BACHELOR THESIS the Political Roots of Armed Conflict and Its

BACHELOR THESIS The Political Roots of Armed Conﬂict and its Geographical Dimensions in Afghanistan and Pakistan

Christopher Moik Student number: 0917610 University of Innsbruck Department of Political Science ao.Univ.-Prof. Mag.Dr. Gerhard Mangott 29.11.2013 Eidesstattliche Erklärung

Ich erkläre hiermit an Eides Statt durch meine eigenhändige Unterschrift, dass ich die vor- liegende Arbeit selbständig verfasst und keine anderen als die angegebenen Quellen und Hilfs- mittel verwendet habe. Alle Stellen, die wörtlich oder inhaltlich den angegebenen Quellen entnommen wurden, sind als solche kenntlich gemacht.

Datum, Unterschrift Contents

1 Introduction 1

2 Theoretical Framework 2 2.1 Related Work ...... 2 2.2 Research Design, Questions and Hypotheses ...... 4

3 Methodological Background 6 3.1 Basics and Coding ...... 6 3.2 Statistical and Geographical Analysis ...... 7

4 Political, Military and Social Context 8 4.1 The Cold War and the Radicalization of Islam ...... 8 4.2 The Afghan Civil War and the Rise of the Taleban ...... 12 4.3 The Answer of the West: From Bush to Obama ...... 14 4.4 The Pivot to Counterinsurgency ...... 17 4.5 Overview of Battle-related Deaths since 1989 ...... 19

5 Armed Conﬂict in Centers and Peripheries 22 5.1 Social Structure in Pakistan ...... 22 5.1.1 General Imbalances between Rural and Urban Areas ...... 22 5.1.2 Political, Ethnic and Economic Structure ...... 24 5.2 Social Structure in Afghanistan ...... 27 5.2.1 General Imbalances between Rural and Urban Areas ...... 27 5.2.2 Ethnic and Social Structure ...... 27 5.3 Patterns of Violence in Centers and Peripheries ...... 30 5.3.1 Empirical Results of Violence on the Macro Level ...... 30 5.3.2 Empirical Results of Violence on the Micro Level ...... 32

6 The Durand Line and Armed Conﬂict 40 6.1 The Origins of the Border ...... 40 6.2 The Border and the Taleban ...... 41 6.3 Violent Events and their Distances to the Border ...... 43

7 Armed Conﬂict in the Mountainous Areas 46 7.1 The Geography of Mountains in Afghanistan and Pakistan ...... 46 7.2 Military Warfare and the Mountains of Afghanistan and Pakistan ...... 48 7.3 Violent Events on Mountains ...... 49

8 Discussion and Conclusion 51

Appendix 53

References 55

Figures 62

1 INTRODUCTION

1 Introduction

Twelve years ago the original mission to overthrow the Taleban government by the United States and its allies was quickly carried out. However, Afghanistan is still in chaos today. President Bush and his comrades got themselves into a region which had been the theater of the Great Game, the Cold War, regional power seeking, a bloody civil war and the rise of militant Islamism. The neighbouring territory Pakistan has also been playing an important role in Afghanistan in these periods, especially since the 1980s. But Pakistan did not only cause problems in Afghanistan but also had problems with India and within its own borders. Today, we know that the broader region, especially the Pashtun areas, is characterized by a huge complexity and a mixture of Islamic terrorists and militants, warlords, criminal networks and other local militias. Militant groups, such as Al-Qaida fighters, the Afghan Taleban conducted by Mullah Omar, the Haqqani network, militants around Hektmatyar or Mehsud’s Tehreek- e Taleban Pakistan (TTP), are frequently tracking through the tribal areas and engaging in armed conflicts with government authorities (Wagner 2010: 106). The primary epistemological interest of this paper is the political and geographical dynamics of physical violence in Pakistan and Afghanistan between 2008 and 2009,1 especially violence in the context of the internationalized intra-state conflict between government authorities and Taleban and other militant groups. This paper argues that a focus on geographical factors might give, in fact, interesting insights into the fundamental dynamics of violence. From a scientific perspective I try to examine the intensity of violence on three theoretically important planes and its geographical extent, particularly (1) the distance between periphery and center, (2) the international boundary between Afghanistan and Pakistan and (3) the impact of rough terrain on violence intensity. Not only battle-related deaths, but also violence against civilians and terrorist attacks should be included in the analysis. Based on the Armed Conflict Location and Event Dataset (ACLED)2 I will examine approximately five thousand cases of violent incidents and their political, military and social extent. First of all, I will provide a theoretical approach of my line of argumentation that includes a short overview on the current state of research and the specifications of my research questions and hypotheses. Subsequently, in section 3 the empirical procedure and statistical and geographical categories are described in detail. Section 4 portrays the historical context of the conflict in political and military manners. For this I will deduce the rise of Islamic terrorists and militants from historical events and political and social backgrounds. After that I will con- cretize the aggravation of the situation in the whole region, especially the counterterrorism and counterinsurgency operations since 2001. Sections 5 - 7 will be the main part of the empirical analysis, which contains the validation of the hypotheses. Finally the conclusion summarizes the essential findings of the analysis and gives a short assessment of future prospects of the violent dynamics in the region.

1Thanks to Prof. Gerhard Mangott, my supervisor and one of my intellectual paragons, for engaging my scientiﬁc interest in this topic. 2composed and published by the Trinity College Dublin, the University of Colorado and by the Centre for the Study of Civil War, a sub-research group of the International Peace Research Institute in Oslo (PRIO).

1 2 THEORETICAL FRAMEWORK

2 Theoretical Framework

The following subsections introduce the theoretical framework of my study. For this I will point out diﬀerent research papers which deal with violence in Pakistan, its geographical and social contexts and other general observations about geographical factors and wars. Thereby I will emphasize some important connecting points of the studies limiting my own research questions and approaches. Derived from these considerations are my research questions and hypothesis, which are speciﬁed in the next subsection.

2.1 Related Work

There are several researchers who focus on the geographical dimension of physical violence. For instance, the geographers Daanish Mustafa and Katherine E. Brown did not indicate military and political reactions to terrorism, in contrast to this work, but they expose specific responses in the every day life of the population based on the same dataset as my study. They raised the interesting question of how terrorist actions fill public space and in which way they depend on cultural and religious factors. As a result they point out that on the one hand public spaces are influenced by ”the predominant syncretic Sufi tradition entailed public carnivals, parades and procession” at different places and, on the other hand, affected by terrorist violence that is ”linked to their project to seek a ’perfected humanity’ and to the performance of ’civilian life’ ” (2010a: 1-22). The same authors in a different study analyze terrorist incidents and the correlations with them and public space and economic and social indicators. They argue that efforts ”to neutralize TTP control depend in no small measure on the population’s understanding of the limited vision of life the Taleban offers and, upon that recognition, their ability to freely communicate these views without fear” (Daanish, Brown 2010b: 496-512). Consequently, they propose an extensive strategy to fight terrorism and counterinsurgency incorporating political and cultural communication in a public space protected by security forces. A study by O’Loughlin, Witmer and Linke, based on the same dataset, provides a human geographical analysis of the locations of violence, temporal dimensions and other general observations. In summary, the most important rationales of the comprehensive study are that (1) violence is still concentrated in the mountainous areas and in the southern part of Afghanistan, which can be assigned to the Pashtun territory (71 % of the incidents); (2) that the number of violent attacks located in urban areas increases, such as Kabul and Peshawar, and (3) that between the years 2008 and 2010 the conflict spread over the Tribal Areas of Pakistan (2010: 437-471). In addition to the substantial findings, the study contains empirically relevant tools to observe the geography of violence that could also be applied in this work. Another trial of Pakistani violence and its social background is presented by the economist Sadia Mariam Malik (2011). Based on data from the Pakistan Institute of Peace Studies in Islamabad she examined statistical correlations between violent incidents, food insecurity, the number of Madrasas, employment rate, literacy rate and so on. She had found out that landlessness and food insecurity positively correlate with violent attacks. But this result does not mean that it is the main reason for violent conflict although poverty can create a spiral of violence. It is possible that landlessness and food insecurity are effects and not the primary motives of the conflict. All that these investigations into social and economic indicators have in common is that they might explain psychological dispositions and aggressions, but they do not

2 2 THEORETICAL FRAMEWORK describe the basic process of recruitment and violence. But an interesting outcome of this topic, contrary to general assumptions and studies, is that Malik did not find a significant coherence between violence and the number of Madrassas. These results supports the result of a study by Christine Fair (2008), who tried to answer the question under which conditions recruitment of militants occurs. She figured out that a ”singular focus upon madaris as recruiting grounds for militants is inadequate” and that Madrassas ”may be more important for other kinds of militants such as sectarian terrorists and suicide attackers”. She refers rather to factors such as ”mosques, tabligh, friends and family” and suggests an intensified focus on the household’s decision-making process to explain militant recruitment. However, these findings do not necessarily mean that Madrassas have no impact on violent activities at large, because some studies show that in periods of military crisis and when facing a military collapse the Taleban recruited militants through Madrasas, particularly in the 1980s (Abuza 2003: 11) and 1990s (Noor 2008: 141-168). Generally, the process of recruitment is a far more complex issue, as Abuza stated, and is principally based on networks linked with ”kinship, mosque, madrassa, and friendship” (Abuza 2008: 63-90). A different work on civil wars and their links to geographical factors is provided by Buhaug, Gates and Lujala (2009: 544-569). Examining 228 civil wars they include variables such as distance between center and periphery, rough terrain, international borders, the capability of the army and natural resources. Their results show that (1) ”conflicts that occur in the periphery are much less likely to be resolved within a short period” than conflicts close to the goverment’s homebase (capital), (2) that national borders have a significant impact on conflict duration, because they are used as safe havens and that (3) natural resources actually have no influence on the duration of the conflict. So, the longer the distance between the center of the government and the conflict zone the longer the conflict duration. The reason for this is that transportation costs and logistical problems increase and the ”knowledge of the local environment” and the ”support from the local population” decrease. Additionally, they understand that the impact of rough terrain on conflict duration is not significant. However, these findings cannot expound the role of rough terrain as such. Another comprehensive study by Fearon and Laitin (1999) tries to clarify the role of minority groups combined with its size relative to the majority population, rough terrain and social indicators. Thereby the term rough terrain is characterized by inaccessibility, rural and local qualities and its function as a geographical base for insurgents. They concluded that ”urban and widely dispersed minority groups are far less likely to see large-scale violence than [...] groups [are] that have some sort of rural base”. Additionally, the findings show evidence that ”group size [. . . ], country size, the presence of ethnic brethren who dominant a neighboring state, relative ethnic homogeneity, and rough terrain are associated with higher probabilities of rebellion” (1999: 21). According to other conclusions of their paper, the empirical results of the variable ”bad terrain” assume that it increases the efficiency of counterinsurgents. But this does not lead to higher success rates, but rather benefits an upwardly violent spiral on both sides. Including the factor ”minorities who live in rough terrain” the variable further increases the likelihood of guerilla war (1999: 27).

3 2 THEORETICAL FRAMEWORK

2.2 Research Design, Questions and Hypotheses

One of the main issues will definitely be the intensity of violence, its geographical implications and political, military and social causes - which is actually quite a sensitive realm. Thus, on the one hand, intensity of violence can be quantitatively measured by the number of violent events and, on the other hand, by counting the event-related deaths. To explain such a multifactorial phenomenon as violence I argue that a broad concept of explaining violence in Afghanistan and Pakistan must incorporate geographical factors. Hypothesis 1 deals with distances between conflict zones and political and military centers and tries to figure out the impact of centers and peripheries on violent intensity. The second assumption is concerned with the influence of the national border between Afghanistan and Pakistan on the intensity of violence and its political, military and social implications. As a last geographical factor that might affect the violent intensity, the variable rough terrain is included in the analysis. Supplementary to the impact of these three geographical factors on violence it is estimated that they are linked among themselves in some degree, such as the national border and the altitude.

Figure 1: Theoretical model

In the following paragraphs the research questions, their corresponding hypothesis and feasible theoretical explanations are speciﬁed. In doing so I will create theoretical hypothesis based on several assumptions that should be veriﬁed later in the empirical analysis.

Hypothesis 1. As previously mentioned, Buhaug, Gates and Lujalathey verified that long distances between the conflict zone and the center of the state make it difficult for the government to strike effectively against insurgents. In this paper, however, I do not measure the influence of distance on conflict duration or military effectiveness, but rather the effects on violence intensity and the number of violent events. Additionally, this hypothesis describes several political bases on a macro level. A common theoretical definition of a political base does not exist. In the majority of cases political bases represent larger cities - in a political sense, centers of authority, governmental institutions and the monopoly on the use of structural and material force. However, in Pakistan and Afghanistan different actors bargain and compete for territorial control, public spaces and, of course, for social support. So, I will also examine local centers in the periphery. So, the first question is whether lower or wider proximities between bases and conflict zone have a significant impact on violence intensity. What are the causes and effects of violence in and around cities and in peripheries on a political, military and social level? What are the typical characteristics of violent conflicts in the peripheries and centers? There are good reasons to believe in this hypothesis as well as in the diametrically opposite outcome. Shorter proximities may require more heavily armed troops and a higher effectiveness.

4 2 THEORETICAL FRAMEWORK

Terrorist attacks in public space might increase the violent intensity and public attention. Furthermore, political and military headquarters may increase the incentives for rebels to attack them because of imprisoned insurgents, arms depots, cars and so forth. It would definitely be an interesting question whether a high presence of security forces can guarantee stability or instead cause violent spirals in cities or peripheries. Alternatively, lower proximities between political bases and the conflict zone might decrease the violent intensity, because of the secure locations of governmental forces, a higher military custody or further difficulties for non-governmental militants. It might be an important question to ask whether cities or political centers attract violence. Terrorist groups can cause a lot of human damage and use cities as self-representation to leave the invisible space and to communicate in public. Moreover, a concentrated settlement as a target of violence might cause higher death rates or higher material damage. Another reason for a higher intensity of violence in cities might be that they function as political, administrative, intellectual, religious or economic centers of the surrounding periphery and as intersections of different networks. Theoretically speaking, this imbalance of power can cause a specific dynamic of counterbalancing. But there is not only an imbalance of power between the center and the periphery, but also imbalances among local centers and their peripheries themselves. Because of the ethnic constellations, specific Pakistani cities, such as Karachi, and provinces play an important role in the political devolution processes. Political, intellectual, religious, economic or public institutions can be seen as a kind of power projection, which manifests physical and structural hard power as well as soft power.

H1: The shorter the distance between the conﬂict zone and the political and local centers, the higher the intensity of violence.

Hypothesis 2. My second hypothesis tries to figure out which role the national boarder in the context of violent intensity plays. According to several analysts, the national border between Pakistan and Afghanistan functions as an entrance to a safe haven for Afghan Taleban and other non-governmental armed networks. This leads to the conjectures that violent events are notably located at the border region and that the proximity to the national border, as may be the case, also causes higher death rates. So, the question is what impact the variable border on violent intensity has and what the causes and effects of violence are? To prove this issue quantitatively I will first provide a simple frequency analysis and will then measure a correlation function that comprises the distances between incident locations and the national border as the independent variable and the violent intensity as the dependent variable. There are several reasons, in addition to the means of safe haven, why national bound- aries matter. The border between Afghanistan and Pakistan crosses a culturally and socio- economically homogeneous ethnic group, in particular the Pashtuns, and have also provoked political disputes between the Afghan and Pakistani governments throughout history, confined national sovereignty and still plays an important role in smuggling drugs, weapons and other commodities. Furthermore, the national border defines political authority and perhaps solidarity. Rivalry or solidarity between Afghan Taleban and Pakistani Taleban is possibly rooted in the issue concerning the national border. These factors must be considered in the empirical analysis. To formalize my assumptions one hypothesis should be postulated:

H2: The closer one gets to the national border between Afghanistan and Pakistan, the higher the intensity of violence.

5 3 METHODOLOGICAL BACKGROUND

Hypothesis 3. In the third hypothesis I will prepare another potential factor to violent intensity, in particular the impact of rough terrain. There is good reason to assume that rough terrain, which is spread over wide areas of Pakistan and Afghanistan, influence the occurrence of combats or the violent intensity. Especially the mountainous terrain is suspected of affecting the conflict and violent intensity. Although Clausewitz’ theoretical analysis of strategic warfare is outdated to some extent and often confined to inter-state wars, he postulated that rough terrain, such as mountains, marshes or forests, favors partisan warfare (Lowther 2007: 37). As mentioned above, there is also some empirical evidence for this assertion. Similarly, I will find out whether violence caused by asymmetric conflict in Afghanistan and Pakistan appears in mountainous areas more often. Apart from beneficial defensive positions for insurgents and appropriate military responses, for instance guerilla warfare or the use of drones, mountains also have something to do with critical infrastructure, such as mountain passes and streets, power stations, military bases or locations concerning telecommunication and logistics. Thus the analysis has to focus on these aspects considering the different impacts on the intensity of violence. To answer this question the height above see level of violent events should be appointed as independent variable and that of violence intensity as dependent variable. But this time the graph is expected to be a Gaussian distribution, because at a specific point high terrain makes it impossible to combat.

H3: The higher the height above sea level the higher the intensity of violence, but at a speciﬁc point of the sea level violence decreases.

3 Methodological Background

In the following sub-sections I will elaborate on the methodological procedures and their backgrounds that are used in the empirical analysis later. First of all, I will elucidate the dataset and its codings and where to ﬁnd the original documents. It is an important to understand what is meant by the given data. To process the data I will use two software programs, in particular the statistical program SPSS and for the geographical tools the open source software QGIS. In general every QGIS map consists of diﬀerent layers that represent certain features of the geographical conditions.3

3.1 Basics and Coding

Unlike other datasets ACLED does not provide information on a macro level, for instance deaths per year, but rather reports single events of a speciﬁc conﬂict. The dataset provides a broad range of information, such as involved actors, event type, location, date, fatalities, the source of information and several other notes (drone strikes etc.). To admit a violent event into the dataset a minimum death rate is not required. The analysis concentrates mainly on the event types ”battles” between military organized groups and ”violence against civilians”. The dataset originally coded a single event that caused batte-related deaths as well as dead civilians as two events, to be precise a ”battle” and ”violence against civilians”. So, the dataset includes a total

3The layers of the administrative units, the elevation and relief are downloaded from http://www.diva-gis.org/gdata, some layers are provided by http://www.naturalearthdata.com/ and http://www.gadm.org/country

6 3 METHODOLOGICAL BACKGROUND number of 4,945 violent events between 2008 and 2009. For this paper I recoded such cases as one event but indicate the number of deaths in a separate column. One problem concerning the data is that some indications of the involved political actors are quoted as ”unidentified armed groups”, which is obviously the result of the complex conflict situation. To cope with this issue I will further elaborate on the conflict actors and its relationships. Nonetheless the dataset specifically indicates different political actors and distinguishes between governmental actors (international, national and local actors), non-governmental armed groups (Taleban, TTP, Al-Qaida etc.) and civilians. But the most interesting part of the dataset for this study is the information about the location which does not only provide the administrative units but also detailed geographical coordinates. For this they used the software GIS, georeferenced GIS layers and additional information from other sources about locations. If a source only reports information about the region of a violent event, the coordinates of the closest town or a georeferenced town associated with that region are used. One of the key problems is that the given numbers of fatalities are usually not accurate and can be erroneous. Some numbers are estimated or differ from other information sources (if so, the lowest number of deaths is listed) or are in some cases rounded. Although not all numbers are exact, it can be argued that they can display trends and tendencies. Local, national and international media, upgraded with NGO reports were used as sources. It must be assumed that not all information is made completely public, but, here too, it can show trends. For further information the data, as well as the codebook, can be found on the homepage of ACLED (www.acleddata.com).

