Education A Dark Force? A Qualitative Investigation of Education and Domestic Terrorism in the Middle East North Africa

Master Thesis in Peace and Conflict Studies Spring 2017 Department of Peace and Conflict Research Uppsala University Education A Dark Force? A Qualitative Investigation of Education and Domestic Terrorism in the Middle East North Africa Lucienne Heyworth Supervisor: Karen Brounéus 1

ACKNOWLEDGEMENTS This thesis would not have been possible without the support of select people. First, I would like to thank Rotary and the Rotary Club of Fitzroy specifically, for supporting my application to the Peace Fellowship and allowing me this opportunity as a Peace Fellow. I would also like to thank Karen Brounéus who has been an important mentor and who s diligence in providing comments and guidance on multiple drafts, as well as general encouragement throughout the process, has been very much appreciated. Thanks also to Christie, Rickard, Mikael and Alden, who have provided hours of entertainment and frivolity and without whom I may have cried more than laughed this semester. My husband Saif also deserves special mention. His unwavering encouragement and belief in and support of me throughout in the Masters programme has meant so much. Finally, I thank Yoda, from whom my causal mechanism owes some credit. 2

ABSTRACT Extant research on terrorism has predominantly examined conditions associated with transnational terror. Considerably less is known about home-grown or domestic terrorism despite its accounting for much of the non-state violence seen in the international system. While some have examined the relationship between education and political violence, less has been done to investigate qualitatively the relationship between education and domestic terrorism, particularly under the condition of corruption. Comparing the cases of Morocco, Libya, Jordan and Egypt between 1970-2010, I find that increases in education bear little connection with levels of domestic terror. However, empirics suggest that increases in education may play a role in individual abilities to recognise and react against, sometimes violently, perceived state corruption. Findings suggest a need for further disaggregated data on the perpetrators of terrorist violence to better understand the complex relationship between education and domestic terror. Key words: Domestic Terrorism, Education, Grievance, Frustration, Morocco, Jordan, Egypt, Libya 3

LIST OF ABBREVIATIONS... 5 LIST OF FIGURES... 6 LIST OF TABLES... 6 1. INTRODUCTION... 7 2. PREVIOUS RESEARCH ON EDUCATION AND TERORRISM... 10 2.1 TERRORISM... 10 2.2 EDUCATION... 14 2.3 THE RESEARCH GAP... 17 3. THEORETICAL FRAMEWORK... 19 3.1 DEFINITIONS... 19 3.2 THEORY, ARGUMENT AND CAUSAL EXPLANATION... 20 3.3 HYPOTHESIS... 26 4. RESEARCH DESIGN... 27 4.1 METHOD... 27 4.2 CASE SELECTION... 27 4.3 OPERATIONALISATION... 30 4.4 TIME-PERIOD AND DATA COLLECTION STRATEGY... 34 4.1 STRUCTURE OF ANALYSIS... 34 5. CASE I MOROCCO... 36 5.1 BACKGROUND & NATIONAL CONTEXT... 36 5.2 EDUCATION... 36 5.3 DOMESTIC TERRORISM... 38 6. CASE II LIBYA... 40 6.1 BACKGROUND & NATIONAL CONTEXT... 40 6.2 EDUCATION... 40 6.3 DOMESTIC TERRORISM... 42 7. CASE III JORDAN... 44 7.1 BACKGROUND & NATIONAL CONTEXT... 44 7.2 EDUCATION... 44 7.3 DOMESTIC TERRORISM... 46 8. CASE IV EGYPT... 48 8.1 BACKGROUND & NATIONAL CONTEXT... 48 8.2 EDUCATION... 48 8.3 DOMESTIC TERRORISM... 50 9. ANALYSIS... 52 9.1 WITHIN CASE ANALYSIS... 52 9.2 BETWEEN CASE COMPARISON... 55 9.3 EXTENDED ANALYSIS ALTERNATIVE EXPLANATIONS AND ADDITIONAL OBSERVATIONS... 62 9.4 LIMITATIONS AND BIAS... 65 10. CONCLUSIONS... 68 11. BIBLIOGRAPHY... 70 4

LIST OF ABBREVIATIONS TI GDP GNI USD HDI UCDP GTD ITERATE MENA UNDP UIS WENR TQJBR UNESCO Transparency International Gross Domestic Product Gross National Income United States Dollars Human Development Index Uppsala Conflict Data Program Global Terrorism Database International Terrorism: Attributes of Terrorist Events Middle East North Africa United Nations Development Fund UNESCO Institute for Statistics World Education News and Reviews Islamist Militants of the Tanzim Qaidat al-jihad fi Bilad al-rafidayn United Nations Educational, Scientific and Cultural Organisation 5

