y-12 Jun -12Jul-12 Au g-12 Sep -12Oct-12 No v-12 De c-12 Egypt 2 0 1 0 2 2 1 3 3 5 4 10 0 3 0 2 7 7 1 4 2 3 7 8 Syria 0 4 0 0 1 0 15 0 0 0 2 6 1 1 0 0 0 1 0 0 0 0 Tunisia 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 3 0 0 0 1 2 4 0 0

(32)

27

Figure 7: Amount of usage of term Human Rights by political actors in Egypt, Syria and Tunisia.

Term Mubarak

The term Mubarak is used the most in Egypt, as to be expected given that Mubarak was the former dictator of Egypt and therefore a target of the protest and revolution within Egypt. One plausible reason why the use of the term Mubarak increased in August 2011 in Egypt and then in September 2011 in Syria is that the trial against the former president of Egypt continued in August, after it had to be paused due to Mubarak’s health condition earlier in 2011 (Spiegel, 2011). It is likely that political actors in Syria referred to the beginning of the trial against the former ruler in Egypt to call for the overthrow of their own ruler, Bashar Al Assad. Maybe, political actors in Syria used the trial against Mubarak to call for a trial against Bashar Al Assad, too. In both countries, it has been mentioned in tweets more often than the minimum amount necessary. Therefore, the data seems to confirm a possible spill-over effect from Egyptian political actors to Syrian ones.

(33)

28

Figure 8: Amount of usage of term Mubarak by political actors in Egypt, Syria and Tunisia.

Term Revolt/Protest/Revolution

The terms revolt, protest and revolution were pooled because the Arabic word that has been used throughout the tweets can be translated with all three English words. The data shows that it has been used consistently by political actors in all three countries throughout the whole time period consulted in this research. However, that data shows an increased use of the term in Egypt and Syria in Summer 2011 (July and August) with its maximum in September 2011. Political actors in both countries increasingly used the terms again in early 2012 (January and February). Observing that the use of the term in both countries occurred at the same time and that it has been mentioned in both countries more often than 10 times, suggests that Syrian political actors might have been influenced trough a spill-over effect by tweets of Egyptian political actors.

(34)

29

Figure 9: Amount of usage of term Revolt/Protest/Revolution by political actors in Egypt, Syria and Tunisia.

Term Freedom

Throughout the whole time period that is covered in this paper, the most-used term was freedom, especially in Egypt and Syria. Both countries have their highest use of the term by political actors in tweets around similar times, for example in April/May 2011 and September/October 2011. Even though it was mentioned more often in Syrian tweets in total, the data shows that it has also been mentioned more than 10 times in the specific months in Egypt. The term freedom seems to be a term associated with protests and the Arab Spring in general and its use in similar months in Egypt and Syria does seem to confirm possible spill-over effects.

0 50 100 150 200 250 300 Ja n-11 Fe b-11 M ar-11 Apr-11M ay-11 Ju n-11 Jul-11_Aug-11_Sep-11_Oct-1 1 Nov-11 Dec-11Ja n-12 Fe b-12 M ar-12 Apr-12M ay-12 Ju n-12 Jul-12_Aug-12_Sep-12_Oct-1 2 Nov-12 Dec-12

Jan-11 Feb-11Mar-11Apr-11 Ma

y-11

Jun-11 Jul-11 Aug-11Sep-11Oct-11Nov-11Dec-11Jan-12 Feb-12Mar-12Apr-12 Ma

y-12

Jun-12 Jul-12 Aug-12Sep-12 Oct-12Nov-12Dec-12 Egypt 1 15 15 32 47 26 63 62 105 41 19 34 52 42 17 38 34 45 13 21 17 37 28 35 Syria 0 2 0 0 36 33 262 40 30 13 35 21 20 12 12 8 6 4 5 3 14 11

Tunisia 0 3 0 2 0 3 1 0 0 2 2 0 5 3 4 3 1 0 0 1 1 2 0 5

(35)

30

Figure 10: Amount of usage of term Freedom by political actors in Egypt, Syria and Tunisia.