3.2 Statistical and Geographical Analysis

In contrast to qualitative studies the violent intensity will be quantitatively measured, on the one hand, by the number of deaths including battle-related deaths as well as dead civilians and, on the other hand, by the number of violent events. The reason for this dual use (death rate and number of violent events) is that the precise number of the death rate of every single event is not invariably given. So, if some cases do not detail the exact number of deaths, those cases are excluded in the statistical analysis of the average death rates. By contrast, the measure of the frequency of events includes those violent events without the detailed number indeed. To test my hypothesis I modify the dataset by using the geographical software QGIS and add some further information. This means, for instance, that the information about the spatial coordinates make it possible to generate data about the distances between violent events and the national border or the figures regarding the altitude of every single event. This information should be incorporated into the dataset. To test my hypothesis I prefer, on the one hand, descriptive statistics to deal with the number of violent events and, on the other hand, functions combined with a correlation analysis. In other words, the highest intensity will then be expected if the height above sea level increases and at the same time the distances between border and events and between bases and events decrease. The first hypothesis clarifies the connections between political and military centers and conflict zones, its accessibilities and political, military and social interdependencies. To test this question, at first the analysis identifies the location of violent hotspots and the centers of the political authority. To quantify the impact of centers on violence intensity I will

7 4 POLITICAL, MILITARY AND SOCIAL CONTEXT build distance circles with a radius of 100 km around bases and will measure how many events occur within decimal steps. Additionally, the correlation between distance and the average number of deaths are generated. This shows us whether the violent intensity increases or decreases with a lower distance to political and military centers. The inﬂuence of the variable national border on the violent intensity requires an evaluation of the absolute distances between the violent events and the border between Afghanistan and Pakistan. For this purpose the QGIS software aﬀords tools to reach this intention. To test the hypothesis of the third hypothesis, QGIS tools enable me to detect the height above sea level values of every single violent event. The impact of the variable elevation on the violent intensity is also illustrated by frequencies and correlations.

4 Political, Military and Social Context

According to Nasreen Akhtar (2008: 49-73) war in Afghanistan and Pakistan’s involvement are historically constituted of three periods. The first period started with the communist coup in 1978 and the internationalized civil war led by the Soviets in 1979 backing the Afghan government against the rebellion and ended with the Geneva accords in 1988 and the withdrawal of the Red Army (1988-1989). The second phase depicts the civil war between different Afghan groups and the rise of the Taleban government until its fall by the United States and their allies in 2001. The third phase describes the internationalized civil war between the governmental actor Afghanistan as well its international allied conflict parties and non-governmental conflict parties, specifically the Taleban. Simultaneously the United States and some of their allies have been operating against the transnational terrorist network Al-Qaeda with some support from Pakistan and Afghanistan.

4.1 The Cold War and the Radicalization of Islam

Afghanistan and Pakistan have suﬀered from large scale violence for several decades. Known as The Great Game Afghanistan was often considered as the theater of rivalry between Great Britain (GB) and czarist Russia. After Afghanistan had become fully independent from GB in 1919, the Afghan monarchy persisted until 1973 when the previously royalist prime minister Daoud Khan overthrew king Mohammed Sahir Schah. But under the king Daoud had to resign as prime minister because he was one of the proponents of the integration of the Pakistani Pashtun areas into Afghan territory. This rhetoric resulted in the closure of the border by the Pakistani government and a harsh rivalry between the two states, which inevitably led to the political rapprochement between Afghanistan and the Soviet Union. When Daoud staged a coup in 1973, he was supported by the leftist parties, which he later thrusted aside to bring the Afghan government closer to the West. But in 1978, after several communists had demonstrated in Kabul and had been arrested by Daoud and the assassination of Mir Akbar Khyber, members of the military and the government, which were partially controlled by the Soviets, were able to overthrow Daoud and established a communist government,4, known as the Saur Revolution which was actually a revolution from above (Goodson 2001: 52, 53). Because of social reforms,

4There is no evidence that die Soviets initiated this coup, although, they probably knew the plan and endorsed it.

8 4 POLITICAL, MILITARY AND SOCIAL CONTEXT an economic decline and the repressive approach to potential counter-revolutionaries, an armed resistance and several revolts emerged in the rural areas in 1979, so that the Soviets had to intervene after they had signed a cooperation treaty with Afghanistan in December 1978, backing Taraki’s regime with military support if needed. From the perspective of the Soviet foreign policy there were many reasons for intervening: First of all, Moscow feared that the revolts against a communist government might affect Soviet Muslims in their southern republics in Central Asia and that they had to bring the rebellion down as fast as they could. But the Soviets were generally worried about the rise of political Islam as had happened in Iran, and in particular the Islamic revolution in 1979. Moreover, they intended to control and strengthen the Afghan government, because of Pakistan’s and Iran’s adversarial attitude towards the Soviet Union (SU). Thereby the SU also aimed to undermine anti-communist and pro US Pakistan by supporting the disintegration of the Pashtun people and Balochis, whose weakness towards Pakistan could have brought them closer to the Soviets. In addition to this, the SU was worried about China’s expansion in Asia, its relations with the United States and its feasible encircling of the Soviets on its eastern borders. Last but not least the intervention must be seen in the context of Soviet-American rivalry in the Middle East and the US balancing strategy through Egypt, Israel, Saudi Arabia, Pakistan, China and the Arabian Sea. But the issue of the US-SU rivalry was far more complex. The Soviets assumed that the Carter administration was not capable of engaging in the Soviet intervention in Afghanistan and, concerning Soviet’s introspection, they were more capable as they had been ten years earlier (Goldman 1984: 384-403). Unlike the Soviet’s expectations, Pakistan, the United States and its alliances used the Soviet intervention as an opportunity to accelerate their own strategic interests when they decided to respond by supporting the rebellion. Overnight Pakistan became the center for the counter-intervention and provided financial and logistical support as well as retreat areas for the rebels. In general, Pakistan’s involvement is attributed to domestic and international considerations: Domestically, Zia ul-Haq, then head of the military regime, was quite aware that his country lacked national unity that could have been constrained through an interpolation in Afghanistan. Zia promoted a policy of Islamization, for instance by increasing the number of Islamic schools from 900 Madrassas in 1971 to 8,000 official and approximately 25,000 unofficial Madrassas in 1988 (Rashid 2011: 146) or by establishing an Islamic law court (Shuja 2007: 25-35). Thereby he reinforced the role of Islam in society and in the army in order to create greater stability and homogenization within Pakistan and to get the support from religious parties and groups for his own political survival (Haqqani 2005: 144). Internationally, Zia’s regime was isolated with respect to the rivalry between Pakistan and pro-Soviet India that had gained good ties with Afghanistan to surround Pakistan and hoped for international support (Akhtar 2008: 49-73). As a consequence Pakistan pursued to undermine the influence of India in Afghanistan by installing a Pakistan friendly regime. The USA and Saudi-Arabia provided money for the rebels and Pakistan, especially its Inter-Services Intelligence (ISI), in conjunction with the CIA instructors helped to set up the Mujaheddin and trained them (Alavi 2002: 24- 31). More precisely the CIA financially obtained the weapons which were, as a consequence, distributed to the rebel’s factions by the ISI (Runion 2007: 112). At the beginning of the Soviet invasion the revolts were initially conducted by different groups, such as ”local tribal khans, traditional ulema, Sufi pirs, and army deserters like Captain Khan” and ”not the Islamist parties

9 4 POLITICAL, MILITARY AND SOCIAL CONTEXT of Gulbuddin Hekmatyar and Burhanuddin Rabbani” (Tomsen 2011: 136). In Peshawar seven different Mujaheddin parties were located. Around the city Kandahar the Durrani network, mainly located in the west of Afghanistan, represented the Mujaheddin (Rashid 2011: 39), who were far less supported by the ISI and CIA than the Wahhabite Arab-Afghans and Islamic Pashtuns in the eastern parts. Another actor whose political origins are ascribed to the Soviet invasion was Osama bin Laden, who went to Afghanistan in 1984, then encouraged the Mujaheddin with economic, cultural and social capital and transformed the Mujaheddin Services Bureau into an international network in order to recruit well educated terrorist Islamists (Bodansky 2001: 12). Bin Laden also established the so-called Islamic Salvation Council to collect financial aids for the Services Buraeu of the Mujaheddin and built and enlarged different bases across Afghanistan and Pakistan.5 One of the most important beneficiaries was the Ghilzai Pashtuns under Hek- matyar, who also got massive support from Abdallah Yusuf Azzam (Tomsen 2011: 249, 250, 251), the father of the Islamic and Afghan Jihad and mentor of Bin Laden. Besides, Azzam was highly influential in exporting the religous views of the Muslim Brotherhood and in recruiting voluntary Arab militants against the Soviet occupation. But those names are just a few among many influential persons and institutions, as the chart of Tomsen shows:

Figure 2: Network of the ISI and the Arab groups (Source: Tomsen 2011, p. 250)

The radicalization of Islam in this region in the 1980s was definitely not only the product of international support, but also an outcome of a political and religious competition on a national and local level. On the local level the conflict in Afghanistan gradually tended to shift the power from traditional tribal leaders (known as Maliks) to Islamic proponents and religious leaders (Mullahs), especially in the Pakistani province FATA (Federally Administered Tribal Areas), where refugee camps and the bases of resistance became more and more influental

5Examples of bases are the headquarters in the Khurram Tribal Agency, the bases of Haqqani in North Waziristan, Hekmatyar’s headquarters in Khost, Sayyaf’s base in Paktiya and many others.

10 4 POLITICAL, MILITARY AND SOCIAL CONTEXT

(Wagner 2010: 105). Ideologically Islamic fundamentalism at this time was highly shaped by interpretations, such as Deobandi, Wahhabi and Salafist views respectively, which were notably spread over the refugee camps and religious institutions in Pakistan. Those religious schools can be aggregated to a radical and conservative Sunni Islam, that equals state and religion (Weatherbee 2009: 171). Yet there was little social support from Pashtun people, because of their moderate Sufi tradition, their social code, called Pashtunwali, and their self-awareness as partially autonomous groups in the Pashtun belt. Even the Mujaheddin often saw the Arabs as foreign intruders and as a burden (Bearden 2001: 17-30). The struggle for power in the tribal areas proceeded on the ground of local checks and balances between the tribal leaders, who have enjoyed some political autonomy. In this first phase, the movement against the government, once a secular group mainly consisting of army deserters fighting for freedom, then became a fundamentalist radical Islamic movement. But the radicalization of the Islam by the Mujaheddin in the 1980s was dominated by Afghan-Arabs and Islamic Pashtuns, especially Wahhabis, whose power later would be replaced by the far more Deobandi-minded Taleban. Whereas 120,000 troops of the Red Army were deployed in 1985, usually dispersed across the country, the Mujaheddin parties were composed of estimated 150,000 to 250,000 fighters. Until 1986 the Soviet air strategy, primarly using MI-24D helicopters, performed very well to some degree, but when the US Congress decided to supply Stinger anti-aircraft weapons, the power shifted gradually to the favor of the Mujaheddin (Bearden 2001: 17-30). Gorbachev ended Brezhnev’s move of and withdrew the Red Army, which had lost 14,500 combatants, 20,000 troops through other exposures as well as having suffered 50,000 wounded between 1979 and 1989 (Young, Kent 2013: 427). As a result, the radical Jihad and its self-proclaimed victory over the Soviets became more and more popular and the civil war has gone on. In total 120,000 Afghan and Russian combatants died (Correlates of War) and by 1991 six million people had had to flee to other countries (Ruiz 2005: 2). Just as many reasons can be quoted for the Soviet intervention as for their failure to defeat the resistance. On a social level the Soviet strategists and theorists, always thinking about a hierarchical society from a Marxist perspective, were not able to deal with the horizontal fragmentation and the non-industrialized structure of the Afghan society. Furthermore, on a military level the conventional forces of the SU were not at all able to deal with the guerilla tactics in the Afghan mountains. The political scientist Ivan Arreguín-Toft states that, interestingly, the lack of a hierarchical structure of the Mujaheddin was a disadvantage for the Soviets trying to bring the Mujaheddin parties down (2005: 197). Another reason for the effectiveness of the Mujaheddin was the international aid to the insurgents and the asymmetric balancing by Pakistan, Arab states, the United States and its alliances. As a consequence of these backings the Mujaheddin fighters were well equipped and were able to cope with the Red Army militarily. Other important reasons for the failure of the Soviet Union were the effects of the Afghan war on domestic aspects and vice versa. The war affected the belief in the Red Army because it showed its vulnerability. As a consequence it had only little popular support within the Soviet Union„ which led to tensions between the central state and its republics in the Caucasus region and its associated states in Eastern Europe. Besides, the SU gained no support from the Afghan population and was even faced with hostile attitudes within the Afghan army. But the war did not only desintegrate the Soviet Union from its sphere of influence, but also resulted in

11 4 POLITICAL, MILITARY AND SOCIAL CONTEXT a split among the political and military elites (Reuveny, Prakash 1999: 693-708). To sum up, the SU was highly constrained internationally as well as domestically within Afghanistan and the Soviet Union itself and was not able to create a decent strategy for dealing with this kind of warfare.

4.2 The Afghan Civil War and the Rise of the Taleban

Figure 3: Most populated cities and the international environment of Afghanistan and Pakistan

City

Disputed Territory

After the SU had left Afghanistan, the Afghan government was still dependent on Soviet aids and was able to maintain the governmental institutions for some time. But when the Russian government withdrew its support for Afghanistan in 1992, Mohammad Najibullah and his forces were quickly overthrown by the Mujaheddin - obviously the end of the SU was the end of Najibullah too. Immediately after the fall of Najibullah Afghanistan became a failed state that was highly characterized by a violent factionalism and an ongoing civil war. The northern parts of the country were dominated by the Uzbek Dostum and his militants, who had good relations with Uzbekistan and Russia and the western parts were controlled by the Tajik Ismail Khan, who collaborated with Iran and Turkmenistan. Aside from these groups, the east and the south of Afghanistan were headed by different regional warlords and Mujaheddin factions, such as Haji Qadir and Haji Shamalai, Jalaluddin Haqqani, Mavlavi Mansoor, who had close ties with Pakistan. The cities Kandahar and Kabul were also divided by different clans and networks, such as Hekmatyar’s Hizb-i-Islami, the Iran-orientated Hezb-e-Wahdat and the Rabbani government between 1992 and the Taleban’s takeover of Kabul in 1996 (Qassem 2009: 96, 97). Concerning Pakistan’s foreign and domestic policy, the situation in Afghanistan led to five general transformations during the 1990s: Firstly, in 1990 the United States implemented the

12 4 POLITICAL, MILITARY AND SOCIAL CONTEXT so-called Pressler Amendment, which abandoned the financial and military support for Pakistan because of its nuclear program and the withdrawal of the Soviets. As a consequence, heavily indebted Pakistan under Benazir Bhutto and Nawaz Sharif had to arrange eight different treaties with the IMF during the following years (Mohiuddin 2007: 120). Secondly, the drawback of the United States from Afghanistan thereby allowed Pakistan to create its own strategy towards Kabul. Tehran and Islamabad (and to some extent Riad), continued the proxy war in Afghanistan to establish a friendly government. Whereas Iran supported the Shiite, Tajik and Uzbek groups and from 1996 onwards the Northern Alliance composed of the same groups, Pakistan supported Hekmatyar in the first half of the 1990s and then the Taleban, which overthrew the Rabbani government in Kabul in 1996. Both states were anxious to install a friendly regime in their neighbouring country (Milani 2010: 156). One of the main reasons why the Pakistani regime backed the Taleban of all groups was that they formed a counterbalance to the ethnonationalistic groups in the Pashtun region. India still had strong support within the Afghan society and the Taleban were considered as the only government which can endorse Pakistani interests towards India. Thirdly, in this phase Pakistan transformed its security strategy regarding India towards nuclear parity and asymmetric balancing by sending former Afghan insurgents to the India-controlled Kashmir area, where consequently the Indian army had to intervene permanently and a spiral of violence appeared again (Dittmer 2005: 193). The crisis between India and Pakistan eventually culminated in the Kargil War in 1999, despite or even because of the nuclear tests. Fourthly, the civil war in Afghanistan also jeopardized the stability in Pakistan. Not only in Kashmir did more violence occur, but also the border region to Afghanistan became more unstable. Sectarian violence between Shiites and Sunnites increased, ethnic and political rivalry became more intensive and weapons were easier to get on the black market. Fifthly, as far as domestic policy is concerned, the PPP (Pakistan Peoples Party) of Bhutto increasingly lost support from the population, whereas religious parties gained more and more popularity, such as the party Jamiat Ulema-i-Islam (JUI), which was admitted as the coalition partner to the Bhutto government. In 1990, the PPP did not emerge victorious for the first time in democratic elections (at least in West Pakistan). Instead the fractional alliance Islami Jamhoori Ittehad (IJI), consisting of the Pakistan Muslim League, National Peoples Party and the Jamaat-e-Islami, won the parliamentary elections. In the meanwhile, the situation of the Afghan civil war between 1992 and 1994 became some kind of stalemate between the different competing groups, especially between Hekmatyar and his Tajik rival Ahmed Shah Massoud, who then became minister of defense in 1992. The Peshawar Accords appointed Rabbani as interim president in 1992, who refused to end his term in 1993 and tried to defend his position in Kabul against the different factions. At this time, Kabul became the main scene of the war and suffered 50,000 fatalities and another 150,000 were wounded (Goodson 2001: 75), while Dostum and the other groups switched back and forth between the two parties. Rabbani and Massoud managed to drag India, Russia and Iran in against the groups supported by Pakistan. Dostum was able to temporarily gain full control over Kabul in 1995, but when the Taleban attacked and took over Kandahar and Herat in March and September and Kabul in September 1996, they controlled roughly 50% of the country and four of the major cities in Afghanistan.6 Pakistan and the ISI supported one

6For a detailed analysis of the complex situation during the 1990s, see also Abdulkader Sinno (2009), Peter Tomsen (2011) or Larry P. Goodson (2001).