LIST OF FIGURES FIGURE 1 THE HYPOTHESISED RELATIONSHIP BETWEEN EDUCATION AND DOMESTIC TERRORISM... 23 FIGURE 2 PROPOSED CAUSAL CHAIN LINKING EDUCATION AND DOMESTIC TERROR... 23 FIGURE 3 CAUSAL MECHANISM STEP 1 - FROM EDUCATION TO EXPECTATIONS... 24 FIGURE 4 DISRUPTION CAUSED BY CORRUPTION TO TRADITIONAL EXPECTATIONS... 25 FIGURE 5 EDUCATION ENROLMENT RATIOS: MOROCCO 1970-2010... 37 FIGURE 6 EDUCATION ENROLMENT RATIOS: LIBYA 1970-2010... 41 FIGURE 7 EDUCATION ENROLMENT RATIOS: JORDAN 1970-2010... 45 FIGURE 8 EDUCATION ENROLMENT RATIOS: EGYPT 1970-2010... 49 FIGURE 9 CASE COMPARISON: TOTAL NUMBER OF DOMESTIC TERRORIST ATTACKS 1970-2010... 58 LIST OF TABLES TABLE 1 CASE SELECTION: LEVELS ON THE INDEPENDENT VARIABLE... 28 TABLE 2 INDICATORS FOR THE INDEPENDENT VARIABLE... 31 TABLE 3 INDICATORS FOR THE DEPENDENT VARIABLE... 33 TABLE 4 DOMESTIC TERROR ATTACKS: MOROCCO 1970-2010 BY ATTACK TYPE AND TARGET... 39 TABLE 5 DOMESTIC TERROR EVENTS: LIBYA 1970-2010 BY ATTACK TYPE AND TARGET... 42 TABLE 6 DOMESTIC TERROR ATTACKS: JORDAN 1970-2010 BY ATTACK TYPE AND TARGET... 47 TABLE 7 DOMESTIC TERROR ATTACKS: EGYPT 1970-2010 BY ATTACK TYPE AND TARGET... 51 TABLE 8 CASE RANKINGS ACCORDING TO ENROLMENT RATIO 1970-2010... 57 TABLE 9 CASE RANKING ACCORDING TO OVERALL LEVEL OF EDUCATION... 57 TABLE 10 CASES RANKING ACCORDING TO OVERALL NUMBER OF DOMESTIC TERROR ATTACKS... 58 TABLE 11 FINDINGS: EXPECTED AND OBSERVED OUTCOMES & SUPPORT FOR THE HYPOTHESIS... 60 TABLE 12 CASE COMPARISON: ATTACKS DIRECTED AT THE STATE (GOVERNMENT, POLICE, MILITARY)... 61 6

1. INTRODUCTION Previous research often argues that education is linked to peace. Quantitative studies suggest that education mitigates individual and group disempowerment and disenfranchisement by increasing contentment with domestic circumstances and raising opportunity costs for violence (Collier & Hoeffler 2004; Østby & Urdal 2010; Hegre 2003; Hibbs 1973; Huntington 1968). However, recent research suggests that education does not necessarily translate to peace: with some scholars finding higher levels of education amongst perpetrators of terrorism, suggesting education, at times, drives violence (Krueger & Malečková 2003; Barrebi 2007; Shafiq & Sinno 2010; Brockhoff, Krieger & Meierrieks 2010; Shafiq & Ross 2010). Until now, few have investigated the relationship between increasing levels of education and domestic terrorism. I conduct a qualitative analysis of four countries in the Middle East North Africa region to assess the effect of increases in education on domestic terror between 1970-2010. Considering changes in education levels, I investigate the role of education in generating frustrations in weak states, specifically under the scope condition of corruption. I argue that corruption plays a specific role in turning higher levels of education amongst some into a threat to domestic peace by altering the opportunity costs for violence and delivering increasing capacities within which to recognise the effect of corruption in limiting the ability fulfil potential individual potential through traditional channels. By considering the effect of increased education levels in a causal pathway linking education to expectations and frustration to ultimately its manifestation in terrorist violence, I seek to answer the research question Under what conditions does increased (higher levels of) education lead to domestic terrorism? I seek to contribute to research on education and terrorism in four ways. First, I consider the potential negative effects of increased levels of education on specifically domestic terrorism. Second, I investigate a causal process toward violence, suggested by previous quantitative research, using a qualitative case comparison of Morocco, Libya, Jordan and Egypt. Third, I add to existing theoretical work and contribute to current gaps on the causal processes linking education and violence by considering the effects of increased levels of education under the specific condition of corruption. Finally, by conducting the above, I respond to calls within the discipline on terrorism to focus on investigation of more nuanced and uncertain relationships between terror and its determinants (Krieger & Meierrieks 2010). 7

Building on the work of Dollard and Miller (1931), Gurr (1970) and Crenshaw (1981) who propose the value of the frustration-aggression mechanism, relative deprivation and group grievance models to explain the resort to violence, I argue that raising levels of education affects individual expectations, and that when expectations and reality are mismatched, increases in education can work to heighten dissatisfactions, grievances and frustrations that can lead some to resort to terror. The greater awareness brought about by education can also heighten the likelihood of a resort by some to acts domestic terror when met with and exacerbated by corrupt, unfair and unjust systems, whilst further education increases the value of such individuals for enlistment in terrorist groups. This causal link between education, corruption and terrorism leads to the hypothesis that... Given corruption, countries with high levels of education will experience higher levels of domestic terrorism. To test my hypothesis, I conduct a structured focussed comparison of cases. Empirics are reported first, followed by analysis of each case after which I compare cases and assess the validity of assumptions. Empirics suggests little support for the hypothesis and limited support for the causal mechanism. This study provides, however, valuable insight into the relationship between education and domestic terrorism suggesting that the relationships complexity is the result of a possible interplay between varied factors rather than consequential on the presence only of corruption. Findings adds credence to growing interest in the policy world on the potential negative impacts of education and that seek to nuance the otherwise dominant propositions of education as a silver bullet for complex social and economic problems (Ghanem 2017). Findings are also in line a new and emerging wave in debates on education andterror which argue for more nuanced empirical investigation by suggesting the interplay of a multiple of factors affecting individual and group deliberations on violence (Krieger & Meierreiks 2010; Brockhoff et al. 2015; Shafiq & Sino 2010; Kavanagh 2011). The only recent separation of data on terrorism into domestic and transnational types has so far limited investigations into the determinants of domestic terror and thus much more can be done to test the applicability causal mechanisms for domestic terrorist violence. I suggest specifically further research comprising qualitative and quantitative components to better bridge gaps between macro level circumstances and individual motivations. In the next section I briefly review previous literature on terrorism determinants and education and violence, and explore the limited scholarship on education and terror. Next, I present the 8