Term Change

Analysing the data for the use of the term change in tweets by political actors of all three countries, it seems that the term has not been used in any temporal correlation between the countries. While the highest number of uses for Tunisia and Egypt was in early 2011 (January), actors in Syria used it later in 2011 (September). Therefore, no correlation between the use of the term in Egypt/Tunisia and Syria can be proven. Although the term was not mentioned at least 10 times in a specific month, its total use is higher than 10 times in total. However, the period of time between the uses of the term is greater than 2 months between the countries. Therefore, the data might show hints for a possible spill-over effect to Syrian political actors but does not comply with the defined criteria for this analysis. 0 20 40 60 80 100 120 Ja n-11 Fe b-11 M ar-11 Apr-11M ay-11 Ju n-11 Jul-11_Aug-11_Sep-11_Oct-1 1 Nov-11 Dec-11Ja n-12 Fe b-12 M ar-12 Apr-12M ay-12 Ju n-12 Jul-12_Aug-12_Sep-12_Oct-1 2 Nov-12 Dec-12

y-11

y-12

Jun-12 Jul-12 Aug-12Sep-12 Oct-12Nov-12Dec-12 Egypt 1 5 3 7 17 4 14 8 16 16 9 5 13 13 9 15 15 23 2 12 13 12 12 27

Syria 0 17 5 0 18 10 97 38 10 21 14 4 2 3 11 4 2 2 0 0 1 0

Tunisia 0 1 0 0 0 0 0 0 0 0 0 0 0 1 2 3 1 0 0 1 0 0 3 0

(36)

31

Figure 11: Amount of usage of term Change by political actors in Egypt, Syria and Tunisia.

Term Cooperation

The last term that is examined is the term cooperation. When tracking the use of the term, no differentiations were made regarding the context the term has been used in by the various political actors. However, it seems plausible that the term either refers to cooperation among several groups with the same interests within each country or between the different states. While the term is mostly used in Syria in September 2011, it is then mainly used in Tunisia and Egypt in May and June 2012. Because the period between September 2011 and May 2012 exceeds the threshold of two months and the mentioning of the term cooperation in all three countries is below the minimum of 10, the use of the term cannot provide any evidence for a possible spill-over effect between the political actors in Egypt, Tunisia and Syria.

0 2 4 6 8 10 12 Jan-11Fe b-11 M ar-11 Apr-11M ay-11 Ju n-11 Jul-11_Aug-11_Sep-11_Oct-1 1 Nov-11 Dec-11Ja n-12 Fe b-12 M ar-12 Apr-12M ay-12 Ju n-12 Jul-12_Aug-12_Sep-12_Oct-1 2 Nov-12 Dec-12

y-11

y-12

Jun-12 Jul-12 Aug-12Sep-12 Oct-12Nov-12Dec-12

Egypt 10 3 0 4 3 3 6 4 1 4 1 4 5 2 5 7 2 4 3 5 3 1 2 8

Syria 0 1 0 0 0 0 7 3 0 1 0 0 0 0 1 0 1 0 0 0 0 1

Tunisia 5 2 0 1 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 1 0 0 0 0

(37)

32

Figure 12: Amount of usage of term Cooperation by political actors in Egypt, Syria and Tunisia.

5.2.3. Summary of results for hypothesis 2

After analysing the terms and hashtags I identified as most relevant as evidence for a possible spill-over effect from Egypt and Tunisia to Syria, I summarise the findings and explain if the hypothesis 2 can be verified or needs to be falsified.

The data for some of the hashtags and terms (including the terms Change, Revolt/Protest/Revolution, Mubarak, Violence and Egypt and the hashtags Egypt) seems to support hypothesis 2 because they were used often either over a short period of time, and first appeared in Egypt and Tunisia before gaining traction in Syria, or they were mentioned in the same months in Egypt/Tunisia and Syria. Thus, though they fulfill the two conditions, no definitive evidence can be found that the increasing use of the terms in Syria is directly correlated with an increasing use