13 4 POLITICAL, MILITARY AND SOCIAL CONTEXT group after another to bring the Rabbani government down. In the end, the Taleban were the most powerful group to do this for Pakistan. Besides Saudi Arabia and the United Arab Emirates, Pakistan recognized them as official government of Afghanistan in 1997. At the beginning of the Taleban era in 1996 the Taleban were able to get some support from the Pashtun people, because the Taleban disarmed many warlords, militant networks and criminals and gave hope for Pashtuns to get back their autonomy that they had lost before. But the Taleban and their understanding of society decreased the popular support to a large extent and increased the international pressure by the United States, especially through some feminist groups, human rights groups and the United Nations. This led to a cut in development aids by western states as well as threats by Iran, Russia and the republics of Central Asia. General Dostum, who still controlled six provinces in the north of Afghanistan, became the last option for those states to keep the Taleban away from their borders. In 1996, the Northern Alliance (or United Front) was formed by the former rivals Ahmad Shah Massoud and his Tajik group and the Uzbek military leader Abdul Rashid Dostum and his Uzbek combatants to fight against the Taleban rule. In this situation, the alliance got support from the United States, Russia, Iran and the Central Asian states. In October 1997, Kofi Annan mounted a ”Six plus Two” meeting including the United States, Russia and the six neighbouring states of Afghanistan to convince Iran and Pakistan of an arms embargo, which unfortunately failed. This led to other consultation processes in Geneva and to the ”Committee on Afghanistan”, which primarly included Islamic states. After Al-Qaeda carried out terrorist attacks against the US embassies in Kenya and Tanzania in August 1998, Clinton and his military forces took action and bombed a training camp of Osama Bin Laden with targeted missiles in the north of Afghanistan, which was harshly condemned by the Taleban leader Mullah Omar. As far as the organizational structure of the Afghan Taleban is concerned, the Taleban leadership was not able to build a real operational government, political institutions or an effective administration. Mullah Mohammed Omar was the head of the Taleban and the armed forces, but a hierarchical chain of command did not exist. His ministers in Kabul were frequently sent as commanders to the combat frontlines as soon as one of them had gained too much influence. Moreover, the Taleban were not aware that they would need a system which could integrate the fragmented society. Instead, they thought that their multiethnic composition and their Islamic ideology would sufficiently represent the Afghan society and that an institutional system would not be required. Whereas the government and ministers respectively were located in Kabul, Mullah Omar and the proper political center were in Kandahar. Not only Pakistan’s support made the Taleban so powerful, but also their structure as a Mullah movement with Mullah Omar as their charismatic and mystic leader, who is considered to be a prophet and can mobilize his fellowship (Johnson, Mason 2007: 71-89).

4.3 The Answer of the West: From Bush to Obama

President Clinton’s Secretary of State Warren Christopher stated in 1996: ”President Clinton has rightly identiﬁed terrorism as one of the most important security challenges we face in the wake of the Cold War. [. . . ] As if the threat is not already severe enough, we now face an even more alarming danger: the terrorist armed with weapons of mass destruction.” (Christopher 2008: 130). Considering the nuclear proliferation of Pakistan and India and the unstable

14 4 POLITICAL, MILITARY AND SOCIAL CONTEXT circumstances in the region, this statement makes an appropriate point demonstrating the concerns of the United States. After the first attack on the World Trade Center (WTC) in 1993 and on the embassies in 1998 the US administration was already well aware of the dangers of terrorism, but could never estimate its own vulnerability that was revealed on 9/11. The US intervention in Afghanistan in 2001 is not as tricky to explain as the Iraq war in 2003. The Iraq intervention primarily rested upon external threats, the supremacy of the administration in security questions, the powerful impact of neoconservative politicians and think tanks within the Bush government and the Wilsonian foreign policy (Taliaferro, Lobell, Ripsman 2010: 3). Instead, the intervention in Afghanistan was a direct response to the external attacks on the WTC, which caused the so-called rally-around-the-flag effect and gave massive support from the US population to the acting politicians (Schubert, Stewart, Curran 2002: 559- 582). This event would have made any President, regardless of whether they were democratic or republican, react to the attacks on the WTC. One day after the attack the UN security council adopted the UN resolution 1368, which gave the United States and its allies permission to self-defence according to Article 51. The Bush administration argued that not only terrorists would be targeted, but also states that support them. This leads to the implication that regime changes became an important factor of Bush’s foreign policy. At the beginning of the Operation Enduring Freedom - Afghanistan (OEF-A), which started on 7 October, the Bush administration did not schedule to send their own boots on the ground, except some US special forces. Instead, the Northern Alliance was supported by the US aircraft and missile attacks and was heavily armed by the USA and UK. Only one month after the beginning of OEF-A the alliance gained control over Kabul and after another month over the Taleban’s last base Kandahar. As the fighting of the Northern Alliance in November went along, the United States and its allies were installing their first ground base near by the city of Kanda- har. At this time approximately 1,000 special operation soldiers were deployed in Afghanistan. In addition to the NATO forces, the International Security Assistance Force (ISAF) was established by the UN in December 2001, initially to support and stabilize Afghanistan, especially Kabul, and later under the NATO command to assist the OEF-A offensively. The Bonn Agree- ment marked the beginning of the construction process of the state and the interim government, which has been led by Hamid Karzai since December 2001. In 2002, the Loya Jirga elected Karzai as president, whose second presidential term ends in 2014. In general, the political culture of the post-Taleban system was actually more characterized by cooperation than by rivalry. But this cooperation was based on a pragmatic perspective and lacked institutional stabilization. First of all, the constellation of the Afghan society would require a consensual democracy instead of a presidential system that gives the executive branch too much power. Moreover, the ethnic groups opposed the centralization of the political system and favored a more decentralized alignment and more ethnic autonomy. Another problem emerged among the Pashtuns when Karzai became President. The motive to appoint the Pashtun Karzai was to get the support from the Pashtuns, however, Karzai caused a division between the Karzai proponents and adversaries among the Pasthun people and a division between Pashtuns and non-Pashtuns in the Afghan parliament (Barfield 2011: 54-65). After the United States had imposed sanctions against Pakistan because of its nuclear program, the aftermath of the Soviet retreat and Musharraf’s military coup in 1999, Pakistan became an important partner of the US overnight once again. Pakistan’s decision to cooperate

15 4 POLITICAL, MILITARY AND SOCIAL CONTEXT with the United States was attributed to the intention (1) to extend its position in Kashmir towards India, (2) to intensify the nuclear program, (3) to improve the economic circumstances in the country and (4) - perhaps the most crucial aspect - to accommodate with the administration’s maxim ”with us or against us”, which threatened the Pakistani regime considerably (Freedman 2009: 6). From 2002 on Pakistan got approximately one billion US dollars of military aid per year, several hundred millions for the economic development and another 32 billion USD from the international coalition, which superficially benefitted Pakistan’s security sector (Wagner 2010: 171). In return, Pakistan ensured its air space and military bases, sent thousands of troops to the border region to Afghanistan and provided logistical and intelligence support. In the year following 9/11, Pakistan arrested 420 foreign Al-Qaeda activists, of whom 332 were extradited to the United States and 34 to other countries (Gaier 2011: 173-187). Nevertheless, the relations between Pakistan and the USA have remained tense for several reasons since the beginning of the conflict: Pakistan admittedly took action against Al-Qaeda, but refused to fight against the Afghan Taleban because of their intention towards India, which is clearly a problem for the US. However, the considerable aids to Pakistan did not change the unstable situation sustainably. Furthermore, Pakistan tried to reach an equal status with India towards the United States, which has obviously not been achieved yet.7 Under the Bush administration, general Stanley McChrystal and David Petraeus became the spiritus rectors of counterterrorism, COIN and the Joint Special Operation Command (JSOC). Implementing the conception of McChrystal meant to improve the flexibility and speed of the US forces, to connect all parts on the battlefield in a holistic information network, to implement new technologies in order to take advantage at night, to enhance the collaboration between military and intelligence and to establish a flatter hierarchical structure (McChrystal 2013: 2-8; Robinson 2012: 110-122). Technically COIN and counterterrorism are linked with each other and contain many more facets, such as the role of strengthening the democratic institutions, handling corruption and patronage, incorporating multilateral approaches and development programs, accelerating low intensity warfare and coaching the local governmental security forces. The dual application of OEF led by US forces and ISAF under NATO reflected the combination of military and civilian strategies, which aimed for confidence-building measures among the population. Methaphorically speaking, the objective of COIN is to win the ”hearts and minds” of the local people (Hühnert 2011: 451-472). But it turned out that this strategy was hardly effective in practice. In 2002, the so-called Provincial Reconstruction Teams (PRTs) were deployed to support local security and the central government and to reconstruct the damaged areas. After the Taleban government was defeated, the US forces focused on counterterrorism and Al-Qaeda. But the fall of the Taleban government did not mean their final end. Especially the Iraq intervention required troops and intelligence from Afghanistan and paved the way for the reemerging Taleban in 2004 and the expansion of the COIN strategy. The aspect of equality with India was getting worse when President Obama introduced his Af-Pak strategy, which changed several key aspects of the entire relationship between Wash- ington and Islamabad: (1) Since then Pakistan has been part of the problem and no longer an equal partner. The problem is that the Pakistani foreign policy planned to gain strategic

7The United States and India signed the so-called U.S.–India Civil Nuclear Agreement, which guarantees an intense cooperation in nuclear aﬀairs. Pakistan urged the US administration to get its own special status, which was rejected by the US because of A.Q. Khan’s unresolved issue of his nuclear network.

16 4 POLITICAL, MILITARY AND SOCIAL CONTEXT depth in Afghanistan towards India, but its purpose led to the opposite outcome, particularly to a strategic depth of the Taleban in Pakistan. (2) Another shift was the understanding of the US strategists that a sustainable success to fight insurgencies needs to strengthen democratic institutions, the rule of law and a comprehensive economic development. From the beginning of Pakistan’s history the political system has been marked by a power struggle between the democratic and military sphere, by an elitist corruption and by a constrained control over governmental institutions, such as the ISI. (3) On a military strategic level Obama focused on the usage of drones, which are obviously quite effective. According to the Bureau of Investigative Journalism (2013) Obama authorized 325 drone attacks, especially signature strikes, out of a total number of 376 strikes between 2004 and 2013, which caused the deaths of 2,525 - 3,613 people (407 - 926 civilians). A study by Jäger and Siddique (2011) statistically examined the deterrence effect of drones on Taleban groups in Afghanistan and Pakistan. They show that the drone strikes have a significant impact on the prevention of terrorist attacks, especially in Pakistan, whereas the impact in Afghanistan has little significance. However, the Taleban respond to the drone strikes with attacks on civilians. Moreover, a study by the Stanford Law School and the Global Justice Clinic (2012) shows that drone attacks cause substantial negative psychological and emotional effects on civilians and thus undermine the American reputation in the region. (4) Although the number of troops was bulked up by Obama in the mid-term, the strategy intended to lower US commitment in the whole region and to increase the local engagement in the long-term. The reason for this is that the United States wants to adapt the security strategy to new circumstances, such as the budgetary burden of the war, the need to balance China in the Pacific region, the loss of popular support, both at home and abroad and the necessity of a multilateral approach. (5) Obama also piled the pressure on Pakistan to fight against the Pakistani Taleban, the Haqqani network and Hekmatyar. The challenges to the Pakistani government, conducted by Obama, caused political tensions between the two. Although the Pakistani government informed Obama of the locations to attack with drones, Pakistan used the drone issue to mobilize domestically. Another problem causing tensions was the military offensive in North and South Waziristan because of its high human, financial (especially in respect of the floods in 2010) and, of course, tactical costs. Additionally, the Operation Neptune’s Spear, which killed Osama bin Laden in the Pakistani garrison town Abbottabad in May 2011, affected the relations between Islamabad and Washington. If the ISI or the Pakistani government had known about Bin Laden’s residence, it would prove the cooperation between them. If not, it shows the incapacity of the security forces (Wagner 2012: 178) - similar to the issue of A.Q. Khan. Both possibilities would be inconvenient for the US government. On the other hand, the Pakistani government complained that they were not informed about the US led operation and that the United States had violated the national sovereignity of Pakistan.

4.4 The Pivot to Counterinsurgency

The longstanding support for Islamist groups during the 1980s and 1990s led to the emergence of many Pakistani groups and to war with the Pakistani government. The TTP especially became a big challenge for the Pakistani government. The TTP was formed by the Islamic tribal leader Baitullah Mehsud and a shura consisting of 40 Taleban leaders in South Waziristan in 2007.

17 4 POLITICAL, MILITARY AND SOCIAL CONTEXT

They oppose the government of Pakistan, its allies as well as the Shia community and fight for a crude interpretation of Sharia law. Another group that associated and cooperated with the TTP is the Tehreek-e-Nafaz-e-Shariat-e-Mohammadi (TNSM), which basically operates in the Swat Valley (Abbas 2008: 1-3). In addition to the Islamist Taleban groups in the Pakistani Pashtun areas, there are also Islamist Punjabi groups, ”such as the Sipah-i-Sahaba Pakistan (Soldiers of the Companions of the Prophet, SSP); Lashkar-i-Jhangvi (Army of Jhangvi, LeJ); and Jaysh-i-Mohammad (Army of Mohammed, JeM)”, which are responsible for several terrorist attacks in major cities and co-responsible for an increasing number of terrorist attacks by 746% between 2005 (2,267 dead, 4,500 injured) and 2008 (Ghufran 2009: 1092-1114). Under the pressure of the United States the Pakistani government started its offensive against the TTP in South Waziristan and against the Haqqani network in North Waziristan in the fall of 2009 - the latter to the disfavor of Pakistan. Pakistan argued that Haqqani had not attacked the Pakistani government, but the US persisted. The operation was quickly stopped when the Pakistani soldiers were sent to the Indian border after the Pakistani group Lashkar-e-Taiba committed several attacks in Mumbai in November 2008. However, the operations in South Waziristan and Swat District and the clashes between the security forces and Islamist militants at the Red Mosque in Islamabad in 2007 marked a tipping point of the general attitude within the Pakistani state. Since then the state has increasingly stepped up their efforts to get different factions of the ISI under control. Moreover, Jayshree Bajoria (2011) argues that Islamist groups ”are breaking away from the state”. One example of this is the arrest of members of the Islamist militant group Quetta Shura, which was founded by Afghan Taleban around Mullah Omar, by the ISI in February 2010. Another shift was the fact that the military increasingly considered the Pakistani Taleban factions as their enemy instead of India. After years of asymmetric balancing the Pakistani government is faced with its consequential problems: (1) Instead of a Pakistan friendly regime, strategic depth in Afghanistan and social support from the Afghan people the Taleban fighters gained strategic depth in Pakistan. (2) The Pakistani deployment of asymmetric balancing led to the continual intention of India to support Afghanistan and to surround Pakistan. (3) The support for the Taleban did not only promote violence in Afghanistan and Kashmir but also caused instability on the Pakistani territory. This made it impossible to practice political, economic and cultural activities and caused a high intensity of violence and extensive violations of human rights. (4) Another problem right now is that fighting squanders enormous sums of money which would be necessary to reform and modernize the country in order to challenge poverty. Additionally, India knows that Pakistan will lose the economic race and that this will have consequences on the regional balance of power in favor of India. (5) Asymmetric balancing did not only lead to tense relations with the Pakistani rival India, but also with its important partner China because the Xinjiang province is increasingly affected by Islamist terrorism. Because of Pakistan’s geographical proximity to the Persian Gulf and to Iran and China’s considerable expenditure on importing oil from the Persian Gulf through the Malacca Strait, the Chinese invest in the port of Gwadar, pipelines and other infrastructure projects in Pakistan. However, an ongoing instability would undermine these investments and the strategic position of Pakistan within the region. This brings up the question of how close the relations between Afghan and Pakistani Taleban are. As can be seen in the scientific literature, analysts do not have a consistent opinion on this issue. The Afghan and Pakistani Taleban groups deny a political cooperation because

18 4 POLITICAL, MILITARY AND SOCIAL CONTEXT the Afghan Taleban do not see the Pakistani government as their enemy. There were several tensions between different Taleban factions and in some cases these tensions even culminated in violence, such as clashes between Afghan and Pakistani Taleban in the Kunar province in October 2013. However, there are many indications that they cooperate. Firstly, several leaders of the Pakistani Taleban fought in Afghanistan during the 1980s and 1990s and many of them are said to have close ties with Afghan Taleban. Secondly, many Pakistani Taleban factions were established just after the Afghan Taleban were overthrown by the United States and its allies. One of the incompatibilities of the Pakistani Taleban declaring war against the Pakistani government was that Pakistan supported the US-led operation against the Taleban in Afghanistan. Thirdly, there were some attacks on governmental forces in the border region of which both groups took over responsibility. Fourthly, the different Jihadi groups often share the same financial sources, religious and political institutions and combatants and help each other in Afghanistan and in Kashmir. Fifthly, the dataset shows that there were only 48 cases of 4,735 within two years in which different non-state groups clashed, including Afghan and Pakistani Taleban and unidentified militias. But most cases of intergroup violence occured between the two associated Pakistani groups Tehrik-e-Taliban and Lashkar-e-Islam. There were only eight cases in which the Afghan Taleban clashed with Pakistani Taleban or another militant group. The point is that the Durand Line separates political authority on a national level, but obviously it does not disconnect the linkages between the Taleban factions.

4.5 Overview of Battle-related Deaths since 1989

Figure 4: Battle-related Deaths in Pakistan (including Kashmir) and Afghanistan (1989-2012) (Data: UCDP)

The following ﬁgures, based on the macro dataset UCDP, illustrate the trends of violence in Pakistan as well as in Afghanistan, which reﬂect, to some degree, the political dynamics in the region. The graph shows the trend within the Pakistani military that it does not con-

19 4 POLITICAL, MILITARY AND SOCIAL CONTEXT sider India but Pakistani groups as main enemy. The large-scale violence with a loss of 6,533 combatants occured in 2009, which primarily rested upon the military offensive against the Pakistani Taleban (TTP, TNSM) and between 2004 and 2006 the offensive in Waziristan with 3,000 battle-related deaths. The Pakistani graph shows that the end of the Soviet intervention marks the beginning of an intensified usage of insurgents causing 19,121 battle-related deaths8 in the India-controlled Kashmir area between 1990 and 2011. The insurgents played an important role in the Kargil War, which resulted in 3,155 battle-related deaths in total in 1999. Additionally, the graph indicates that the peace process between India and Pakistan under President Musharraf and the US intervention in Afghanistan decreased the application of the insurgents in Kashmir. These findings would support the thesis that the rivalry between Pak- istan and India is linked with the deployment of Islamist groups and the fact that the Pakistani government has tried to reduce support for Islamist militias and to get specific factions of the military and ISI under control since 2001. The right graph demonstrates the violent dynamics of the Afghan and American governments and non-state actors, but exludes the intergroup conflicts and the prehistory of the Afghan civil war, particularly the Soviet occupation in the 1980s. Under the Rabbani government violence peaked with a total number of 8,829 battle-related deaths in 1994. Whereas Rabbani had to operate on several fronts against the different competing groups, his fight against the Taleban caused the highest numbers of battle-related deaths in 1995 and 1996. The highest number of deaths (300) caused by intergroup conflict occured in the same year. Maybe this intergroup rivalry between at least seven different factions (1994-1997) benefitted the Taleban. In 1994, the Taleban took up arms against Hekmatyar’s Hizb-i-Islami. After the high losses of the combats between Rabbani and the different factions in 1994, the Taleban entered the stage against Rabbani at the right moment in 1995, caused further 5,612 battle-related deaths in 1995 and 1996 and emerged as the strongest power in Afghanistan. As the figures 4 and 5 show, the conflict among the non-state actors ended and the Northern Alliance (UIFSA) was formed after the Taleban took over Kabul in 1996. During the Taleban period the number of battle-related deaths declined from 4,538 to 2,877 (1998-2001). But the Taleban celebrated their victory by killing 5,799 civilians in 1999 (see figure 5). In total the Taleban killed 7,464 civilians between 1997 and 2012. Violence continued to decrease when the United States in- tervened in Afghanistan and reached its minimum point in 2003 when the US turned their attention to Iraq. But then the number of battle-related deaths grew rapidly from 2003 with 353 battle-related deaths to 7,225 in 2012 (with a short interruption of 4,532 in 2008). The intergroup conflicts in Afghanistan reemerged after the defeat of the Taleban by the Northern Alliance and the forces of the international coalition. The conflict among non-state actors in Pakistan reemerged in 2007, especially between tribal groups. When the Pakistani government began its campaign in South Waziristan, the TTP also got involved in the intergroup conflict and killed 1,493 civilians between 2007 and 2012. In total, a number of 138,868 battle-related deaths was caused by the civil wars of state and non-state actors and intergroup conflicts in Afghanistan, Pakistan and Kashmir between 1989 and 2012. Including the Kargil War and the violence against civilians in Afghanistan and Pakistan, 11,558 more people were killed, thus making a total number of 150,426 deaths in 23 years.