theoretical framework in support of the argument that given corruption, increased education will lead to an increase in domestic terrorism. In section three I describe the research design of this study and the operationalisation and measurement of key variables. Section four presents the results of empirical data for each case. In analysis, I consider each case briefly in relation to the causal mechanism and hypothesis, before engaging in the comparative analysis of cases. I also outline alternative explanations and additional observations and acknowledge any bias and limitations. In the final section I provide conclusions and suggest avenues for further research. 9

2. PREVIOUS RESEARCH ON EDUCATION AND TERORRISM In this section I introduce previous research into the determinants of terrorism and provide an overview of research on education and violence. I also specify the research gap to which this study seeks to contribute. 2.1 Terrorism Research on terrorism has burgeoned in recent years making a comprehensive literature review beyond the scope of this study. What follows is an overview of relevant theoretical and empirical research that acts as a frame to this investigation. Domestic Versus International Terrorism Several streams of scholarship seek to understand terrorism. Theoretical underpinnings apply to greater or lesser extent to either domestic or transnational terrorism. Domestic terrorism is considerably more common phenomena than its transnational counterpart i, yet its coverage in available datasets to date has been limited (Enders, Sandler & Gaibulloev 2011: 319). The first dataset to include and differentiate between domestic and international terrorism over time became available in the late 2000s (Brockhoff et al. 2015: 1211). The result of data constraints has been predominance of empirical investigations into the determinants of terror which focus on the transnational (Freytag et al. 2011: 59). Differences between domestic and international terrorism are important however, as a focus on one or the other without acknowledgement of their differences can produce misleading results ii (Freytag et al. 2011: 59). Nevertheless, the core theoretical arguments of existing research canvassed below demonstrate predominant understandings of the terrorist phenomena. Terrorist Recruitment and Selection The factors that influence an individual s likelihood to seek out or be recruited to a terrorist cause are important. Investigations into the determinants of terrorism are often framed around the calculations that affect the recruitment and selection of individuals for violence and build on predominantly rationalist choice theories iii. Individuals are assumed as rational actors that make decisions in support of the achievement of positive outcomes whilst avoiding negative ones. Opportunity costs calculations in relation to terrorism, are then the calculations made by any individual of the potential benefits and disadvantages of participation in terrorist activity and include weighting potential benefits, such as money, power, position, or recognition against the 10

disadvantages including punishment, pain or shame. An individual s opportunity costs calculations also consider the benefits and disadvantages presented by success or failure in achieving the aims of terrorist activity. Empirical research that seeks to explain the determinants of terror often rely explicitly or implicitly on the assumption of terrorists as rational actors and consider specifically the economic, political, social and ideological factors that may influence individual and group motivations by effecting or altering opportunity cost calculations. Terrorists as rational actors and the importance of opportunity cost calculations thus act as an important frame in considering the prevalent determinants of terrorism canvassed below. Theorised Determinants of Terror Political Many empirical studies argue political and demographic causes for terrorism (Krueger & Malečková 2003; Tavares 2004; Abadie 2006; Kurrild-Klitgaard et al. 2006; Dreher & Gassebner 2008; Kruegar & Laitin 2008; Basuchoudary & Shughart 2010; Choi 2010; Kris-Katos et al. 2011; Freytag et al. 2011). Political theories often argue the role of political systems in shaping the likelihood of terrorist violence. The positive and negative implications of varied regime types are often highlighting, for affecting the opportunity costs of violence. Democracies are argued to offer greater opportunities for non-violent dissent and to thus lessen the number of terrorist attacks. Democracies are unable, however to realise harder (more punitive) counter terrorism measures as a result of civil liberty obligations (Lai 2005). Autocratic regimes in contrast, can more easily exercise repression, but in doing so risk generating grievances resulting from greater political disenfranchisement (Krieger & Meiericks 2010). The effect of government capacity on terrorism has also been explored in research pointing to the importance of characteristics of states. The use of opposition repression and punitive terrorist policies or the unpredictable use of restraint as highlighted as factors that can determine either high or low opportunity costs for operation of terrorist groups (Fearon & Laitin, 2003; Barrebi & Lakdawalla 2007; Li, 2005; Li & Shuab 2004; Drakos & Gofas, 2006 in Lai 2007). Social welfare policies have also been argued to reduce the risk of terrorism by reducing economic insecurity, inequality, poverty and religiouspolitical extremism (Burgoon 2006). Others have suggested that social policies can have the opposite effect and lead instead to disgruntlement and the clustering of certain groups whilst also providing resources to channel energies toward extremism (Kaus 2001). 11