0 1 2 3 4 5 6 Ja n-11 Fe b-11 M ar-11 Apr-11_M ay-11 Ju n-11 Jul-11_Aug-11_Sep-11_Oct-1 1 Nov-11 Dec-11Ja n-12 Fe b-12 M ar-12 Apr-12_M ay-12 Ju n-12 Jul-12_Aug-12_Sep-12_Oct-1 2 Nov-12 Dec-12

y-11

y-12

Jun-12 Jul-12 Aug-12Sep-12 Oct-12Nov-12Dec-12

Egypt 0 1 0 0 2 2 2 0 1 3 0 1 0 1 1 0 0 4 1 3 3 2 3 5

Syria 0 0 0 0 0 2 4 0 3 0 2 1 2 1 1 0 0 1 0 0 1 1

Tunisia 0 1 0 0 0 0 0 0 1 0 0 0 0 0 1 1 5 0 0 0 0 1 3 0

(38)

33

in Egypt and Tunisia beforehand. The data only allows me to assume that there has been some indirect communication and emulation between political actors in Egypt, Tunisia and Syria. However, the data of some of the analysed terms, including the terms Tunisia, Freedom, Cooperation, as well as the hashtag Syria, seem to not provide enough evidence for a spill-over effect and therefore do not confirm my hypothesis.

Besides, one needs to remember that only 11 of the total 40 hashtags and 63 terms were analysed in detail. Not all hashtags and terms could be directly linked to political events, protest and the uprising of the Arab Spring. As previous research of scholars like Weidmann (2019) and Salehyan & Gleditsch (2006) has shown, the dynamics of ethnic divisions and spill-over effects can increase the potential for the breakout of a civil war. Therefore, my second hypothesis can neither be completely verified nor falsified. The data suggests that there is the possibility of a spill-over effect but due to the above-mentioned limitations, the data does not allow for any definite conclusions. Therefore, hypothesis 2 can only be verified to a limited extent. A list of all political actors whose tweets were included in this analysis can be found in the Appendix (Table 4).

5.3. Hypothesis 3: The influence of ethnicity across national borders

5.3.1. General explanation of the data

The following section presents the explanation of the data that was used in order to find evidence for hypothesis 3. The third hypothesis “Syrian ethnic group leaders were influenced by group

leaders of the same ethnicity in neighbouring conflicts in order to mobilise for protest” was

planned to be conducted with the same data as hypothesis 2. After the political actors’ tweets were identified with the help of the API crawler for 2011 and 2012, I tried to determine the ethnic identity of group leaders within the conflict parties of the Arab Spring and uprising protest in Egypt, Syria

(39)

34

and Tunisia with the help of secondary resources, such as newspapers, reports and journals. Unfortunately, I could only identify the ethnic identity of a small number of political actors. Therefore, the hypothesis is tested by means of an inductive example. The results of this cannot provide strong support for the hypothesis but rather indicate a trend of whether transnational ethnic influence has played a role in motivating Syrian ethnic leaders to mobilise for protest. The inductive example considers two political actors whose ethnic identity of Sunni Arab could be confirmed: Abd-al-Mun'im Abu-al-Futuh from Egypt and Moaz al-Khatib from Syria. This test analyses whether a spill-over effect of the use of terms and hashtags in the tweets by the two political actors can be validated. Once again, not all 40 hashtags and 60 terms are examined. The hashtags and terms that are included in this test need to fulfil two aspects: they are used at least once in the tweets by one of the political actors and its content shows a correlation to political events happening in the frame of the Arab Spring. This leads to 19 hashtags and terms that are analysed in this inductive example to evaluate whether indications for a spill-over effect from Abd-al-Mun'im Abu-al-Futuh (Egypt) to Moaz al-Khatib (Syria) can be confirmed.

(40)

35

5.3.2. Analysis of the terms and hashtags used by Abd-al-Mun'im Abu-al-Futuh (Egypt) and Moaz al-Khatib (Syria)

Figure 13: Amount of use of hashtags/terms by Abd-al-Mun'im Abu-al-Futuh (Egypt) and Moaz al-Khatib (Syria).