8The number of battle-related deaths also includes dead civilians.

20 4 POLITICAL, MILITARY AND SOCIAL CONTEXT

(a) Intergroup conﬂict in Afghanistan and violence against civilians by the Taleban

(b) Intergroup conﬂict and violence against civilians in Pakistan

Figure 5: Civil war between non-state actors (l) and violence against civilians (r) (Data: UCDP)

21 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

5 Armed Conﬂict in Centers and Peripheries

5.1 Social Structure in Pakistan

5.1.1 General Imbalances between Rural and Urban Areas

According to the Pakistan Statistical Yearbook 2011 9, Pakistan had a total population of estimated 179.2 million in 2012 (World Bank data 2012), of which 36% live in urban areas. The growth of the urban population in Pakistan had an average rate of 3% per year between 1990 and 2011. By contrast, Afghanistan has a much lower population with 32.3 million, but a higher average growth rate of 6% of the urban population (UNICEF 2011), of which approximately 2.7 million refugees live in Iran and Pakistan. Since 1979 the Afghan population has doubled from 15.5 million to 32 million in 2012 and is expected to grow to 80 million in the next 40 years (Eikenberry 2013: 53). In sum, the ten most populated Afghan cities are almost as large as half of Karachi - the largest city in Pakistan with an estimated population of 11.1 million. Comparatively, the total population number of the ten largest cities in Pakistan is as high as the population of Afghanistan. In contrast to other developing countries, the urbanization of Pakistan is not due to internal migration trends but rather a product of the higher birth rates in the cities. According to the CIA World Factbook (2014a & b) about 54.8% of the Pakistani population are under the age of 25, which is in combination with the lack of a job market an explosive mixture. In Afghanistan 64.2 % of the Afghan population are under the age of 25 and 42% under the age of 15 and the lack of institutions is even more obvious. Not only did the urban population grow continuously, but the total Pakistani population also grew from 40 million in 1950 to 140 million in 1996 and an estimated population of roughly 200 million is expected in 2020 (United Nations Data 2012), however, the growth rate decreased from 2.4% between 2001 and 2007 to 1.8% between 2008 and 2010 (Wagner 2012: 79). This population growth is rooted in a complex mixture of several factors and has effects on urbanization, economy and social indicators. As Fakhari (1998: 557-574) stated, the main reasons for the high growth rate of the population are declining death rates, ”husbands’ preferences, religious reasons, and lack of knowledge”. Looking at the poverty line the statistics show significant differences in the poverty rate and other economic and social indicators between the rural and urban areas in Pakistan. In 2006, 27% of the rural inhabitants lived below the national poverty line, compared with 13.1% of the urban population. The data of Afghanistan is worse than the Pakistani figures. 37.5% of the Afghan rural population live below the poverty line as well as 29% of the urban population (World Bank data). The long years of conflict and poverty have influenced the cities and rural areas. Population growth and migration caused changes in the ethnic composition of society, such as the increasing weight of the Pashtuns in Balochis- tan (Wagner 2012: 112). One problem of the social and economic standard is the absence of differentiated social systems, particularly a market. Instead of an innovative entrepreneurship, the economy is formed by some influential families and the military that remains the largest employer, landowner and carrier and corresponds with the biggest industrial complexes (2012:

9The yearbook of the Federal Bureau of Statistics primarily refers to the last oﬃcial national census in 1998. There are several reasons why the government has tried to prevent a national census since 1998. Firstly, an increasing urbanization would possibly result in a redistribution of the political mandates to the disadvantage of some political elites. Secondly, it can be assumed that the increasing population of speciﬁc districts, for instance the Pashtun-dominated districts in Karachi, could provoke tensions between the ethnic groups again (Mielke, Schetter 2013: 155).

22 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

85). As Siddiqa (2007: 241) stated, under Ayub Khan ”38 business houses controlled over 60 per cent [. . . ] of business assets”, 42 in 1970 and 44 in 1997.

Table 1: Social Indicators of Pakistani Provinces (Data: Census 1998) K.P. FATA Punjab Sindh Balochi- Islama- Total (No. 5) (No. 6) stan bad Population 17.74 3.17 73.62 30.44 6.56 0.8 132.35 (m) -Pop. 1951 4.55 1.33 20.54 6.04 1.16 0.09 33.74 -Pop. 1961 5.73 1.84 25.46 8.36 1.35 0.11 42.88 -Pop. 1972 8.38 2.49 37.6 14.15 2.42 0.23 65.3 -Pop. 1981 11.06 2.19 47.29 19.02 4.33 0.34 84.25 Male (%) 51.22 52.01 51.74 52.88 53.40 53.93 52.03 Female 48.78 47.99 48.26 47.12 46.6 46.07 47.97 Area 74,521 27,220 205,345 140,914 347,190 906 796,096 (sq km) Population 238.1 116.7 358.5 216 18.9 880.8 166.3 Density (Persons/sqkm) Population 47.2 - 42.52 42.76 46.67 37.9 43.4 under 15 (%) Urban 16.87 2.7 31.27 48.75 23.90 65.7 32.5 Population (%) Migrant Popu- 0.647 - 6.701 2.832 0.249 0.397 10.829 lation (m) Literacy Rate 35.4 17.4 46.5 45.2 24.8 72.8 43.9 -Male 51.3 29.5 57.2 54.5 34 80.6 54.8 -Female 18.8 3 35.1 34.7 14 62.3 32 Unemployment 26.8 - 19.1 14.4 33.4 15.7 19.6 Rate (%) Size Average 8 9.3 6.9 6 6.7 6.2 6.8 Household Mother tongue -Urdu 0.7 0.1 4.5 21 0.9 10.1 7.5 -Punjabi 0.9 0.2 75.2 6.9 2.5 71.6 44.1 -Sindhi 0.04 0.01 0.1 59.7 5.5 0.5 14.1 -Pushto 73.9 99.1 1.1 4.1 29.6 9.5 15.4 -Balochi 0.01 0.04 0.6 2.1 54.7 0.06 3.5 -Saraiki 3.8 - 17.3 1 2.4 1.1 10.5 -Others 20.4 0.4 0.9 4.9 4.1 6.9 4.6

The aﬀected provinces Balochistan (no. 1), FATA and Khyber Pakhtunkhwa (KP) represent the structurally weak periphery of Pakistan. Some characteristics of the peripheral area can be deduced from demographic, social and economic indicators. Apart from the province Islamabad, KP, FATA and Balochistan have the lowest population with 27.47 million which is 20.7% of the total number of 132.35 million. They have the lowest literacy rate, the lowest percentage of the urban population, the highest unemployment rate and the highest proportion of people under 15. FATA and Balochistan have the lowest population density with an average number of 116.7 and 18.9 persons/sq km. FATA does not have the administrative status of a Pakistani province, but rather has a special set of rules with the Pakistani state, which dates back to British rule. The social coexistence is usually organized by the informal social code Pashtunwali and some Islamic traditions. On the local level, FATA is an important strategic area for Pakistan because it acts as a buﬀer zone between the so-called settled districts and the border to Afghanistan.

23 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

But one difficulty is that dynamics in FATA can only be partially controlled by the Pakistani government because of the special administrative status of FATA and the interrelations between the different Pashtun tribals in Pakistan and Afghanistan. In contrast to this, the provinces Islamabad, Punjab and Sindh had a total population of 104 million in 1998. They also have high urbanization levels, below-average unemployment rates and relatively high literacy rates. But the most interesting province concerning politics is Islamabad as the political center of Pakistan and its geographical location. Islamabad was founded by general Ayub Khan after his military coup to relocate the capital from Karachi closer to the military center of Rawalpindi, which had already served as an important military cantonment for the British during the Great Game. However, Rawalpindi is not only a large garrison town, but also a vital center of the Pakistani rail network and science programs, such as the nuclear program Kahuta Project in the east of Rawalpindi in the city of Kahuta, which was established under President and then Prime Minister Zulfikar Ali Bhutto. Another milestone for Islamabad was the beginning of the construction of the Karakorum Highway, which connects the Chinese province Xinjiang with Islamabad in order to improve the trade between the two states.

5.1.2 Political, Ethnic and Economic Structure

Considering the locations of Rawalpindi and Islamabad on the Pothohar Plateau in the north of Punjab, in the east of Khyber Pakhtunkhwa and in the west of Azad Kashmir, the military can control its political power base. This is due to the fact that the military level of command depends on the land owning social classes in Punjab and that the social structure of the army is dominated by Punjabi people. Additionally, both cities are highly protected by the military, such as the garrison in Abbottabad, which became famous for being Bin Laden’s last hideout. Although the social structure of the military had been changing since the 1990s, in 2005 44.3% of the army personnel came from Punjab, 22.4% from KP and FATA, 23% from Sindh, 1.5% from Balochistan and 9.7% from Northern Areas and Azad Kashmir. In 1991, 63.8% of the military personnel came from Punjab. Between 1991 and 2005 the proportion of personnel from Sindh grew from 8.8% to 23% (Nawaz 2009: 26). Within Punjab personnel is mostly recruited from urban centers in the south and the middle of Punjab. On the one hand, this development can improve the image of the Punjabi dominated army within the Pakistani society because of its increasing heterogenity. On the other hand, there are more Islamist groups in southern Punjab than in the north, not least because of poor circumstances with which people live, so that this development could also entail negative aspects (Wagner 2012: 64, 65). From the beginning of Pakistan’s history one main pillar of the state was erected on the alliance between the immigrant Indian Muslims Mohajirs and the traditional elite consisting of Punjabis. While the Punjabis primarily dominated the military, the Mohajirs found their place in the public administration because of their urban and intellectual background. The Mohajirs, who are the Urdu-speaking Pakistanis, are located mainly in the former capital and economic center Karachi and in the urban areas of Punjab, especially in Lahore and Rawalpindi. The domination of this ethnic axis is reﬂected in the language rates of Islamabad in table 1. 71.6% of Islamabad’s population speak Punjabi and 10.1% Urdu - the two largest groups of the capital in 1998. The Mohajirs and Punjabis feared an alliance between the representatives of Sindh, Bengal and the North West Frontier Province, who refused to centralize political power to the

24 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES favor of the Mohajirs and Punjabis. This and several disagreements about the role of religion and the official language were some of the reasons why the democratic process was not ivolved and the state was bureaucratically led during the 1950s. The decision to change the capital is also attributed to the exposure to India because of two fundamental reasons. Firstly, the geographical location of Islamabad improved the connection between the political center and the main conflict zone of the rivalry, Kashmir, and secondly, Islamabad and the area around the city function as a kind of soft power projection towards Pakistan’s major rival. Balochistan, a highly perhipheral province, often suffered from the interests of different groups for an independent state or more political autonomy. The central government has been neglecting the province politically, socially and economically for a long time, so that the Baluch are the poorest Pakistanis today. However, the province is the richest in natural resources. Balochistan provides natural gas, minerals and also the strategically important coastline. The most intense violent conflict between Baloch militants and the central government occured in 1973 lasting until 1977 and caused 8,000 battle-related deaths in total. Since the beginning of the new millenium disputes between Baloch groups and the government have been reemerging. But the nature of the conflict has changed slightly: Firstly, the present groups are not supported by the Baloch society and are highly fragmented. Secondly, the contemporary conflict is not manifested by the incompatibility between groups that fight for more provincial autonomy and the central government, but rather by the incompatibility with the fate over the natural resources. As a consequence, a series of attacks were targeted at critical infrastructure, especially in 2004 at gas pipelines that supply energy to the more populated provinces. In 2002, China and Pakistan started to enlarge the strategically important port of Gwadar, which makes it easier for China to import oil from the Persian Gulf by avoiding the Strait of Malacca. As a reaction militant groups attacked Pakistani security personnel, Chinese workers and engineers. In recent years immigration from Afghanistan increased the number of Pashtuns, who live in Balochistan, with the result that at least as many Pashtuns live there as Balochis. During the 1950s the bureaucratically led government focused on urban industry and manufacturing instead of agriculture because this new political elite, who initially had been immi- grants from urban areas in India, were not keen on agriculture and had no indigenous base in the rural areas. The axis of the military and bureaucracy was established when Ayub Khan came to power in 1958. Ayub Khan and his regime did not only constrain the press but also shaped the economic activities of the whole country autocratically and implemented regulative laws, protected the domestic economy and controlled the import flows, which led to a lack of competitiveness and efficiency and to a low export rate. Until the 1970s the national economy was heavily dependent on agriculture which made more than 50% of the total exports. This industrial policy led to major changes in the 1960s. Bureaucracy and military as the fundamental pillars of the state were now linked with the emerging industrial and agricultural elites who sought their personal wealth through a patronage system, but who were controlled by Ayub Khan’s regime. Furthermore, high growth rates were accompanied by a process of urbanization, the green revolution, a rising educated middle class and the emergence of different social and economic groups. But the middle class remained small and at the same time capital was concentrated on the upper class and shifted from rural to urban areas. Because of the green revolution a high number of the rural population lost their land and were increasingly dependent on the landowners for agricultural equipment, chemicals, water canals etc. As a

25 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES result rural poverty increased despite the immense growth of agricultural production. By the end of the 1960s the patronage system favored 43 families who dominated the industry, banks and insurances and gave them significant influence on economic policies of the government, but the new social classes and the growing poverty strenghened the political leader of the Pakistan People’s Party Zulfiqar Ali Bhutto, who came to power in 1973. Bhutto’s power was based on the rural landowners of his party and the polarized lower and middle classes in the urban areas. This initially led to a leftist economic policy against private capitalism and a nationalization of companies. However, this and previous measures did not really facilitate the redistribution of wealth in order to attain social equity, but rather led to a higher degree of state patronage, a lower rate of the productivity of investment, a decline in the private sector and consumption, a lack of a diversified industry, a higher expenditure for security and administration, higher inflation and a low growth rate of the national economy (Akmal 2004: 69-102). As a consequence the poverty rate and dissatisfaction among the society increased, which resulted in demonstrations against Bhutto and his replacement by Zia-ul-Haq in 1977. Zia, whose legitimacy primarily was based on Islamic ideology and authoritarian military rule, pursued the denationalization of the economy and received financial support from abroad, especially from the West and Middle East, due to the war in Afghanistan. During this time the military and bureaucracy consolidated their economic and political power and achieved high growth rates, but when foreign credits were abandoned because of the Soviet withdrawal from Afghanistan, the structural weakness of Zia’s economic agenda became obvious. In the meantime Zia’s Islamization of the society and the growing middle class in the urban areas led to a rise of Islamic political parties, such as Nawaz Sharif’s Islami Jamhoori Ittehad (IJI) which were supported by the ISI (Zaidi 2005: 5173-5181). Together with other Islamic parties they formed the coalition Pakistan Muslim League which has mainly been supported by the conservative urban population of Punjab and new economic groups to challenge the PPP. Additionally, the PML got votes in Khyber-Pakhtunkhwa and Balochistan, where the coalition became a significant oppositional party. However, the democratic era in the 1990s was not really democratic, but rather a chaotic and corrupt bargaining between the political parties and their linkages with part of the economic middle class and the traditional institutions of the state. As some scholars claim, the point is that social, economic and political developments and the process of urbanization are interlinked with each other and have led to major changes since the 1960s. Firstly, urbanization came along with the appearance of new sectors that replaced the economic weight of the rural agriculture, such as jobs in industry, services and culture. Secondly, these new sectors were accompanied by the emergence of ”urban lower middle classes, the educated and professional groups, and traders and medium and small industrialists” (2005: 5173-5181). Thirdly, the Pakistani state has become more and more inclusive since the 1980s. The new social classes increasingly shape domestic politics through their linkages between the institutions and parties. The provincial level was an important entry into the Pakistani system for political newcomers. But obviously they were not able to form a unified civil society transforming the political system into a democracy. In addition to ethnic cleavages intra-group cross-cutting cleavages, such as economic classes, sectarian groups, ideologies, and opportunistic attitudes complicated the political formation process. As in many other developing countries, the Pakistani state still lacks effective political and economic institutions that can organize the rapid increase in the population and solve other political and social problems.

26 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

The case of Pakistan shows that the dominating groups share social and geographical characteristics. In other words, the borders of a specific geographical space coincide with the borders of political, military and social networks. The geographical locations and their conditions of the peripheral territories play an important role in Pakistan’s strategic environment. East Pakistan, for instance, served as a strategic outpost to encircle India. Balochistan functions as a buffer zone towards Iran and Afghanistan and makes crucial resources and the coastline available. Gilgit-Baltistan (formerly called Northern Areas; No. 7 Fig. 7) connects Pakistan with China and has rich deposits of natural resources, especially hydro energy and minerals. The Balochistan conflict, the secession of East Pakistan in 1971, the ethno nationalistic groups in the Pashtun region as well as the ethnic rivalry between Sindhi and Mohajir are examples of struggles for power between the government and the provinces and between different ethnic and ideological groups. But they can also be seen as conflicts between the center and the periphery.

5.2 Social Structure in Afghanistan

5.2.1 General Imbalances between Rural and Urban Areas

After 34 years of civil war the capital suffers from intense damage to buildings and also migration waves, especially internally displaced people, of which many do not have a supporting family or an adequate social standard. Kabul has been a center since ancient times because of its geographical location and its logistical relevance. To the north several passes lie across the Hindu Kush and to the south-east the Kyber Pass towards Pakistan. Moreover, Kandahar to the south-west is reachable from Kabul. As table 2 shows, Kabul has the highest population number, population density, urbanization level, literacy rate and the highest proportion of people who work in the service sector. The province has the lowest underemployment rate, the lowest maternal mortality rate and, of course, the lowest proportion of the agricultural sector. By contrast, the least populated provinces ranked among the highest ratios of the agricultural sector, the lowest population density rates and among the highest numbers of maternal mortalities. Other major cities and provinces have peripheral structures too, such as high poverty rates, high underemployment rates and relatively high maternal mortalities. But those provinces locate big cities which are a kind of local centers of the periphery, as can be seen in figure 6 and by the population density, the urbanization level and the ratio of the sectors in table 2. Throughout history different regions of present Afghanistan were either incorporated in different empires, such as the ancient Persian empire or the regional empires of the Kashuns and Ghaznavids, or were autonomously political units. There are four major regions and their historically local cities: ”Herat in the west, Kandahar in the south, Balkh (Mazar-i-sharif) in the north, and Kabul in the east” (Barfield 2010: 48).