Demographic Demographic factors are also suggested to have positive and negative effects on the generation of terrorism (Gassebner & Luechinger 2011; Travers 2004; Basuchoudhary & Shughart 2010; Krieger & Meierrieks 2010). Young populations are seen by some as raising the likelihood of terror (Tavares 2004; Urdal 2006) while others point to less violence in countries with growing populations (Dreher & Fischer 2010). High numbers of youth between 15-24 has also found empirical support for having a positive effect on (to increase) terrorism. Urdal (2006) argues youth bulges increase risks of spontaneous violence, including an increased risk of terrorism (Østby & Urdal 2010: 14). The influence of ethnicity and identity has also been shown to have positive links to the likelihood of terror. Ethnic/linguistic fractionalisation is shown to increase tensions that can lead to violence but is argued as distinct from religious fractionalisation (Gassebner & Leuchinger 2011: 239). Limitations to the explanatory role of ethnic/linguistic fractionalisation are highlighted however by pointing to the superior measure of ethnic tension (Basuchoudhary & Shughart 2010). The role of modernisation processes has also been highlighted for generating marginalisation and alienation and for exacerbating grievances by placing pressure on economic capacities, lifestyles and ideas (Ross 1993). Grievances manifest by the tensions of modernisation are argued ripe for manipulation by terrorist groups (Krieger & Meierricks 2011). Finally, the presence of minority groups and specifically minority group economic discrimination has been found as a positive correlate with violence and to include positive links with domestic terror (Enders et al. 2011: 319). Economics That terrorism is rooted in economic determinants is perhaps most widely-argued by academics and politicians alike (Enders & Sandler 2006; Caplan 2006; Bueno de Mesquita 2008; Shughard 2010). Economic determinant arguments are typically two-fold focusing on grievance and opportunity costs. The focus on grievance argues the role of economic complaints in generating terrorist violence. Economic development, measured in Gross Domestic Product (GDP), GDP per capita or Gross National Income (GNI) is used to highlight the extent to which poor economic circumstances are responsible for grievance (Gassebner & Leuchinger 2011). This thinking is extended by the work of scholars such as Gurr (1970) in his theory of relative deprivation. In it he argues that frustration is caused not only by poor economic circumstances, but by discrepancy between what individuals think they deserve and what they receive. The second group of arguments use the rationalist choice model to understand the decisions of 12

would be terrorists relating specifically to incentive structures for violence. Assuming individuals engage in violence based on a determination of relative costs, benefits and opportunities, scholars argue poor economic circumstances lessen the opportunity costs of violence and therefore increase the incidence of terror (Freytag et al. 2011). It follows then that successful economic development (measured in GDP per capita) significantly reduces the genesis of terror (Blomberg & Hess 2008; Azam and Delacroix 2006; Lai 2007 in Azam & Thelen 2008). Recent empirical investigation cast doubt on the value of economic arguments in explaining terrorism by pointing to only weak links (Gassebner & Leuchinger 2011: 238; Krueger & Malečková 2003; Kurrild-Klitgaard et al. 2006; Basuchoudhary & Shughard 2010; Krueger & Laitin 2007; Freytag et al. 2011). In their seminal work, Krieger and Meiericks (2010) find broadly conflicting all existing research in support of poor economic conditions as determinants of terror (Krieger & Meiericks 2010). Gassebner and Leuchinger (2011) investigating 65 proposed correlates using economic development measures similarly find only weak support. More recent investigation points to, for greater explanatory power, links between poor socioeconomic conditions measured in insufficient welfare policies, economic discrimination and low levels of economic openness in generating the incidence of terror (Burgoon 2006; Blomberg & Hess 2008; Krieger & Meierrieks 2011; Caruso & Schnieder 2011). Ideology A final category of theory on the determinants of terror focuses on the influence of ideology. Ideology is considered to effect rationalist choice models underlying terrorist recruitment and selection. Bernholz (2004) argues that the supreme values of ideology leave little room for opportunity costs as an explainer of terrorism. Ideology disrupts traditional cost-benefit analysis through the introduction of universalist and non-material incentives (Bernholz 2004). Some religious and ideological motivations are determined particularly notable for effects of time discounting ; in such cases the value of the present declines compared with the value of the future. Pain in the present is tolerable compared to eternal peace in the hereafter thus challenging traditional opportunity cost calculations assuming avoidance of pain/punishment. Notable here are ideologies placing greater emphasis on a next life seen as encourage individuals to focus on the afterlife, rather than material or immaterial costs in the present (Atran 2006). 13