Hasht ags_e gypt Hasht ags_p alesti ne Hasht ags_t ahrir Hasht ags_j an25 Hasht ags_u sa Hasht ags_j or Term s_Eg ypt Term s_de mons tratio n Term s_vio lence Term s_inj ury Term s_Hu man Right s Term s_Pal estine Term s_Mu barak Term s_Ba shar AlAs sad Term s_ma rtyr Term s_Re voltP rotest Revo lution Term s_cor rupti on Term s_stri ke Term s_Fre edom Abd-al-Mun'im Abu-al-Futuh 128 1 109 22 2 4 566 4 3 5 1 14 46 1 5 201 13 2 60 Moaz al-Khatib 0 0 0 0 0 0 0 0 0 0 0 2 0 0 2 4 0 0 0 0 100 200 300 400 500 600

Use of hashtags/terms by Abd-al-Mun'im Abu-al-Futuh (Egypt) and Moaz al-Khatib (Syria)

(41)

36

The data shows that only three of the hashtags and terms were used by both actors, namely the terms Palestine, Martyr and Revolt/Protest/Revolution. The data shows that the term Palestine has been used 14 times by Abd-al-Mun'im Abu-al-Futuh and 2 times by Moaz al-Khatib. The term Martyr has been used 5 times in the tweets by Abd-al-Mun'im Abu-al-Futuh and 2 times by Moaz al-Khatib. The term Revolt/Protest/Revolution was used 201 times by Abd-al-Mun'im Abu-al-Futuh and 4 times by Moaz al-Khatib. That these three terms have been used by both political actors might indicate a potential spill-over effect (especially when they can be traced from Egypt to Syria). Consequently, the following figures show in which month the three terms were used by Abd-al-Mun'im Abu-al-Futuh and Moaz al-Khatib:

Figure 14: Amount of use of the terms Palestine, Martyr and Revolt/Protest/Revolution by Abd-al-Mun'im Abu-al-Futuh. 0 10 20 30 40 50 60 70 Ja n-11 Fe b-11 M ar-11 Apr-11M ay-11 Ju n-11 Jul-11_Aug-11_Sep-11_Oct-1 1 Nov-11 Dec-11Ja n-12 Fe b-12 M ar-12 Apr-12M ay-12 Ju n-12 Jul-12_Aug-12_Sep-12_Oct-1 2 Nov-12 Dec-12 Jan -11 Fe b-11 Ma r-11 Ap r-11 Ma y-11 Jun -11-11Jul Au g-11 Se p-11 Oc t-11 No v-11 De c-11 Jan -12 Fe b-12 Ma r-12 Ap r-12 Ma y-12 Jun -12-12Jul Au g-12 Se p-12 Oc t-12 No v-12 De c-12 Hashtag_Palestine 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Term_Martyr 0 0 0 0 0 1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 Term_RevoltProtestRevolution 0 0 0 3 5 7 0 22 62 28 3 0 5 4 1 5 3 4 1 3 0 6 4 0

The outbreak of the Syrian Civil War; Ethnic division and spill-over effects from neighbouring conflicts

Leiden University

Department of Political Science

Master Thesis

The outbreak of the Syrian Civil War

–

Ethnic division and spill-over effects from

neighbouring conflicts

By Caroline Ohlendorf, s2421615

Submitted to Prof. Dr. Nicolas Blarel and Prof. Dr. Roos van der Haer

Word count: 10 280

Table of Contents

Table of Figures

Table of Tables

1. Introduction

2. Literature Review

2.1. Ethnic division and ethnic grievances as factors accounting for civil

war outbreak

2.2. Political Opportunity Structure as a driver for civil war outbreak

2.3. Transnational influence on civil war outbreak

3. Theoretical Framework

3.1. Hypothesis 1: Ethnic grievances and unequal distribution of power

along ethnic lines

3.2. Hypothesis 2: The role of spill-over effects from neighbouring

countries for the providence of new Political Opportunity Structures

3.3. Hypothesis 3: The influence of ethnicity across national borders

3.4. Main Argument

4. Research Design

4.1. Method

4.2. Case Selection

4.3. Data Collection

4.4. Variables

5. Empirical Section

5.1. Hypothesis 1: Ethnic grievances and unequal distribution of power

along ethnic lines

5.2. Hypothesis 2: The role of spill-over effects from neighbouring

countries for the providence of new Political Opportunity Structures

5.3. Hypothesis 3: The influence of ethnicity across national borders

Use of the terms Palestine, Martyr and Revolt/Protest/Revolution

by Abd-al-Mun'im Abu-al-Futuh