5.2.2 Ethnic and Social Structure

The largest group of the Afghan society are the Pashtuns with a ratio of about 40%. The south and south-west are dominated by diﬀerent Pasthun groups, particularly the Durrani, who can be subdivided into diﬀerent tribes. Another Pashtun confederacy is constituted by the Ghilzai Pashtuns, who are located mainly between Kabul and Zabul, but are dispersed over large parts of Afghanistan as well as Pakistan. In addition to them, there are some smaller

27 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES - 25 62 2.6 0.14 15.2 55.3 92.2 - 39 33 16 9.5 50.4 76.1 0.146 - 22 31 32 3.8 0.15 54.4 55.9 - 1 2 4.3 0.28 16.6 58.5 94.3 33 24 18 0.95 63.3 58.2 18.1 118.6 5 16 27 1.15 31.3 22.4 45.3 42.5 44 32 49 1.24 72.1 29.1 38.5 43.2 29 15 19 1.43 54.9 53.7 23.3 185.8 har (No. 21) (No. 4) (No. 18) (No. 32) (No. 22) (No. 23) 36 28 22 1.78 32.4 51.6 50.2 26.7 58 48 47 3.95 18.9 16.5 50.1 885.3 (2012) (2005) (2007) (2007) (2007) (2005) Kabul Herat Nangar- Balkh Kandahar Kunduz Zabul Nimroz Panjshir Nuristan (Pers./sq km) Table 2: Demographic and social indicators of selected provinces in Afghanistan (Data: Civil-Military Fusion Centre) (m) rate (%) UnemploymentRate (%) 10.7 (2007) 9.9 1.6 8.2 4.4 8.3 3 10.5 14.3 5.1 Population Area (sq km)Population Density UrbanPopulation (%)Literacy Rate 4,462(%) -Male-Female (2012) 54,778 83.7 7,727Underemployment Sectors (%) 17,249-Agriculture 28 66 54,022-Manufacturing,mining, etc. 14.6-Services 8,040 43Poverty 16.9 Rate 17,343(%) 36.1Housholds with 41,005 (2007) electricity 41 (%) 21.9Maternal 34.4 mortality 3,610ratio (Deaths 23.1 per100,000 births) 19.6 9,225 54 24.8 700 38.7 (2005) (2007) 16.2 3.9 26 900 33 10.6 15.9 1,100 40 60.3 21.9 - 1,800 1 22.8 0.5 1,100 - 30 29.7 1,800 11 2,100 37.4 43 6.5 1,600 24.5 2.9 - - 22.6 49.8 2,100

28 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

Pasthun groups, such as the Gurghusht or the Karlanri, of which the latter is located at the border between Afghanistan and Pakistan and in Khyber Pakhtunkhwa. Among the Pashtuns, a minority of the groups lives as nomads and leaves the mountainous areas behind during the winter. Kandahar is a local center of many Pasthun tribes as well as warlords and criminal networks. Under the Taleban rule the city became the factual capital and was stabilized by the disarming and expulsion of different militias which came back after the Taleban were defeated. The second largest group are the Sunni Muslims Tajiks with a total number of 30%, who can also be subdivided into different factions. The Tajiks dominate the urban areas of Kabul, Mazar-i-Sharif and Herat, but they are also dispersed over a wide rural area in the north-east. The urban Tajiks traditionally play a dominant role in all kinds of bureaucracies because of their Persian language. Additionally they are overrepresented within the Afghan military with 40% of military personnel being Tajiks. However, the Pasthuns dominate the command level, especially since the formation of local security forces in the south (Adeney 2008: 535-557). One of the main local centers of the Tajiks is Herat, from where they hold strong relations with the Tajiks in eastern Iran, who facilitate more trade and strategic depth. But the Tajiks also dominate Kabul and the administrative sector, so that Karzai has difficulties prevailing politically. By contrast, the Shia community, Hazaris, living mainly at the Hindu Kush in central Afghanistan, ”were systematically excluded from almost all government positions” but gained ”parity with other groups under the constitution of 2004” (2010: 26), at least formally. Addi- tionally, they suffered the most from racism and ethnic cleansing under the Taleban. They are the third largest group with about 15% of the Afghan population. The Uzbeks and Turkmen are the fourth largest group with 10% of the total population. Both groups moved as nomads to Afghanistan but became sedentary in the north-west to the borders to Iran, Turkmenistan and Uzbekistan. As a consequence of the Russian revolution, more Uzbeks and Turkmen migrated to Afghanistan, especially under repression from Stalin. They speak Turkish, belong to the Sunni Muslims and are not organized tribally. Indeed, they are not sufficiently represented in the governmental administration, but have become a considerably political and military force in the north-west. Another Afghan group are the Aimaqs, who are located in the east of Herat and in the west of Hazarajat and make up 5% of the total population. There are also several smaller minorities living in Afghanistan, such as the relatively isolated Nuristanis or the Shiite Turkish speaking Qizilbash, who are strongly represented in the elite of Kabul. Other groups are the Baluchs, living in the desert to the border to Pakistan, the non-Sunni Pamiris with their own language and many other minorities with local identities. Last but not least, the international workers, primarily located in Kabul, play an important role in Afghanistan. According to the Afghanistan Statistical Yearbook 2012-13, 78.1% (in absolute numbers 1,716) of the NGOs are located in Kabul, of which 257 come from abroad. There is a high presence of international security forces, diplomats and other officials, who often live in separated districts and in their own political, social and economic microcosms (Fluri 2009: 986-994). Although Kabul got more support from the international coalition than other provinces, the city is still faced with damage caused by the conflicts, a high unemployment rate, a lack of water supply and health care, high accommodation costs, a high number of landmines and poor circumstances of the majority of the Afghans. In sum, the society has a multi-ethnic composition and a multiple horizontal fragmentation

29 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES based on ethnicity, language, tradition and, of course, religion. Some groups dominate specific geographical spaces, but because of the internal migrations, slave trades and forced displace- ments many groups are spread over different regions. So, several provinces are just dominated by a plurality and not a majority. There have always been struggles for power between the different groups to dominate Kabul, in particular local commanders who can influence governmental institutions and possess a local power base in a tribal or a local jirga. Controlling the governmental institutions of the center means holding much of the political power, although it does not mean full control over the peripheral regions and their local centers. Another problem has always been the huge political, economic and social imbalance between the central government and the local administrative units. One measure with dealing with this problems is the deployment of the Afghan Local Police (ALP) and the Village Stability Operations (VSP), which are components of the COIN-Strategy in Afghanistan and coexist with ISAF and the Special Operations Joint Task Force-Afghanistan. This measure has three main functions: governance, development and security in villages and cities. Whereas the ALP protects checkpoints around villages and ensures security, the VSP teams try to integrate the villages politically and to develop them socially. But their capabilities are restricted. In terms of political and cultural resources the state lacks educated people who can organize the society on a district level. The VSP only consists of about 800 personnel forming around 60 teams, however, it is an important program in order to deal with rural villages. The ALP should reach a number of 28,500 by February 2014 and 30,000 personnel by December 2014 (U.S. Department of Defense 2013: 69).

5.3 Patterns of Violence in Centers and Peripheries

5.3.1 Empirical Results of Violence on the Macro Level

Figure 6: Hotspots of violence in Afghanistan and Pakistan (Data: ACLED, 2008-2009)

30 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

Figure 6 shows the conflict centers that include all kinds of violent events (battles, violence against civilians and others). Generally, the most violent events occur in the periphery, whereas the large urban areas in the east of the Indus river are only occasionally affected by violent events. Islamabad seems to be less affected than Kabul, but there are more interrelated violent hotspots in the Pakistani province FATA (see no. 2) and its neighboring province Khyber Pakhtunkhwa, formerly known as North-West Frontier Province (see no. 4). By contrast, in Afghanistan the conflict is dispersed over different provinces but within a province often concentrated on local centers. There are also some large Pakistani cities and local centers, which are located in a conflict zone, such as Peshawar, Quetta or Mingora. Of a total number of 4,946 violent events in Afghanistan and Pakistan between 2008 and 2009, 2,584 events occured in Pakistan and caused a total number of 8,049 battle-related deaths and 2,021 dead civilians, making a total number of 10,070 casualties within two years. In 1,764 of 2,584 cases there are either no numbers given or there were no battle-related deaths. In 2,021 cases this is applied to civilians. In Afghanistan a total number of 5,664 deaths were counted between 2007 and 2008 although the information about casualties in the first half of the year 2007 is incomplete. The average of the given numbers of deaths in Afghanistan and Pakistan differ from each other. Whereas in Pakistan an average number of 9.82 combatants and 3.58 civilians per violent event were counted, an average of 6.01 combatants and 5.79 civilians died in violent events in Afghanistan. Looking at the figures of the Pak- istani provinces (see table 3), 92.9% of the violent events occured in FATA, Khyber Pakhtunkhwa and Balochistan, which together make up 56.39% of the Pakistani territory, but where only 20.75% of the Pak- istani population live (in 1998). Including the highly affected Khyber Pakhtunkhwa and FATA, 81.7% of all violent events were located in 12.77% of the Pakistani territory, which is inhabited by 15.79% of the total population. 6.3% of violent events occured in the highly populated provinces of Punjab and Sindh. As can already be seen, Islamabad was much more secure than Kabul. Azad Kash- mir, the disputed territory in the east of Is- lambad, and Gilgit Baltistan (No. 7) were relatively safe with only 0.6% of all violent events. One interesting aspect is the toppling combatant-civilian ratio towards the higher populated provinces, which can be explained Figure 7: Drone Attacks between 2008 and 2009 by the higher concentration of civilians in urban areas. The number of suicide attacks in urban spaces increased in the last years and militants use urban areas and civilians as a pro- tective shield more often. 87 drone attacks of a total number of 92 were carried out in FATA.

31 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

Unfortunately there was no information about drone attacks in Afghanistan given. According to The Bureau of Investigative Journalism (2012) 326 drone attacks were carried out by Great Britain and the United States in Afghanistan in 2008 and 2009. Of a total number of 114 involvements between 2008 and 2009, the United States acted in 108 cases as primary party in FATA, compared with seven involvements in the other provinces. The armed forces of Pakistan, which include all kinds of military organizations on behalf of the government (Pakistan Armed Forces, paramilitary forces, Frontier Corps, etc.), were more than a thousand times in action within two years. In 989 cases one of the two primary parties could not be identified. The most active group fighting against the Pakistani government was the TTP with a total number of 634 events as a primary party. However, the table shows that the prevailing situation of FATA and Khyber Pakhtunkhwa regarding the actors is highly complex. Many unidentified militias and other groups, Al-Qaeda, the TTP-associated Lashkar-e-Islam, Pakistani and Afghan Tale- ban, several tribal militias as well as the international and local actors contribute to a high militarization and a great instability in FATA and K.P. Many of these tribal militias are pro- government and fight against the increasing influence of Pakistani Taleban and other Islamist militants. Violence against civilians often contains cases in which tribal leaders were executed because they had stood on the side of the government or because they were blamed for spying. Other cases of violence against civilians include one-sided violence against the Shia community. Compared with the numbers of Afghanistan (see table 4), the tribal militias in Pakistan were more active than the Afghan tribals. In sum, Pakistani tribal militias were involved in 136 cases, whereas in Afghanistan there were only four cases. Looking at the geographical areas, the most affected provinces in Afghanistan and Pakistan lie in the so-called Pashtun belt. 50% of all violent events occured between Helmand and Paktika. The Taleben were involved in 58.8% of the total number of events. In addition, in 35% of the cases unidentified militias were engaged. Interestingly, the conflict in Pakistan is less dispersed geographically and the conflict actors have a low cohesion on the side of non- governmental actors. In 1.3% of the cases the Taleban acted as a primary party, in 24% the TTP, Lashkar-e-Islam in 4%; the other Pakistani Taleban make up 13.3%, tribal militias 5.2%, other groups 12.7% and in 38.2% of the cases an unidentified militia acted as a primary party. On the side of the governmental actors, there is a higher homogenity in Pakistan. The Pakistan Armed Forces were actively involved in 40% of all cases, the police forces in 9.9%, the US forces in 4.4% and the other international forces were actively involved only in 43 cases (1.6%) as primary party on Pakistani territory. However, in Afghanistan the Afghan army was involved in 17.2% of the total number, the police forces in 15.9%, the US forces in 12% and the other international forces with a percentage of 33.8. To sum up, whereas in Pakistan the side opposed to the government was much more fragmented than in Afghanistan, the governmental side was less fragmented and operated in a less dispersed conflict zone than in Afghanistan.

5.3.2 Empirical Results of Violence on the Micro Level

The descriptive statistics illustrate that there is a dominant pattern of the numbers of violent events among the largest cities in Afghanistan and Pakistan. In five out of eleven cases (Kabul, Islamabad, Rawalpindi, Peshawar, Kunduz) the figures show that there is a quantitative differ- ence between the three main areas (center, semi-periphery, wider periphery). Comparing the three main areas, the cities are exposed to a relatively high intensity of violence per square

32 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

(a) Kabul (b) Islamabad

(e) Kunduz (f) Mingora

Figure 8: Violent events in and around major cities

33 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

(a) Herat (b) Jalalabad

(e) Quetta

Figure 9: Violent events in and around major cities

34 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES

Table 3: Overview of violent events in Pakistani provinces (2008-2009) FATA K.P. Balochi- Sindh Punjab Azad Gilgit Islama- Total stan Kashmir Baltistan bad Frequency of 1,064 1,049 290 89 75 11 4 3 2,584 Violent Events Percentage 41.1 40.6 11.2 3.4 2.9 0.4 0.2 0.1 100 Casualties 4,877 4,235 545 172 220 18 - 3 10,070 -Battle-rela- 4,111 3,412 381 39 85 18 - 3 8,049 ted Deaths -Civilians 766 823 164 133 135 - - - 2,021 Combatant-ci- 1:0.18 1:0.24 1:0.43 1:3.41 1:1.5 - - - 1:0.25 vilian ratio Number of 87 5 ------92 drone uses Actors -Pakistani 526 455 47 5 6 8 - 2 1,049 Armed Forces -Police Forces 35 139 53 12 14 2 - 1 256 -Forces of the 108 5 1 - 1 - - - 114 United States -Other inter- 8 21 8 2 - 4 - - 43 national Forces -Taleban 25 10 1 - - - - - 36 -Tehrik-e- 353 265 3 5 8 - - - 634 Taliban -Lashkar-e- 66 39 ------105 Islam -Other Paki- 133 202 4 2 2 - - 2 345 stani Taleban -Tribal Mili- 86 9 36 5 - - - - 136 tias -Al-Qaeda 16 ------16 -Other groups 156 122 37 6 7 2 - - 330 -Unidentified 294 383 206 57 40 5 3 1 989 Militia -Civilians 267 380 152 67 63 1 4 - 934 kilometer although the number of violent events heavily decreases from the center to the semi- periphery and increases again in the wider periphery. The other six graphs show different patterns. As the graphs of Mingora and Lashkar Gah indicate, the semi-periphery around these two cities is much more affected than the semi-peripheries of the first pattern, however, their numbers decrease from the center to the semi-periphery. Other patterns are provided by the graphs of Herat, Jalalabad, Kandahar and Quetta. The graphs of Herat and Quetta show a stark drop in the number of violent events from the center to the semi-periphery, but then they remain steadily. Within a radius of 5 km around Quetta there is an intensity of 113 events, of which a remarkable number of 79 cases represent violent acts against civilians. A similar pattern can be seen in the figures of the city Karachi which is Pakistan’s largest city and an important economic center. In sum, 70 violent events occured right in the city of which 78.6% of the cases represent violence against civilians. By contrast, the center of Jalalabad is relatively secure with a total number of 22 cases within 10 km in the two years, whereas the number of violent events increases from the semi-periphery to the wider periphery. A distinct exemplary can be seen in the graph of Kandahar, which is highly exposed to violence in the

35 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES 1 15 47 377 799 2360 5,633 1,174 9 3 1 96 416 874 197 104 - 9 5 3 2 75 67 28 - 1 2 77 70 14 27 484 - - 1 83 49 19 17 169 - 1 1 84 55 11 30 174 - - 2 95 55 10 21 220 - - - 18 29 106 523 179 - - - 46 22 39 109 292 - - 75 17 37 10 132 269 - - 5 26 23 38 133 410 - - 6 20 19 53 133 200 - - 5 40 31 46 142 181 - - 4 47 314 720 261 108 Table 4: Overview of violent events in the most aﬀected provinces in Afghanistan - 1 5 92 42 461 222 1,050 (No. 10) (14) (8) (20) (32) (16) (6) (17) (18) (31) Helmand Kandahar Ghazni Paktika Zabul Khost Uruzgan Farah Kabul Kunar Herat Kunduz Wardak Residual Total -Civilians -Police Forces -Other international forces -Hizb-e-Islami -Other group -Al-Qaeda -Forces of theUnited States 80 39 9 17 14 12 9 17 10 12 5 2 15 60 284 Frequency of violent events PercentageCasualties 19.5 13.3Actors -Armed forces 6of Afgh. 76 5.6 5.6 40-Taleban 5.6 27-Unidentiﬁed 4.6Militants 256 20 184 4.5-Tribal Militia 4 32 191Civilians 113 1 14 3.6 99 36 3.5 25 53 1 71 50 3.3 25 77 78 - 49 3.2 10 73 44 30 17.6 13 - 55 52 16 100 29 - 75 27 16 27 1 57 32 53 13 46 32 - 16 54 79 24 - 20 41 33 406 47 - 41 25 22 - 250 27 121 - 1388 15 830 - 17 102 - 541 1 4 -Battle-rel-ated deaths Combatant-ci- 958vilian ratio 1:0.09 459 1:0.56 1:0.28 141 1:0.11 1:0.6 180 1:0.38 384 1:0.18 1:0.52 194 1:0.33 1:0.46 246 1:0.4 0:16 344 1:0.15 165 1:0.29 119 1:0.26 120 414 58 677 4,459