2.2 Education The role of education has been most often considered in the context of political violence and civil war. Some investigation has been conducted into the relationship between education and terrorism, with little on specifically domestic terror. Education and Political Violence Education s role contributing to or mitigating political violence is most often considered in the context of civil war. Measures of education focus on levels obtained but also on education expansion, inequality or content (Østby & Urdal 2010: 2). Education is argued to improve personal and society wide conditions measured in higher incomes, stronger economic growth and democratisation,. By improving individual circumstances education thus ameliorates grievances that might otherwise lead to violence (Brockhoff et al. 2015: 1203). Increased educational attainment is also argued to reduce conflict by raising the opportunity costs of rebellion (Collier & Hoeffler 2004). By increasing thinking, reasoning and empathy skills education increases the likelihood of employment and lessens the need to resort to criminality and violence. Increases in educational attainment thus have the effect of lessening the pool of potential rebel recruits by heightening opportunities for legal work and income generation (Barakat & Urdal 2009). Finally, higher educational attainment is sometimes argued to reduce political violence by encouraging political participation, specifically the access and means to meaningfully contribute to formal and informal pathways for conflict resolution (Alesina & Perotti 1996; Hegre 2003; Hibbs 1973; Huntington 1968). The potential negatives effects of education have, however, been cautioned, specifically increases in education without attention to employment opportunities. The inability of labour markets to absorb increasingly educated populations has potential to agitate grievances that might turn violent (Urdal 2006; Huntington 1968; Choucri 1974; Goldstone 2001; Lai 2005). In line with relative deprivation theories, education inequality related to both access and attainment, can also breed grievances that could [ } cause conflict (Østby & Urdal 2010: 5). Socio-economic or political inequalities born of education inequality have also been highlighted to coincide with identity cleavages and may enhance group grievances and thus facilitate mobilisation for engagement conflict and violence (Østby & Urdal 2010: 5). Some scholars even propose education as a main driver of socio-economic inequality (Ferranti et all 2004 in Østby & Urdal 2010: 5). Expansion of low-quality education has similarly been highlighted as a risk by raising 14

expectations unmatched by employment opportunities (Salehi-Isfahani & Navtej 2008: 6) Finally curriculum has been seen as a potential means to indoctrinate and formalise problematic views in and between societies that can increase the likelihood of violence (Sambanis 2005). Education and Terrorism Current research on education and terrorism often proposes education for lessening the likelihood of terrorist violence. Again, it is assumed that education positively shapes personal attitudes toward extremist ideologies and affects individual perceptions on the legitimacy of violence (Brockhoff et al. 2015: 1203). A negative correlation between education and terrorism (that education lessens the likelihood of terrorism) is also highlighted by pointing to the high incidence of attacks in countries with low literacy levels (Bravo & Dias 2005). Others have argued that increased education provides individuals the ability to better evaluate the exceptionally low probabilities of terrorist success (Brockhoff et al. 2015). Some find however that that a positive relationship (education increasing terror) can be observed specifically for transnational terrorism (Testas 2004; Kavanagh 2011; Kruegar & Maleckova 2003; Barrebi 2007). The ameliorating effect of education on violence has been called into question in the context of terrorism. This less sanguine view was canvassed by Krueger & Malečková (2003) where they argue little reason for optimism that (an) increase in educational attainment will lead to a meaningful reduction in terrorism (Krueger & Maleckova 2003: 142). This seminal work paved the way for others combining qualitative and quantitative analysis have pointed to similar results (Barrebi 2007; Shafiq & Sinno 2010; Brockhoff et al. 2015; Shafiq & Ross 2010). When considering education and terrorism, economic arguments maintain precedence. Macroeconomic decline seems to have some association with a higher incidence of terror, though terrorists are not necessarily drawn from poor or uneducated groups (Kavanagh 2011). Kavanagh (2011) argues that poverty has a conditioning effect on terrorist group participation but increases participation in terrorism only among individuals with high levels of education (Kavanagh 2011: 107). Terrorist group selection preferences and the lower opportunity costs of economically depressed environments are also considered important in explaining increases in terrorist group participation (Kavanagh 2011: 107). Under economically depressed environments education is seen to facilitate terrorism engagement by increasing individual competencies that make recognition of injustices and discrimination easier. Increased reasoning and awareness also results in more educated individuals being more astute in seeking means to overcome injustices 15

(Barrebi 2007). Terrorist organisations can in some cases offer wages and other incentives closer to individual human capital endowments and aspirations, above those provided by traditional labour markets (Brockhoff et al. 2015: 1205). These incentive structures and the ability to better match individual aspirations for influence and remuneration support the recruitment of more highly educated individuals to terrorist groups. This is a finding supported by findings of the number of more highly educated individuals amongst some terrorist groups (Barrebi 2007). Others go further to suggest that higher education levels have given rise to increases in levels of transnational terrorism in specifically the Muslim World. They argue that in such cases education inequality intersects with longstanding feelings of indignity and frustration (Testas 2004: 253). Emerging Scholarship on the Education-Terrorism Nexus Select recent studies have proposed further nuance in the relationship between education and terror. Such studies suggest that whether education reduces or fuels terror depends on sets of moderating and country specific conditions. This new wave of research proposes often nonlinear relationships between the determinants of terror and the analysis of interactions between them (Krieger & Meierreiks, 2010: 147). They argue that more conditional explanations are likely to be more useful to overcome the variance in current empirical findings (Brockhoff et al. 2015; Azfal 2013; Shafiq & Sino 2010; Kavanagh 2011). Brockhoff et al. (2015) address lacking consistency in findings on education and terror in their study of over 133 countries. Conducting a cluster analysis, they distinguish between what they call the favourable and unfavourable circumstances of states. In it they argue that these circumstances moderate the dynamics of [the] education-terrorism nexus (Brockhoff et al. 2015: 1187). Favourable circumstances are where social, demographic and politico-institutional barriers in countries are low, including those governing employment, redistribution and political participation ; thus, the individual experience of social, demographic and politico-institutional factors is positive. Under favourable circumstances socio-political grievances that might aggravate terrorist recruitment and mobilisation are removed and education ought to contribute to (individual and social) progress (Lipset 1958; Barro 1999; Temple 1999; Cohen & Soto 2007; in Brockhoff et al. 2015: 1193). Favourable circumstances mean that education boosts political participation and is expected to reduce terror by lowering the costs of achieving political goals (see also Li 2005; Brockhoff et al. 2015: 1193). 16