36 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES center as well as to a lower extend in the semi-periphery. In almost all cases the numbers of violent events in the semi-periphery are lower than in the centers. Even if a city was not located right in a conflict zone, it attracted violence, such as Islamabad or Quetta. Logically, the intensity of violence dependet on the fact whether a city is located in a conflict zone or not. As opposed to centers outside the conflict zone, the semi-peripheries of cities, which are located in a highly affected region are also affected. But in all cases the intensity decreased in the immediate periphery and in most cases the intensity of violence increased once more within the wider periphery. The high concentration of violent events per square kilometer in the cities contradict the claim that the conflict is merely a rural phenomenon. Both capital cities suffered from violence although twice as many violent events occured in Kabul during the two years as in Islamabad. In the Afghan capital in 55% of the cases the Taleban acted as primary party. The burden of guaranteeing security in Kabul is equally shared between the governmental actors. In 15.8% of all cases the US forces were involved, in 31.3% other international forces, in 16.6% the Afghan forces and in 9.4% the police forces. In 2008, the responsibility for the district of Kabul was transferred from ISAF to the Afghan Security Forces, especially to the police. As part of the transition process the province of Kabul was not delivered until 2011, but the Taleban groups still have several strongholds in the bordering districts. In Islamabad 39 violent events occured in the years 2008 and 2009, of which in most cases the TTP and the governmental forces clashed. A similar degree of intensity can be seen in Rawalpindi, which is close to Islamabad. Looking at the absolute numbers of the violent events the highest intensities per square kilometer can be seen in and around the largest cities of the Pashtun areas, especially in Peshawar, Mingora, Kandahar, Lashkar Gah and Quetta, which are important local centers of the periphery. In the years following 2009 the number of violent events rapidly decreased in Kandahar and moved to the rural areas of the province with the result that the Afghan security forces assumed responsibility for Kandahar City in May 2012 (US Department of Defense 2012). The region around Kandahar is a traditional stronghold of the Afghan Taleban. During the 1990s the Taleban gained some support from the population because they enforced law and order and eliminated competing warlords. However, their terror regime mostly led to the disappearance of popular support. The major military bases of the governmental forces are also located between Kandahar and Farah (No. 6). Most cities within the conflict zone are protected by military checkpoints, which obviously had a low effect on the appearance of battles. Within a radius of 10 km around the central point of Kandahar 70% of the violent events were battles between military organized groups and 30% violence against civilians. In Jalalabad, which is a highly important city of the governmental forces because of its airport, 40% of the events were cases of violence against civilians. In controlled areas, which are highly protected by security forces, the type of events shifted from battles to violence on civilians and infrastructure, especially through hidden actions such as suicide attacks, improvised explosive devices (IEDs) or planting car bombs. In Quetta, which is believed to be a major stronghold of the Afghan Taleban in Pakistan, 62.3% (71 cases) of all violent events (within a radius of 10 km) were violent acts against civilians. The numbers of Mingora, which is the largest city in Swat District, reflected the stronghold of the Pakistani Taleban, who occupied the district and enforced a crude interpretation of Sharia law in 2007 and 2009. Initially the armed forces of the local government were not able

37 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES to defeat the Pakistani Taleban, so that the central government was asked for help. This led to the Operation Rah-e-Haq (2007) and Operation Rah-e-Rast (2009) by 3000-5000 Pakistani troops chasing the Pakistani Taleban away from the Swat District. Right now the city is heavily protected by the Pakistani army. The occupation of the Swat District by the Pakistani Taleban is one example indicating that not only poor economic circumstances can lead to violence but also many other factors play an essential role. As Rabia Zafar (2011) argues the stronghold of the Pakistani Taleban in the Swat District is due to ”a broken judicial system, poor governance, and an institutional vacuum”. As can be seen in the appendix (table 5) two cases (Rawalpindi and Quetta) have a moderate correlation. In further three cases (combatants in Islamabad, total numbers in Rawalpindi and combatants in Jalalabad) we can see small correlations, of which two are significant at a 0.01 level and one at 0.05. The test on civilian deaths does not show any correlation between the two variables. But the values of the significance also show that there are not enough cases of civilian death rates given. Considering the results of the inductive statistics, the following conclusions can be made: (1) The average number of deaths depends on the kind of warfare, but does not depend on the population intensity, because otherwise the average death rate would be higher in urban areas than in the wider periphery. (2) From a center to a wider periphery the average number of deaths almost doubled. But the correlation is not linear because the semi-peripheries show lower average death rates. (3) The cases of Islamabad and Rawalpindi indicate that major political centers in Pakistan attract violence even if they are not located in a conflict zone. But the average number of deaths decreased in the semi periphery and increased in the wider periphery. (4) The cities in a conflict zone do not show any strong correlation between the average number of deaths and distances. They are as much affected as the immediate and wider peripheries, however, the average number of deaths increased a little bit from the center to the periphery. This section pointed out one major problem of violence in Afghanistan and Pakistan, in particular the imbalances between the theoretical categories center and periphery. Governments are faced with serious problems to integrate the local administrative levels into the state and to give them a certain amount of autonomy at the same time. This problem can be seen in Afghanistan as well as in Pakistan. But the Pakistani state is much more robust than the Afghan one and can control its political power resources. One explanation for this is the fact that the population of the urban areas in Punjab and Sindh, which dominate the military, bureaucracy and the political parties, were much less affected by violence than the peripheral provinces. A higher intensity of violent conflict in the periphery means that people cannot conduct economic activities. As a consequence the periphery is less relevant to the economic performance of the state than to the strategic considerations of the political and military elite. Moreover, it shows the social and ethnic asymmetry between the local minorities in the periphery and the dominating majorities or at least pluralities in the urban areas. Another problematic aspect is that there is a lack of the central government’s accountability to provide security and public goods, such as economic development, public and private investments, infrastructure, healthcare and education, in order to acquire the political legitimacy. Some cases do not only show the government’s inability to secure the peripheries, but also the incapacity to protect the local centers. People feel insecure to practice political, economic, religious and cultural activities. Whereas the offensive raids of the military take place in the periphery, the police have to defend

38 5 ARMED CONFLICT IN CENTERS AND PERIPHERIES and secure the urban areas because attacks on cities and political centers mean a threat to a large concentration of people and to the power resources of the state. Good circumstances to lead a functioning state are to secure a large part of the territory, the political, economic and social capital and to gain popular support. On the side of the insurgents the latent and actual power is mainly located in the periphery and within its local centers. The periphery and its local centers include actual and latent power of the insurgents. Politically they mainly act in the underground through terrorist attacks, criminal activities, so-called shadow governors and secret networks.

(a) (b)

Figure 10: Violent events in and around major cities

39 6 THE DURAND LINE AND ARMED CONFLICT

6 The Durand Line and Armed Conﬂict

6.1 The Origins of the Border

The Durand Line, the 2,640 km long border between Afghanistan and Pakistan, came into existence when the British signed an agreement with Abdur Rahman Khan, the Amir of Afghanistan, after a war had led to a stalemate between them in 1892. The Mughal empire tried to demarcate its western frontier beyond Kabul and Kandahar because they knew that every attack from outside on India must start from Kabul and the British thought the same. The survival of the Amir of Afghanistan was intended by the British because they needed Afghanistan as a buffer zone between them and Russia. Nevertheless the British originally tried to establish their defence positions along the geographical axis Kandahar and Kabul. For the British the Amir needed to be friendly as well as a puppet, with the result that both negotiation parties had to make concessions. This led to the split of the tribal areas. The Amir got the strategically important Asmar and the Birmal tract of Waziristan, whereas the British acquired ”Dir, Swat, Peshawar and Chitral” (Omrani 2009: 177-195). The British concept of making the boundary to Afghanistan was similar to the situation in the north of British India, particularly in relation with China. In 1910, China controlled Tibet, which was lost again after the Chinese revolution. But the Chinese attempt concerning Tibet spured the British to intensify their forward policy. The British demarcated the MacMahon Line at the Simla conference in 1914, which is the actual line of control between India and China today. Just like Afghanistan did towards the czarist empire, South Tibet functioned as a buffer zone between British India and China. Actually the Durand Line included a triple border which comprised (1) the northern Afghan border to Russia’s sphere of influence, (2) the Durand Line as the military line of control and (3) an administrative boundary which was located somewhere between the Indus and Punjab (Gopalakrishnan 1982: 99). The British did not fully control the tribal areas between the Durand Line and Indus River. Instead, they deployed a few military bases and did not interfere in local policies within the tribal area, at least at the beginning. Admittedly the concept worked very well concerning the security towards Russia, but this policy caused internal problems with the tribes. When the tribal population reacted collectively to the British influence in 1897, India sent 60,000 troops to counter the insurgency. In order to handle the situation within the tribal belt, the British-Indian government instituted the North-West Frontier Province and executed several restrictive laws which regulated the interactions. Furthermore, they created a mezzanine between the British authority and the tribal groups, in particular the Maliks. The Maliks and other influental tribal leaders were paid by the British government to maintain order. If this method did not work the British officers were able to isolate them by cutting off the trading and communication routes. The Durand Line kept not only the British busy but also the Amir of Afghanistan. For the first time the Amir was able to enforce his law on a well-defined territory. He could implement permanent institutions that projected state power and provided him the capability to convert people to Islam (2009: 177-195). This led to insurgencies in the tribal areas. The Pashtun nationalist Abdul Ghaffar Khan fought with the congress movement for India’s independence and an independent Pashtunistan. In 1947, a popular vote was held to decide whether the Pashtuns should join India or Pakistan. But a separate Pashtunistan and the

40 6 THE DURAND LINE AND ARMED CONFLICT opportunity to choose Afghanistan were not included in the vote, and so about 50% of the population boycotted the referendum. As a result, 90% of the participants in the NWFP voted for joining Pakistan with the outcome that the Pashtuns form a plurality in Afghanistan, but twice as many Pashtuns live in Pakistan. The issue of the Pashtuns caused tensions between Pakistan and Afghanistan, such as Afghanistan’s refusal to vote for the admission of Pakistan into the United Nations. Aside from that, Afghanistan has not been accepting the Durand Line as the official border since the genesis of Pakistan. In 1960, several Pashtuns infiltrated some parts of the Pakistani territory through the border from Afghan tribes. Apart from the heights in the north-east and the desert in the west of Balochistan, several sectors of the border do not match with any physical landscape. Moreover, the boundary crosses through ethnic groups, such as the Pashtuns and Balochis. The local communities of the tribal areas were more integrative than the state with the result that the border did not separate social and cultural allegiances. The British were well aware of these problems, but the main reason why they were just interested in the tribal areas was that they wanted to control the Khyber Pass - the most important corridor through the Hindu Kush and the access to the Indian subcontinent from Central Asia. Originally, the Pashtuns were a marginal note for the British during the Great Game but they decided to join and challenge the British. The identity of the Pashtuns usually refers to tribalism but when they were ruled by foreigners they united and their identity tended to regionalism. Today a social or political unity between the Pashtuns does not exist. Instead of one threat from outside there are different regional and international states and several ethnic, religious, political and other criminal non-state actors.

6.2 The Border and the Taleban

Pakistan’s strategy towards Afghanistan and India is based on the idea of strategic depth. Within that thinking the border region has always been playing an important role. The Durand Line has several interesting characteristics because it crosses different peripheral areas, cultural groups and harsh landscapes. The northern sector is protected by the highest mountains on earth, the Sulaiman mountains and the desert of Balochistan which protects the southwestern sector of the border and, of course, the Hindu Kush. A few passes, such as the Khyber Pass or the Bolan Pass, canalize transportation routes and provide strategically important positions to control border crossings. In the northern parts of Pakistan the Durand Line lies in the south of the Wakhan Corridor, which connects Afghanistan with China. Along this border the Pakistani army deploys about 3,500 military posts. By contrast, the border to India is shaped by a flat area that does not enable an effective line of defense. On the other side of the border the positions of the Indian forces are offensively oriented towards Pakistan. The political events during the 1980s afford the opportunity for Pakistan to take the plan of strategic depth into action. This and the strategic disadvantage towards India led to the concept that the Durand Line could not only function as a line of defence towards the west and north but also towards India. The objective was that the Afghan border region and mountains should function as safe havens for Pakistani and Afghan militants if India overruns Pakistan. But it was never clear how to win back the Pakistani territory if the military could not defend it by itself. Pakistan does not have the conventional capabilities to cope with India’s dominating forces. Actually, it is not in India’s interest to overrun or annex Pakisten because they do not want to deal

41 6 THE DURAND LINE AND ARMED CONFLICT with another problem. The strategy of Pakistan involves an extensive support of local militant groups in order to gain their support if needed. But the militants should not only be used in case of an Indian attack but also as an important capability for asymmetric balancing. To the north the border functions as an impenetrable barrier to large scale invasions by conventional forces. By contrast, in the context of guerilla warfare the Durand Line is highly porous. During the Soviet occupation the border region was used as the center for the anti-Soviet jihad, from where the insurgency was organized and the combatants infiltrated Afghanistan. Several training camps were located at the border region, such as the Salman Farsi camp in the NWFP, Yuldeshev’s camp in South Waziristan, the Yawa camp in the Khost province or camps in the whole region around Peshawar and Muzaffarabad (Sareen 2005: 229). The Pakistani state, Arab networks and Western states primarily supported Islamist groups in the tribal areas, which became a threat to the traditional Pashtun tribes. One of the main pivots of the conflict zone has been FATA and its 800 km long border to Afghanistan, which has certain uncontrolled crossing points. Pakistan supported these Islamist groups because they act as a counterbalance to the ethno-nationalistic groups. The result was that Pakistan destabilized Afghanistan and drove a wedge between tribal elders, khans and maliks and mullahs, of which the latter increasingly consolidate their political power. Traditional jirgas were considerably replaced by ulema shuras, which were dominated by conservative Islamists (Johnson, Mason 2008: 41-77). Not only military camps were established in the Pakistani border region, but also madrassas, recruiting offices and refugee camps, where criminal networks dealt with militant groups and sold drugs, arms and kidnapped people. But not only Islamism and the activities of the Pakistani and Western governments can explain the rise of the Taleban but also the traditional checks and balances within the tribal areas. The border region is populated by the Pashtuns, who consist of about 350 tribes. These 350 tribes form five confederacies, in particular the Durrani, Ghilzai, Gurgusht, Karlanri and Sarbani Pashtuns, who can be further subdivided into clans (khels) and families (kahols). In the last 300 years the Ghilzai Pashtuns were mostly dominated by their Durrani rivals, who traditionally dominated the Afghan state. There have only been three Afghan governments dominated by Ghilzai Pashtuns since 1721, in particular the Hotaki Empire, which was established in 1721 and lasted until 1738, Nur Mohammed Taraki, who became the Afghan leader after the downfall of Daoud in 1978, and Mullah Omar, who became the leader of the Taleban and took over Kabul in 1996. The emergence of the Taleban, who are primarily dominated by Islamists of the Ghilzai and Gurgusht Pashtuns, was a product of the rivalry between the Ghilzai and the Durrani Pashtuns. When the Taleban fought for control over Kabul, they also had the intention to detach the Durrani tribes from power (Johnson, Mason 2007: 71-89). The Durand Line, which cuts the Pashtun area in half, is just a theoretical construct for the Pashtuns. Afghanistan has never accepted the border to Pakistan and the political, social and cultural ties between the tribes still exist. When Hekmatyar and his allies lost power during the 1990s, Pakistan provided support for the Taleban, which played an important role at the border either. The Taleban ordered the Pakistani government to close the camps of hostile Mujaheddin groups in order to control the border region. The loss of a specific border section meant the loss of an economic income by smuggling. During the rule of the Taleban between 1996 and 2001 Pakistan was not able to control them all the time. Several disputes emerged because the Taleban did not accept the Durand Line as the political border. Additionally, they refused to extradite criminals who

42 6 THE DURAND LINE AND ARMED CONFLICT escaped from Pakistan to Afghanistan. However, the Taleban raised the hope that the region at the Durand Line could be used for strategic depth and that the Taleban could guarantee a safe trade corridor to Central Asia. Instead, the porous boundary cost the Pakistani government millions of tax revenues due to smuggling (Wagner 2012: 163). After 9/11 and the fall of the Afghan government the Taleban used the Durand Line to recover their forces. In March 2002, Pakistan sent about 100,000 troops to secure the border and to assist the Special Operation Forces. However, various militants and Al-Qaeda activists could flee across the border. In 2002, first the Taleban had their strongholds in the south of Kandahar at the border region to Balochistan and Waziristan, from where they gradually expanded their line of control. Several infiltration routes crossed the border to Afghanistan, such as the boundary to Balochistan, Zabul, Waziristan, Kurram or to the Swat district. By 2006, the Taleban controlled the northern parts of Kandahar, Helmand, Nimruz and parts of the border region to Swat, Kabul and Paktika (Giustozzi 2009: 2-4). While the Taleban can easily move across the porous boundary, foreign forces are faced with the problem that the Pakistani government must permit operational border crossings. But the forces of the United States sometimes infringed the territorial integrity of Pakistan. A few unauthorized attacks, mostly from the skies, caused tensions between the United States and Pakistan. The first thematized border crossing of foreign troops into Pakistan happened in September 2008 when Navy Seals caught several militants although some crossings had occured there before. Another incident happened in 2010, when the Pakistani government shut down the Khyber Pass after friendly fire killed several Pakistani soldiers.

6.3 Violent Events and their Distances to the Border

As can be seen in figure 10 within the strongholds of the Taleban, such as Kandahar, Helmand and between Zabul and Kabul, the governmental side is quite fragmented, whereas the side of the non-governmental actors is only represented by the Taleban. Considering the Pakistani territory, the governmental side is relatively homogenous and mostly represented by the Pakistani security forces. The United States is the only governmental actor operating on both sides of the Durand Line, although, in the case of Pakistan they acted mainly in FATA. The pattern around FATA shows that the geographical dynamics of violence between the US and its oppo- nents coincides with the geographical reach of the social life. Within a distance of 120 km to the border the United States operated as a primary party in 7.3% of all cases. The armed forces of Pakistan had the highest amount of involvement with 30.9% within that area. The police forces have a higher involvment with 11.9% than the international forces with 11.3% and the Afghan forces with 5.7%. On the side of the non-governmental actors in 19.7% of the violent events within 120 km the Taleban were involved and in 19.7% the TTP. The statistics show a high percentage of unidentified militants with 34.37% of cases within 120 km. Large parts of the wider border region are highly affected. 718 violent events occured within a distance of 20 km to the border and 91 within 5 km. Violent events that are located less than 10 km to the border have an average death rate of 9.5 combatants and 4 civilians, compared to the average number of all events with 8 combatants and 4.1 civilians. Violent events with a lower proximity of 120 km to the border caused 8,441 combatants dead and 1,799 civilians. As can been seen in figure 11 the descriptive statistics partially confirm my hypothesis.

43 6 THE DURAND LINE AND ARMED CONFLICT

(a) Pakistan & Afghanistan (b) Pakistan

Figure 11: Number of Violent Events (y-axis) per Class measured from the Durand Line (x-axis; 1 class = Distance of 20 km)

(a) Governmental Actors (b) Non-governmental Actors

Figure 12: Violent events of political actors between 2007 and 2008

44 6 THE DURAND LINE AND ARMED CONFLICT

The numbers of Pakistan and Afghanistan combined and the figures of Pakistan show that the number of violent events per class (20 km) increased the closer one gets to the Durand Line. By contrast, the Afghan figure illustrates four peaks between 0 and 260 km with more than 180 violent events per class. In total 718 violent events occured within a distance of 20 km away from the border. 191 violent events with a lower distance than 10 km to the border caused 813 deaths. The inductive statistics does not show any strong correlation between distances and the average number of deaths per violent event. The figures of the combatants have very small correlations, but none of them has a correlation higher than +/-0.3. This high fragmentation of the governmental forces in Afghanistan implicates a difficult situation. Firstly, cooperation requires smart mechanisms of coordination, especially the coordination between different governmental forces horizontally and functionally different units vertically, such as stabilization and reconstruction teams and other civilian programs. This seems to be a difficult task but it is manageable. Secondly, the close cooperation between different security forces is often combined with the so-called green-on-blue attacks. These internal friendly attacks on governmental forces, which are mostly carried out by members of the Afghan security forces, are due to the dissatisfaction of Afghan soldiers over foreign involvement, infiltration and psychological pressure by Taleban fighters, anti-Islamic sentiments of some international soldiers and different incidents, such as Koran burnings or massacres on civilians by US soldiers (Long 2013: 167-182). According to Bill Roggio and Lisa Lundquist (2012) this kind of violence has starkly increased since 2008. The boundary gives the Taleban and other militant groups several advantages. Firstly, the border region is characterized by a harsh landscape, mountains, desert areas and small valleys, which make it impossible for both governments to attain full control over the border. Secondly, the characteristics and the location of the border region are arguments for Pakistan’s belief in the concept of strategic depth and for the intention to undermine India’s influence in Afghanistan. Instead, the Taleban got strategic depth beyond the border, from where they can reorganize, consolidate, recruit and train and from where they can infiltrate the bordering regions in Afghanistan. It improves the mobility of the insurgents and their ability to escape from the conflict zones in Afghanistan. Several military operations of the governmental coalition would require more freedom in moving across the border. But Pakistan frequently insists on its territorial integrity. Thirdly, the tribes in the region do not see the Durand Line as the legitimate border, because they have strong ties to tribes of the same or another confederacy on the other side of the boundary. To solve the problem of the porous border a promotion of Afghan nationalism and a massive manpower at the border region on the Pakistani side would be necessary. Both conditions are either not realistic nor promising. As the British diplomat Sir Hilary Synnott (2011: 117) stated, it ”is a problem best dealt with by management rather than by solution”. Practically the Durand Line is in combination with the characteristics of the landscape an open border. Islamist groups were increasingly able to fill the power vacuum in the border region although the offensive of the Pakistani government in FATA indicates that it took action to improve its position.

45 7 ARMED CONFLICT IN THE MOUNTAINOUS AREAS

7 Armed Conﬂict in the Mountainous Areas

7.1 The Geography of Mountains in Afghanistan and Pakistan

The most remarkable mountain range is represented by the Kindu Kush, which stretches about 800 km from north-east to south-west and at a maximum of 350 km from north-west to southeast. The east of the Hindu Kush has over 20 summits higher than 7,000 metres. The main line of the Hindu Kush in the north-east of Afghanistan has several mountain ranges which can be subdivided into several subranges, such as Koh-i-Baba, which reaches from the south of Bamyan to Helmand and Herat and the Safed Koh range, which is located in the south of the Khyber Pass and marks the border line. Safed Koh lies between the valley of Jalalabad and Kurram valley. Koh-i-Baba consists of a southern and northern range, of which the former is called Koh-i Saﬁd and the latter Koh-i Bubuk.

Figure 13: Mountain ranges

Both ranges are sources of main rivers in Afghanistan, particularly Hari Rud, Kabul River and Helmand, which enable agriculture in the lower areas. 95% of the demand for water is used for agriculture, such as the cultivation of poppy, apricots, apples or melons. The river Helmand is the main source for the Afghan poppy cultivation. The income from poppy trade therefore depends on the length of the dry period and alternates seasonally. Most agricultural areas also depend on canal systems of which the many are scarred by the longstanding conﬂict. The ranges Safed Koh and Koh-i Saﬁd are the major strongholds of the Taleban. Safed Koh was also the place of the well known battle of Tora Bora, which is a cave system constructed as a hideout for the Mujaheddin during the 1980s. Several Al-Qaeda members and militants were able to escape across the border. Another famous mountain stronghold is located at Zhawar,

46 7 ARMED CONFLICT IN THE MOUNTAINOUS AREAS located at the border between the ranges Toba Kakar and Safed Koh in Waziristan and which is known for the battles of Zhawar between the Mujaheddin and Soviets. Generally speaking, the mountains and climate determine the conflict season. The data show that the number of violent events per month in Afghanistan in 2008 continuously increased from 35 in January to 168 in October and then suddenly decreased to 58 violent events in December. A similar pattern can be seen in 2009, when the number of violent events increased from 87 in January to 167 in August and then declined to 78 in December. In Pakistan the monsoon season between the end of June and September seems to play a role in the conflict season. During this period the number of violent events increased or decreased to a large extend. But the data show that the Pakistani figures of violent events in the winter are much higher than in the winter months in Afghanistan. Several mountain passes are closed for some days during the winter, such as the Salang pass to the north of Kabul. There are also some passes that are closed for months, such as the Dorah Pass at an elevation of 4,300 metres or the Broghol pass at about 3,800 metres. Controlling these passes was often a strategic objective, not only for governmental actors but also for non- state actors. It can be assumed that some of the passes at the Durand Line are intensely used by the Afghan and Pakistani Taleban, such as the Khyber Pass between Peshawar and Jalalabad, the Dora Pass, which connects the District of Chitral with Khyber Pakhtunkhwa, the Gumal Pass in South Waziristan or the Khojak Pass between Quetta and Kandahar. For both sides the passes, mainly built by the Soviets, are important strategic areas because they provide access to outlying areas with motorized equipment. By contrast, the other side can easily attack supply trucks at passes. The Khyber Pass, which can be entered through FATA, is the major supply route for the international forces. Especially water, food or motorized equipment go through Pakistan. In 2008, the Khyber Pass was highly affected by several attacks of Taleben fighters on NATO convoys killing about 80 truck drivers. In 2010, the Taleban were able to blockade the Khyber Pass ten times. Between 40 and 70% of all supplies cross the Khyber Pass whereas the residual amount is imported by planes or through other streets (Imtiaz 2009: 4-6; Daily Times 2011). The alternative route goes over the Khojak Pass. In total, 80% of all supplies cross the passes at the Durand Line, whereas about 20% go through Uzbekistan and Tajikistan (Gohel 2011: Ev. 58-82). Known as the dependency syndrome, the international forces still depend on Pakistan’s supply routes and the state’s will to secure them. The shape of the landscape also affects the construction of the traffic system. The local centers are connected with Kabul by major roads, in particular the Asian Highway (A76) from Kabul to Mazar-i-Sharif, the A77 from Kabul through Bamian along the Hindu Kush to Herat, the highway to Kandahar, which crosses the A75 to Quetta, and the A1, which connects Kabul with Jalalabad and across the Khyber Pass with the political center of Pakistan. Together they form the so-called ring road system through Kabul and around the western part of the Hindu Kush. The Kabul-Kandahar highway, which is part of the ring road, lies on a flat plateau between the Hindu Kush and the ranges of Safed Koh and Toba Kakar. The 480 km long corridor, which was mainly financed by the international coalition, has been affected by the Taleban since 2003. Initially the highway should have improved the economic development and the transportation between the two largest cities Kabul and Kandahar, but the corridor has become one of the most dangerous zones of Afghanistan and should be avoided by foreigners. This plateau is an important strategic area for the Taleban. First, it is close to the Durand Line

47 7 ARMED CONFLICT IN THE MOUNTAINOUS AREAS and to the mountain strongholds of the Taleban in the north-west and south-east. Moreover, the Taleban installed several checkpoints to generate some income from transportation between Kabul and Kandahar and kidnapping. In addition to this, some parts are highly populated and compared to other parts of Afghanistan give little popular support.

7.2 Military Warfare and the Mountains of Afghanistan and Pakistan

Several authors wrote about guerilla conflict in mountains and refered to the Pashtun fighters in Afghanistan. As a matter of fact, the analysis of civil war in Afghanistan and Pakistan needs to focus the attention on rough terrain, especially the strategic relevance of the mountains. Deputy Secretary of Defense Paul Wolfowitz, an influential neoconservative serving for the Bush administration, believed that the technical superiority of the US forces would defeat the Taleban very quickly. The strategists intended a technologically dominated operation by air with some support from boots on the ground. But it turned out that the given technological equipment could not overcome such mountainous terrain and the capabilities of the Taleban (Bahmanyar 2004: 17). When the strategy shifted to counterinsurgency the Bush administration began to focus on manpower. However, a sophisticated artillery, manpower or a technological superiority is almost useless in such a rough environment. Nevertheless, the United States and its immense defense budget tried to create a series of technological innovations to compensate its problems with the Afghan terrain. In general, both sides have to deal with the same conditions and each side has specific assets and drawbacks. Both sides are concerned with the decreasing oxygen content in the upper mountains and both depend on the same weather conditions. Each side is faced with rugged and steep landscape which is quite exhausting for foot soldiers. Though the local fighters are used to the terrain and the usage of secret paths through the mountains. An important advantage for the Taleban given by the mountains are caves or the so-called karez systems. Some of these caves are rudimentary but they can also be highly complex with several attaching tunnels - naturally occuring or man-made. Several of them are primarly used as hideouts in the mountains, which store food, weapons, some kind of medics, whereas other tunnels are scooped as bunker in conflict zones. The conflict in Afghanistan and Pakistan also requires different kinds of tactical alignments. The equipment of the insurgents can easily be transported through the mountains because they do not use heavy artillery. Instead of large divisions of infantery they consist of a low number of foot soldiers, which enables them to move through the mountains. Mountains are used as hideouts, as starting points for offensive raids on villages and as escape routes. Insurgents use hit-and-run tactics to attack a governmental base and to escape again very quickly. Motorized units mainly go by pickup trucks or horses to get from one place to another. Dostum, for instance, who cooperated with the Special Operation Forces in 2001, is known for his horse cavalry. In several cases he and his forces conducted lightning attacks against Taleban camps on their horses, mostly from mountains and disappeared as fast as they had come (Williams 2013: 32). Concerning the air the Taleban lack helicopters and people who can pilot them. Instead, they get flexibility through their geographical dispersion of militants and hideouts and their knowledge about the terrain. As opposed to conventional warfare, there are no extensive movements of large divisions and artillery. Instead, small units consisting of seven to twelve foot soldiers patrol through the

48 7 ARMED CONFLICT IN THE MOUNTAINOUS AREAS mountains and villages. To overcome the geographical dispersion and flexibility of the militants the United States has technological superiority but sophisticated equipment is vulnerable to extreme weather conditions. The US military developed a system of interlinked bases to transfer information effectively within the whole country. Furthermore, ground soldiers are equipped with GPS instruments to orientate themselves in the valleys or on the mountains. Because of the night visions, which are mounted at all weapons, the US forces dominate the battles at night. At daytime it is much more difficult to detect combatants because the mountains provide a kind of camouflage and the small units of foot soldiers can easily hide behind rocks. The United States has deployed several interconnected air bases which can provide air support anytime. Air support was sometimes needed when military units come into an ambush because of narrow valleys or when they have to destroy karez systems. But in general, the air force play a lesser role because the Taleban do not have many sophisticated ground bases. Another example is the use of drones, which were created in order to dominate the sky and ground within narrow valleys and winding rocks. They enable long flight durations and can observe rough terrain from a safe distance. On the ground the soldiers use small drones to overview hilly landscape or areas behind a retaining wall around a house. US strategists distinguish between the terms physical and human terrain. For the physical terrain the United States has created detailed maps with the aid of satellite imagery. The maps of the human terrain are calculated by genealogical data, network analysis of tribes, clans and families and charts of economic interactions. These two spheres normally interact in a complex way and the objective is to get both spheres under control. Some claim that the Taleban would make the international forces kill civilians with the result that the US could not win over the human terrain (Junger 2010). There are some indications confirming this thesis. Whereas the Taleban are stronger in the physical terrain, the governmental forces have an advantage in the human terrain. However, the governmental forces as well as the Taleban have huge problems in both spheres. According to McChrystal (2013: 2-8), as time passed, every single commando increased its number of campaigns per night from one to six in order to prevent the regeneration of the Taleban. But the hideouts in the mountains as well as the border region in Pakistan enable the Taleban to disappear and to disperse, to regenerate, to get used to the attacks and to become stronger. The strategy of the insurgents has been working very well in the long term. They are locked in stalemate in which both are too strong to lose and too weak to win, while the United States is bleeding to death financially as well as ideologically (Dorte 2011: 451-472). The insurgents can celebrate the withdrawal of the international forces as a victory, just as they consider the withdrawal of the Soviet Union, and can lead to an enormous upturn. The point is that neither the dual use of civilian programs of the COIN strategy, the meticulously developed military tactics, nor the technological superiority of the US forces showed any positive outcome.

7.3 Violent Events on Mountains

Pakistan has an average elevation of 576 m above sea level compared with the average elevation of the violent events of 1052.1 m. By contrast, the average altitude of Afghanistan is about 1,200 m and the average altitude of the violent events is about 1,270 m. 50% of all events in Afghanistan and Pakistan took place between an elevation of about 1,000 m and 3,000 m.

49 7 ARMED CONFLICT IN THE MOUNTAINOUS AREAS

In Pakistan the highest event occured at an elevation of 3,477 m and the lowest at 4 m. The lowest altitude of the violent events in Afghanistan is 279 m, whereas the maximum height of the violent events lies at 4,157. There are some differences between the boxplots, but generally speaking, they are quite similar although Pakistan differs from Afghanistan concerning geology and elevation. 776 violent events took place below an altitude of 500 m. Between the elevation of 500 and 1,000 m, 1,498 violent events occured and a further 1,271 cases up to 1,500 m. Up to an altitude of 2,000 the number of events decreased to 760 violent events and a further 564 events between 2,000 and 2,500 m. But then the number of cases rapidly decreases to only 54 events between 2,500 m and 3,000 m and nine violent events higher than 3,000 m. Violent cases arising higher than the average altitude are more dispersed than the cases below the average altitude. The inductive statistics do not show any high correlations between the altitude and the average number of deaths per violent event. It shows a small negative correlation combined with a high significance between civilians and the average number of deaths in Afghanistan. The higher the elevation the lower the average death rates of civilians, which is due to the low population density in higher areas. Within a radius of 10 km of the crossing point at the Khyber Pass 41 violent events occured. 513 violent events have a shorter proximity than 60 km to the Khyber Pass. As can be noticed by the intensity of violence around the Khyber Pass, the strategic depth of the Taleban leads to the capability that they can disturb the supply routes of the governmental forces, but not vice versa. In order to control the human terrain a large number of battles are taking place in villages on a lower altitude - mostly with police involvement. Offensive operations in the mountains, such as heavily armed patrols, are carried out by the military. But the Taleban fighters mainly dominate physical terrain in mountainous areas and local centers in the periphery. Additionally, they were able to attack transportation routes which connect the center with the periphery and the local centers. The position of the Taleban within the human terrain is much weaker than within in the physical terrain of the mountains. However, controlling the physical terrain of a mountain top militarily is completely different from controlling and securing the human terrain. The governmental actors have their major strongholds in the capital cities, control their physical terrain and thus have a stronger position within the human terrain. Controlling the human terrain means to hold the better political position because population, legitimation and economic and cultural structures, which are concentrated in the centers, are important power resources for the governmental side. Thus the governmental forces can mobilize enough resources in order to run the offensive operations in the mountains and to secure the villages and cities. The interesting question is: Why were the insurgents able to resist militarly even though the governmental forces increased their number of offensive raids? The reason for this is the fact that the mountain strongholds and the save haven in Pakistan enable the insurgents to expand time and space.

50 8 DISCUSSION AND CONCLUSION

8 Discussion and Conclusion

The years 2008 and 2009 marked a new peak of violence in Afghanistan and Pakistan. Both countries were affected by a similar kind of violence but at the same time they differed from each other in terms of political conditions, stability, geographical dispersion, social structures and the relations between center and periphery. In both cases there were some local centers in the conflict zone which were as much affected as their immediate periphery. Although the US forces were able to throw the Taleban quickly out of Kabul, the COIN operations in reaction to the reemergence of the Taleban tremendously failed. The process of state and nation building under the Bush administration and the implementation of an elaborated post-war scenario failed to show positive outcomes (Pressman 2009: 149-179). Furthermore, both states still lack democratic institutions, a critical middle class and social systems. While the Afghan state has always shifted from an authoritarian regime to none at all by 2001, the Pakistani state was governed either by an authoritarian or democratic regime. But even the democratic ones were characterized by power politics, corruption and short-run tactical considerations. Some institutions, such as specific factions of the ISI or the military, gained too much influence within the state over time, however, there are some indications that the Pakistani government tries to regain back control over these groups. The Durand Line does not separate the political and cultural ties of the tribes and is highly porous for the Taleban and other militias, but it limits the geographical reach of the coalition forces. In combination with the rough terrain of the mountains it enables the insurgents to expand time and space in order to resist the technological superiority of the governmental forces. In addition to these domestic problems, both states have been faced with a highly complex regional and international environment in which contrary interests directly promote violence or maintain the political conditions leading to violence. One major problem has always been Pakistan’s support for the Afghan Taleban to gain strategic depth and the deployment of militants as a weapon against India, especially in Kashmir and Afghanistan. Support for Islamist militias as a balancing strategy depended on the relations with India and on the dominating groups within the executive branch and the security apparatus. However, the Pakistani strategy of asymmetric balancing during the 1980s and 1990s has failed tremendously and caused several problems for Pakistan. Right now the Pakistani government has to deal with these problems from the past. The Pakistani elites had to realize that the instability in Pakistan has something to do with their policy towards Afghanistan and India. Pakistan has a high potential to play an influential role within the regional environment, but if it wants to be an equal partner on the global stage, it has to modernize its state politically and economically in order to compete with India and China, to prevent poverty and to build and tax a critical middle class. Another question will be of how the Americans can deal with the activities of Pakistan and the rise of China, its increasing dominance in Asia and its thirst for energy from the Persian Gulf. If the United States decides to contain Pakistan, Pakistan will further fall into the hands of China. For the Chinese government a China friendly Pakistan means to have a better position towards the United States, India and the Persian Gulf. However, China does not simply provide weapons and aid, but is rather investing into the economy, building up infrastructure and improving its soft power within the Pakistani society. The United States will have to

51 8 DISCUSSION AND CONCLUSION cooperate with Pakistan on a political and intelligence level in order to deal with terrorism and nuclear security issues. Another important challenge will be the promotion of conditions that establish political, economic and social institutions by investing in the Pakistani market, by integrating it into the regional economy and by expediting trade with Pakistan. But most importantly the international community should promote the improvement of the Pakistan- India relations, which would also have positive effects on the relations with Afghanistan and asymmetric balancing. Concerning Afghanistan the international forces have to leave the country after a total failure. Additionally, it is not sure whether the Afghan forces can keep the Taleban down and provide stability. In the short-run America has to train and equip the Afghan military and has to hand the authority over the Afghan forces. The United States must consider that they will need a post-withdrawal scenario and someone who can support the Afghan government, but the United States has to find a compromise between costs and stability. It is likely that India will fill the power vacuum by supplying weapons and money. In the long-term the US foreign policy makers have to realize that they cannot involve themselves in domestic affairs. One cannot authorize hundreds of offensive drone attacks killing civilians and tell the Afghans how to organize themselves at the same time. The cases of Afghanistan and Iraq remind us of the fact that military forces are basically good at conquering or defending a territory but they are not an adequate instrument for social engineering. Other measures, such as intelligent operations, containment strategies, some targeted air strikes, multilateral approaches or diplomacy, would have more positive outcomes and much lower human and financial costs than the deployment of boots on the ground.