In states with unfavourable circumstances the effect of raising education is reversed, and it negatively interacts with poor political, social and economic conditions to facilitate the recruitment and mobilisation of individuals for terrorist causes. Unfavourable circumstances include poor economic conditions such as low economic growth, unemployment and slack labour markets. Unfavourable conditions also include institutional constraints, such as corruption, and constrained political participation resulting from poor democratic institutions. Raising education levels under unfavourable (state) circumstances is argued to enhances the development of grievance and disenfranchisement and risks increasng terrorist violence (Brockhoff et al. 2015: 1205). 2.3 The Research Gap An interesting puzzle has therefore emerged between findings suggesting education s pacifying and stabilizing effect on political violence and more recent findings pointing to higher levels of education among perpetrators of terrorism (Østbly & Urdal 2010: 1-19). Despite its being a substantially less common phenomenon the focus of governments, international organisations and media has remained on international or transnational terrorism. Domestic terrorism and its determinants is however overlooked with substantial risk. Burgeoning research in the field of recent years has also begun to suggest the need for further investigation into radicalisation of youth with high access to education. It has proposed a potential role for education in radicalisation when expected returns do not align with corresponding opportunities (Ghanem 2017). Despite this gap, a causal pathway has not yet been extensively developed or qualitatively analysed that considers the effects of levels of education on specifically domestic terror. Further, no causal pathway has been presented that builds on the recommendations of emerging scholarship into education and terror that seeks to focus on dynamics of the relationship within the population of states with unfavourable circumstances and the extent to which the availability of economic, political and social opportunities may affect the relationships dynamics. Nor has anyone (to the authors knowledge) considered the specific effect of corruption on the relationship between education and terror (Brockhoff et al. 2015). I thus intent this study to build on emerging scholarship on education and terror and seek to contribute to existing gaps in four ways. First, I focus on the qualitative investigation of domestic 17

terrorism rather than its international or transnational counterpart. Second, I consider the extent to which the relationship between education and domestic terrorism is affected under the scope condition of corruption. Third by extending on previous large-n studies, I examine qualitatively the four cases of Morocco, Libya, Jordan and Egypt. Each are cases from the Middle East and North Africa region and examples of states presenting with unfavourable domestic conditions. Finally, I seek to contribute theoretically to gaps by considering explicitly a causal pathway considered under the condition of corruption. I investigate whether the presence of corruption plays a notable role in both framing and exacerbating grievances and frustrations from unmet expectations of education amongst increasingly educated populations. 18

3. THEORETICAL FRAMEWORK In this section I first provide definitions of core concepts and outline the theoretical framework of this study. I then present the hypothesis to be tested and provide detailed explanation of the proposed causal mechanism. This section seeks to explain the way in which this study will answer the research question under what conditions do higher levels of education lead to higher levels of domestic terrorism? 3.1 Definitions The definitions below are provided for clarity of the variables and concepts used in this study. I have selected definitions in alignment with current research into education and terrorism to allow for cumulative contribution to the discipline and to strengthen the quality of analysis and access to data. Education Education is a process of receiving systematic instruction, often in institutional settings such as kindergartens, schools or universities. I use a comprehensive definition of education to strengthen measurement and overcome criticism of previous research based on limited measures of education. Level of education is measured by considering enrolment ratios at the primary, secondary and tertiary level as well as average years of schooling and adult and youth literacy rates. The use of three distinct measures is purposeful to gain a greater overall indication of the national levels of education. Enrolment ratios at the primary, secondary and tertiary level re considered. However, levels of secondary education are the most useful and important in this study given that levels of primary education are more equal across countries, and tertiary education has been found as less significant in earlier studies on education and terror. (Kavanagh 2011; Richardson 2011). Levels of secondary education are the basis on which overall levels of education are determined, observation of necessary variance between cases is determined, and are used to comparatively rank cases. Terrorism Terrorism is defined differently in existing research, with differences subject to the purpose and use of research. No one definition is accepted across all fields and purposes. In this study, I define terrorism as the premeditated use or threat to use violence by individuals or subnational groups against non-combatants to obtain a political objective through the intimidation of a large 19

audience beyond that of the immediate victims (Enders et al 2011: 321). This definition conforms with the data set from which empirics are drawn. Under this definition, perpetrators are individuals or groups. State terrorism is not included. Notable, is that terrorism, under this definition requires a political or social goal and involves attacks against civilians, officials or noncombatants. Domestic Terrorism Domestic terrorism is solely considered in this study. Domestic terrorism is defined as an incident where the venue, target and perpetrators are all from the same country (Enders et al. 2011: 321). Domestic terrorism involves only one country and excludes cases where terrorists attack international targets (Brockhoff et al. 2015: 1211). A terrorist attack is domestic when terrorists operate within their own country (Enders et al. 2011: 320). Again, this definition is selected to conform with the data set used and builds upon data of the Global Terrorism Database (GTD) which allows this study to build on an existing empirical base and effectively consider the empirical data available. Corruption Corruption is the use and misuse of power and influence, and the abuse of entrusted power for personal gain. The definition of corruption used conforms with common definitions and measures used by Transparency International (TI), which facilitates access to and comparability of empirics. Corruption can be distinguished into levels including grand, petty and political (TI 2016). The definition of corruption used is conceived as including nepotism. Nepotism is defined as a form of favouritism based on acquaintances and familiar or other relationships whereby someone in an official position exploits his or her power and authority to provide a job or favour to a family member or friend, even though he or she may not be qualified or deserving (TI 2016). Nepotism can fall into any of the three levels of corruption. 3.2 Theory, Argument and Causal Explanation Previous theoretical explanations of the determinants of terrorism form important theoretical starting points and the conceptual basis of my investigation. I use theoretical arguments explaining the generation of grievances and frustration building on economic, political and social determinants to explain the causal pathway linking education and terror and the role of corruption in the causal pathway to violence. 20

The Population of Cases - States with Unfavourable Circumstances At noted, emerging scholarship into education and terrorism proposes the distinction of states with favourable and unfavourable national circumstances that may condition the impact of education on terror. Considering circumstances at the national level is important in the investigation of determinants of domestic terrorism given that domestic events are most likely linked to domestic circumstances. National circumstances frame and affect the daily life of citizens and are therefore likely to play the most immediate role, if any, in generating grievance or frustration. Unfavourable infers a negative effect of conditions on the lives and opportunities of citizens. Unfavourable conditions are thus considered to have a positive effect on terrorism, to increase its likelihood, by increasing the likelihood of grievances and frustrations. Unfavourable circumstances at the national level include poverty, inequality, closed political systems and poor social mobility. These circumstances create the frame within which individuals will consider engagement in terrorist violence, and where levels of education may impact the recruitment, selection and interest of individuals in terrorism by affecting their opportunity cost calculations for violence (Freytag et al. 2011). To build upon and contribute to previous work on the determinants of domestic terror, this study focusses its investigation on and selects cases from within the population of states exhibiting unfavourable circumstances. This case population is purposeful as the proposed positive relationship between education and terrorism is argued as relevant specifically to states meeting three basic and unfavourable national circumstances. The first is economic and builds on the argument that poor economic conditions can directly affect individuals levels of deprivation, [and] feelings of injustice that can lead to violence (Chen 2003; Paxon 2002; Lie & Shaub 2004; in Burgoon 2006: 126). While empirics are contested, I argue that unfavourable economic conditions will affect individual attitudes and therefore patterns of domestic terrorism by framing experiences of poverty and inequality. The second is political and builds on findings that weak political and civil rights, government instability, high levels of repression and weak political institutions are collectively and consistently associated with higher levels of terrorist production and are a necessary condition under which the positive relationship between education and domestic terrorism is supposed to hold (Krueger & Malečková 2003; Krueger 2007; White 1989; 1993; Lichbach 1987; Krueger & Laitin 2008; in Kavanagh 2011: 108). The third relates to socio-demographic circumstance and the social, demographic and politico-institutional barriers governing employment, redistribution and political participation. Here I build on previous research 21

pointing to the potent combination of poor economic and institutional conditions for terror and argue that poor economic circumstances and minimal opportunities to escape it combine to incite grievances and channel frustrations toward violence (Krieger & Meierricks 2010). These economic and socio-political circumstances are thus important conditions within which the proposed causal mechanism, and the positive relationship between education and domestic terrorism is argued to take place. Great Expectations, Greater Frustrations - Potential Negatives Impacts of Education I propose that corruption, a previously understudied factor, can assist in illuminating the positive relationship between increased levels of education and domestic terror. The theoretical underpinnings of this study build on the work of Dollard and Miller (1939), Gurr (1970) and Crenshaw (1981). In 1939 Dollard and Miller suggested that frustration leads men to act with aggression. Gurr (1970) built on this hypothesis by proposing a psychological approach describing how discontent can manifest into violence - the primary source of human capacity for violence appears to be the frustration-aggression mechanism the anger induced by frustration is a motivating force that disposes men to aggression, irrespective of the neutralities (Gurr 1970; in Richardson 2011: 6). Gurr argues that the gap between what is expected and what is achieved generates discontent. Relative deprivation is a term used to denote the tension that develops from a discrepancy between what ought and is ; and that disposes men to violence (Gurr 1970). Crenshaw (1981) and Ross (1993) propose a similar model of grievances for groups, arguing that grievances and frustration are born out of marginalisation, relative deprivation and inequality, with the potential to result in violence. The hypothesis of this study builds on these works. In this thesis, I argue that when education is obtained under conditions of state corruption that hinder personal development, this causes frustration, spurred by a form of relative deprivation. The resulting aggression, domestic terrorism, is directed locally, motivated by local conditions and involves local actors (Ross 1993; Gurr 1970; Crenshaw 1981). In sum, I argue that that in countries fraught by corruption, higher levels of education will lead to higher levels of domestic terrorism. 22

Corruption IV: Education + DV: Domestic Terrorism + Figure 1 The Hypothesised Relationship between Education and Domestic Terrorism Causal Mechanism Education is often seen as building the potential, capabilities and competencies of individuals. With increased competency, skills and knowledge, comes enhanced expectations of individual value and opportunities to contribute. Corruption fundamentally alters this state of play in some societies by disrupting input to outcome calculations of education, or the payoff of educational investment. By creating a system of who rather than what you know, corruption disrupts typical expectations of a return on educational investment and by doing so increases the value of non-market means to gain returns (Brockhoff et al. 2015: 1205). When raised education levels, and the expectations that accompany them, are met with the barriers of corruption, I argue that some will be left frustrated. Under such circumstances, frustration may lead to aggression, expressed in an increase in levels of domestic terror. Further, that higher expectations born out of increased education are likely to increase frustrations on the part of a select group of individuals if the promise of better circumstances remains unmet. And that recognition of this relative inequality and deprivation is likely to increase the prospects of engagement in violent conflict behaviour (Urdal 2006; Goldstone 2001). This causal mechanism and relationship between higher levels of education and domestic terror can be divided into three steps. The first accounts for the increases in expectations born of increased education and the second explains how high expectations can lead to frustration when corruption is present. The third demonstrates the hypothesised shift for some, from frustration to acts of domestic terror. This causal process is summarised in the diagram below. IV: High Education STEP 1 STEP 2 STEP 3 High expectations Frustration DV: Domestic terrorism Figure 2 Proposed Causal Chain Linking Education and Domestic Terror 23

Step 1 - The Role of Education in Raising Expectations As noted, education is widely assumed to increase individual and society-wide prospects of socio-economic participation, individual productivity and intellectual capacity (Brockhoff et al. 2011: 1181). As education levels increase, so too do expectations, at the individual and collective level about achievement of economic, political or social gains. Agency, the ability to contribute, is similarly linked to increases in life satisfaction. Studies on the value of education continue to support and propagate this view. Education thus links to productivity and the promise of greater opportunities. A common expectation born of the investment in time, effort and often money for education is that investment is repaid in benefits in standards of living, respect and employment. This oft considered chain is shown in the figure below. Causal Process Higher levels of education Increased skills & competence Higher levels of employment/ social/political participation Improved economic / social status Higher life satisfaction Figure 3 Causal Mechanism Step 1 - From Education to Expectations Step 2 - The Role of Corruption in Frustrations Corrupt and nepotistic practices mean that attainment of higher levels of education and merit as the basis for obtaining associated gains in social, political and economic opportunities cannot be confidently relied upon nor assured. Who you know rather than what you know becomes currency. Corrupt systems thus create exclusion and injustice which can be seen by some aspiring to advancement through education as unfair limitations on the ability obtain anything from political positions, employment opportunities, value recognition or participation in political and social processes. I argue that the distortion to traditional expectations and outcomes is caused by corruption that affects individual and group calculations on participation in violence. First, corruption affects the perceived fairness and equity of state systems. Second, by heightening who, rather than what, you know, corruption renders less valuable the role of education in delivering benefits in social status, employment and individual agency. Corruption therefore makes reliance on traditional pathways less dependable and heightens consideration of alternatives and therefore lowers the opportunity cost of participation in alternative, even illegal activities to achieve benefits. 24

Otherwise stated, corruption has the effect of removing the reliability of opportunities that would otherwise be delivered by increases in education. The relevance of this to individual participation in domestic terrorism is two-fold and relates to grievance and incentives. First, corruptions distortion of the system is likely to heighten grievance on the part of some more highly educated individuals resulting from the devaluation of their educational investment and limiting of opportunities. Second, the incentives presented by participation in the non-legal economy, including incentives provided by terrorist groups which might better match individual expectations for remuneration or status, become more appealing. If financial or other material and non-material incentives cannot be as reasonably achieved through traditional pathways, then opportunities elsewhere are considered. The opportunity costs calculations for more highly educated individuals under corrupt systems therefore are altered, including increases to the valuable prospects of a career in terrorism. A final consideration is that corruption is likely to increase grievances specifically directed against the state, it s institutions and elite seen as directly responsible for perverting and using the system for personal gain. Corruption, manifest in the nepotistic or structurally exclusionary operations of state and by extension often society, are thus a critical barrier to achievement of individual potential based on merit. This disruption to traditional expectations is visualised in the figure below. Disruption to the causal process caused by corruption Higher levels of education Increased skills Higher likelihood Improved & competence of employment/ economic / social participation status Higher life satisfaction Figure 4 Disruption Caused by Corruption to Traditional Expectations Step 3 - Moving from Frustration to Violence Under corrupt conditions, increases in educational attainment serve then only to raise collective and individual expectations along with the potential for these to become graver frustrations. Indeed, in the absence of agency to make good the promises of educational investment, frustrations are likely to be high and raising levels of education under corrupt systems is instead likely to generate only greater awareness of potential unrealised, stoking aggression and violence rather than decreasing it (Proctor 2015; in Burde et al. 2016). 25