52 8 DISCUSSION AND CONCLUSION

Appendix

Table 5: Correlations between distances (<120 km) and average number of deaths

Combatants Civilians Total Combatants Civilians Total Islamabad (coef.) 0.333* -0.118 0.123 Kabul 0.137 0.207 -0.093 (Sign.) (0.013) (0.498) (0.258) (0.160) (0.256) (0.297) Rawalpindi 0.541** 0.173 0.381** Kandahar -0.122 0.082 -0.130 (0.002) (0.442) (0.006) (0.155) (0.607) (0.090) Peshawar 0.090* 0.197** 0.133** Lashkar Gah -0.123 0.327 -0.073 (0.029) (0.001) (0.000) (0.116) (0.171) (0.335) Mingora 0.124* -0.172* 0.058 Kunduz -0.250 -0.415 -0.221 (0.014) (0.019) (0.172) (0.168) (0.354) (0.189) Quetta 0.503* 0.103 0.196 Jalalabad 0.238** -0.113 0.175** (0.012) (0.828) (0.111) (0.000) (0.187) (0.000)

** Correlation is signiﬁcant at a 0.01 level * Correlation is signiﬁcant at a 0.05 level

(a) Islamabad (b) Peshawar

Figure 14: Scatterplots of distance (x) and the number of deaths per event (y)

53 8 DISCUSSION AND CONCLUSION

Table 6: Correlations between distances (<120 km) and average number of deaths

Combatants Civilians Total Distance to boarder -0.189** -0.31 -0.138** (0.000) (0.395) (0.000)

** Correlation is signiﬁcant at a 0.01 level * Correlation is signiﬁcant at a 0.05 level

Figure 15: Distance to the Durand Line and average number of deaths

Table 7: Correlations between altitude and average number of deaths

Combatants Civilians Total Altitude of 0.034 0.076* 0.062** all violent events (0.175) (0.036) (0.003) Altitude in 0.120** 0.110** 0.154** Pakistan 0.001 0.009 0.000** Altitude in -0.050 -0.261** -0.070* Afghanistan 0.175 0.000 0.036 Distance to Khyber -0.179** 0.085* -0.094** Pass (0.000) (0.019) (0.000)

** Correlation is signiﬁcant at a 0.01 level * Correlation is signiﬁcant at a 0.05 level

54 REFERENCES

(a) Boxplot of events per country (b) Altitude and average death rate in Pakistan

Figure 16: Violence in mountainous areas

References

Abbas, Hassan (2008): A Proﬁle of Tehrik-i-Taliban Pakistan; in: Combating Terrorism Center Sentinel, U.S. Military Academy, Vol. 1, No. 2 (Jan.), New York, pp. 1-3.

Abdul-Karim Khan (2006): The Pashtun Man; in: Men of the Global South, Adam Jones (Ed.), London, New York, pp. 368-371.

Abuza, Zachary (2003): Militant Islam in Southeast Asia. Crucible of Terror; London

Abuza, Zachary (2008): The social organization of Terror in Southeast Asia: the case of Jemaah Islamiyah; in: Countering the Financing of Terrorism, Thomas J. Biersteker, Sue E. Eckert, Oxon, New York, pp. 63-90.

Adeney Katharine (2008): Constitutional Design and the Political Salience of ”Community” Identity in Afghanistan: Prospects for the Emergence of Ethnic Conﬂicts in the Post-Taliban Era; in: Asian Survey, Vol. 48, No. 4 (July/Aug.), pp. 535-557.

Afghanistan Statistical Yearbook 2012-13 : Number of non-governmental organization local and international by provinces; (17.11.2013).

Akhtar, Nasreen (2008): Pakistan, Afghanistan, and the Taliban; in: International Journal on World Peace, Vol. 25, No. 4 (December), pp. 49-73.

55 REFERENCES

Akmal Hussain (2004): Institutions, Economic Structure and Poverty in Pakistan; in: South Asia Economic Journal, Vol. 5, No. 1, pp. 69-102.

Alavi, Hamza (2002): Pakistan between Afghanistan and India; in: Middle East Report, No. 222 (Spring, 2002), pp. 24-31.

Arreguín-Toft, Ivan (2005): How the Weak Win Wars. A Theory of Asymmetric Conﬂict; New York, Cambridge.

Bajoria, Jayshree (2011): Pakistan’s New Generation of Terrorists; Council on Foreign Re- lations, 13 May, (20.10.2013).

Bahmanyar, Mir (2004): Afghanistan Cave Complexes: 1979-2004. Mountain strongholds of the Mujahideen, Taliban & Al Qaeda; Oxford.

Barﬁeld, Thomas (2010): Afghanistan. A Cultural and Political History; Princeton.

Barﬁeld, Thomas (2011): Afghanistan’s Ethnic Puzzle. Decentralizing Power Before the U.S. Withdrawal; in: Foreign Aﬀairs, Vol. 90, No. 5, (Sept./Oct.), pp. 54-65.

Bearden, Milton (2001): Afghanistan. Graveyard of Empires; in: Foreign Aﬀairs, Vol. 80, No. 6 (Nov./Dec.), pp. 17-30.

Bodansky, Yossef (2001): Bin Laden. The Man who declared War on America; Roseville.

Buhaug, Halvard; Gates, Scott; Lujala, Päivi (2009): Geography, Rebel Capability, and the Duration of Civil Conﬂict; in: Journal of Conﬂict Resolution, Vol. 53, No. 4 (August), pp. 544-569.

Bureau of Investigative Journalism (2013): Casualty estimates Reported deaths and injuries, Pakistan 2004–2013, (28 Oct. 2013).

Christopher, Warren (2008): Secretary of State Warren Christopher, “Fighting Terrorism: Chal- lenges for the Peacemakers,” Address to the Washington Institute for Near East Policy An- nual Soref Symposium, May 21, 1996; in: Evolution of U.S. Counterterrorism Policy, Yonah Alexander, Michael Kraft (Eds.), Westport, pp. 130-139.

CIA The World Factbook (2014a): Afghanistan, Age Structure; (11.11.2013).

CIA The World Factbook (2014b): Pakistan, Age Structure; (11.11.2013).

Civil-Military Fusion Centre (2012): Afghanistan Provincial Indicators; (13.11.2013).

Daanish Mustafa; Brown, Katherine E. (2010a): Spaces of Permormative Politics and Terror in Pakistan; in: Environment, Politics and Development Working Paper Series, Department of Geography, King’s College London, No. 33, pp. 1-22.

56 REFERENCES

Daanish Mustafa; Brown, Katherine E. (2010b): The Taliban, Public Space, and Terror in Pakistan; in: Eurasian Geography and Economics, Vol. 51, No. 4, pp. 496-512.

Daily Times (2011): Tensions Flare Between U.S. and Pakistan After Strike, by Eric Schmitt, 28.11.

Dittmer, Lowell (2005): Conclusion. On the Current Status and Future Outlook of the South Asian Nuclear Security Embroilment; in: South Asia’s Nuclear Security Dilemma. India, Pakistan, and China, Lowell Dittmer (Ed.), New York, pp. 189-213.

Eikenberry, Karl W. (2013): Demographics; in: Afghanistan. Country Study Guide, Vol. 1, Strategic Information and Developments, International Business Publications (Ed.), Wash- ington D.C.

Fakhari A. Siddiqui (1998): Population Growth and Development Prospects for Pakistan; in: The Pakistan Development Review, Vol. 37, No. 4 (Winter), Part 2, pp. 557-574.

Fair, Christine C. (2010): Obama’s New “Af-Pak” Strategy: Can “Clear, Hold, Build, Transfer” Work? in: The Afghanistan Papers, The Centre for International Governance Innovation, No. 6, July, pp. 1-23.

Fearon, James D.; Laitin, David D. (1999): Weak States, Rough Terrain, and Large-Scale Ethnic Violence; Paper released at the Annual Meeting of the American Political Science Association in Atlanta.

Fluri, Jennifer (2009): ”Foreign Passports Only”: Geographies of (Post)Conﬂict Work in Kabul, Afghanistan; in: Annals of the Association of American Geographers, Vol. 99, No. 5, pp. 986- 994.

Freedman, Lawrence (2009): A Choice of Enemies. America Confronts the Middle East, London.

Gaier, Malte (2011): Von Ground Zero bis Abbottabad: Der Frontstaat Pakistan seit 9/11; in: Die Welt nach 9/11, Thomas Jäger (Ed.), Wiesbaden, pp. 173-187.

Ghufran, Nasreen (2009): Pushtun Ethnonationalism and the Taliban Insurgency in the North West Frontier Province of Pakistan; in: Asian Survey, Vol. 49, No. 6 (Nov./Dec.), pp. 1092- 1114.

Giustozzi, Antonio (2009): Koran, Kalashnikov, and Laptop: The Neo-Taliban Insurgency in Afghanistan; New York.

Gohel, Sajjan M. (2011): Afghanistan—Pakistan (AfPak): Britain’s Challenges, Dilemmas and Strategic Imperatives; in: The UK’s foreign policy approach to Afghanistan and Pakistan, Fourth Report of Session 2010–11, Vol. 1, House of Commons, Foreign Aﬀairs Committee, pp. Ev. 58-82.

Goldman, Minton F. (1984): Soviet Military Intervention in Afghanistan: Roots & Causes; in: Polity, Vol. 16, No. 3 (Spring), pp. 384-403.

Gopalakrishnan, Ramamoorthy (1982): The Geography and Politics of Afghanistan; New Delhi.

57 REFERENCES

Goodson, Larry P. (2001): Afghanistan’s Endless War. State Failure, Regional Politics, and the Rise of the Taliban; Washington.

Haqqani, Husain (2005): Pakistan. Between Mosque and Military; Washington D.C.

Hühnert, Dorte (2011): Der 11. September 2001 und Afghanistan - Ein strategisches Dilemma; in: Die Welt nach 9/11, Thomas Jäger (Ed.), Wiesbaden, pp. 451-472.

Imtiaz Ali (2009): NATO’s Khyber Lifeline; in: Terrorism Monitor, In-Depth Analysis of the War on Terror, Vol. 7, No. 1 (Jan.), pp. 4-6.

Jäger, David A.; Siddique, Zahra (2011): Are Drone Strikes Eﬀective in Afghanistan and Pak- istan? On the Dynamics of Violence between the United States and the Taliban; Discussion Paper No. 6262, Dec., Bonn.

Johnson, Thomas H.; Mason, M. Chris (2007): Understanding the Taliban and Insurgency in Afghanistan; in: Orbis, Vol. 51, No. 1 (Winter), pp. 71-89.

Johnson, Thomas H.; Mason, M. Chris (2008): No Sign until the Burst of Fire. Understanding the Pakistan-Afghanistan Frontier; in: International Security, Vol. 32, No. 4 (Spring), pp. 41-77.

Junger, Sebastian (2010): War: Ein Jahr im Krieg; München.

Long, Austin (2013): ”Green on Blue”: Insider Attacks in Afghanistan; in: Survival: Global Politics and Strategy, Vol. 55, No. 3 (June, July), pp. 167–182.

Lowther, Adam (2007): Americans and Asymmetric Conﬂict: Lebanon, Somalia, and Afghanistan; Westport.

Malik, Sadia Mariam (2011): An Empirical Investigation of the Relationship between Food Insecurity, Landlessness, and Violent Conﬂict in Pakistan; PIDE Working Papers, No. 71, Islamabad.

McChrystal, Stanley (2013): Generation Kill. A Conversation with Stanley McChrystal; in: Foreign Aﬀairs, Vol. 92, No. 2 (March/April), pp. 2-8.

Mielke, Katja; Schetter, Conrad (2013): Pakistan. Land der Extreme; München.

Milani, Mohsen (2010): Iran and Afghanistan; in: The Iran Primer. Power, Politics, and U.S. Policy, Robin Wright (Ed.), Washington, DC.

Mohiuddin, Yasmeen Niaz (2007): Pakistan. A Global Studies Handbook; Santa Barbara.

Nawaz, Shuja (2009): Pakistan’s Security Complex; in: Pakistan. Consequences of Deteriorat- ing Security in Afghanistan, Kristina Zetterlund (Ed.), Stockholm, pp. 15-31.

New York Times (2013): U.S. Disrupts Afghans’ Tack on Militants; by Matthew Rosen- berger, October 28, (29.10.2013).

58 REFERENCES

Noor, Farish A. (2009): The Uncertain Fate of Southeast Asian Students in the Madrasas of Pakistan; in: The Madrasa in Asia. Political Activism and Transnational Linkages, Farish A. Noor, Yoginder Sikand, Martin van Bruinessen (Eds), Amsterdam, pp. 141-168.

O’Loughlin, John; Witmer, Frank D. W.; Linke, Andrew M. Linke (2010): The Afghanistan- Pakistan Wars, 2008–2009: Micro-geographies, Conﬂict Diﬀusion, and Clusters of Violence; in: Eurasian Geography and Economics, Vol. 51, No. 4, pp. 437-471.

Odierno, Raymond T. (2012): The U.S. Army in a Time of Transition. Building a Flexible Force; in: Foreign Aﬀairs, Vol. 91, No. 3 (May, June), pp. 7-11.

Omrani, Bijan (2009): The Durand Line: History and Problems of the Afghan-Pakistan Border; in: Asian Aﬀairs, Vol. 40, No. 2, pp. 177-195.

Pakistan Bureau of Statistics (2011): Pakistan Statistical Yearbook 2011; (4.11.2013).

Pressman, Jeremy (2009): Power without Inﬂuence. The Bush Administration’s Foreign Policy Failure in the Middle East; in: International Security, Vol. 33, No. 4 (Spring), pp. 149-179.

Qassem, Ahmad Shayeq (2009): Afghanistan’s Political Stability: A Dream Unrealised; Surrey, Burlington.

Rashid, Ahmed (2011): Taliban. Afghanistans Gotteskämpfer und der neue Krieg am Hin- dukudsch; 2nd edition, Munich.

Reuveny, Rafael; Prakash, Aseem (1999): The Afghanistan war and the breakdown of the Soviet Union; in: Review of International Studies, Vol. 25, pp. 693–708.

Robinson, Linda (2012): The Future of Special Operations. Beyond Kill and Capture; in: Foreign Aﬀairs, Vol. 91, No. 6 (Nov./Dec.), pp. 110-122.

Roggio, Bill; Lundquist, Lisa (2012): Green-on-blue attacks in Afghanistan: the data; in: The Long War Journal, August 23; (15.10.2013).

Ruiz, Hiram A. (2005): Afghanistan and Refugees; in: Immigration and Asylum. From 1900 to the Present, Matthew J. Gibney, Randall Hansen (Ed.), Vol. 1, Santa Barbara.

Runion, Meredith L. (2007): The History of Afghanistan; Westport.

Sareen, Sushant (2005): The Jihad Factory: Pakistan’s Islamic Revolution in the Making; New Delhi.

Schubert, James N.; Stewart, Patrick A.; Curran, Margaret Ann (2002): A Deﬁning Presiden- tial Moment: 9/11 and the Rally Eﬀect; in: Political Psychology, Vol. 23, No. 3, Special Issue: 9/11 and its Aftermath: Perspectives from Political Psychology, (Sept.), pp. 559-583.

Shuja, Sharif (2007): Pakistan. Islam, Radicalism and the Army; in: International Journal on World Peace, Vol. 24, No. 2 (June), pp. 25-35.

Siddiqa, Ayesha (2007): Military Inc. Inside Pakistan’s Military Economy; London.

59 REFERENCES

Sinno, Abdulkader (2009): Explaining the Taliban’s Ability to Mobilize the Pashtuns; in: The Taliban and the Crisis of Afghanistan, Robert D. Crews, Amin Tarzi (Ed.), Harvard.

Stanford Law School; NYU School of Law (2012): Living Under Drones. Death, Injury, and Trauma to Civilians From US Drone Practices in Pakistan; September.

Synnott, Hilary (2011): Examination of Witnesses; in: The UK’s foreign policy approach to Afghanistan and Pakistan; Fourth Report of Session 2010–11, Volume I, House of Commons, Foreign Aﬀairs Committee, London, pp. Ev 106-119.

Taliaferro, Jeﬀrey W.; Lobell, Steven E.; Ripsman, Norrin M. (2010): Introduction: Neoclas- sical realism, the state and foreign policy; in Neoclassical Realism, The State, And Foreign Policy, Lobell, Ripsman, Taliaferro (Eds.), 2nd Edition, Cambridge, New York, pp. 1-42.

The Bureau of Investigative Journalism (2012): Drone Warfare. Revealed: US and Britain launched 1,200 drone strikes in recent wars; Chris Woods and Alice K Ross, 4.12., (13.10.2013).

Tomsen, Peter (2011): The Wars of Afghanistan. Messianic Terrorism, Tribal Conﬂicts, and the Failures of Great Powers; New York.

UCDP Department of Peace and Conﬂict Research: UCDP Battle-Related Deaths Dataset v.5-2013, 1989-2012; (15.10.2013).

UNICEF (2011): Afghanistan Statistics; (4.11.2013).

United Nations Data (2012): Pakistan. Total population, both sexes combined, Year 2020; (5.11.2013).

U.S. Department of Defense (2012): Report on Progress Toward Security and Stability in Afghanistan; Report to Congress, (Dec.), (14.11.2013).

U.S. Department of Defense (2013): Report on Progress Toward Security and Stability in Afghanistan; Report to Congress, Nov., (27.11.2013).

Wagner, Christian (2012): Brennpunkt Pakistan. Islamische Atommacht im 21. Jahrhundert; Berlin.

Weatherbee, Donald E. (2009): International Relations in Southeast Asia. The Struggle for Autonomy; Maryland, Plymouth.

Williams, Brian Glyn (2013): The Last Warlord: The Life and Legend of Dostum, the Afghan Warrior Who Led US Special Forces to Topple the Taliban Regime; Chicago.

World Bank Data (2012): Pakistan; (9.11.2013).

60 REFERENCES

Young, John W.; Kent, John (2013): International Relations since 1945. A global history; 2nd Edition, Oxford.

Zafar, Rabia (2011): Development and the Battle for Swat; in: al Nakhlah, The Fletcher School Online Journal for issues related to Southwest Asia and Islamic Civilization, Spring. (12.10.2013).

Zaidi, S. Akbar (2005): State, Military and Social Transition: Improbable Future of Democracy in Pakistan; in: Economic and Political Weekly, Vol. 40, No. 49 (Dec. 3-9), pp. 5173-5181.

61 LIST OF FIGURES

List of Figures

1 Theoretical model ...... 4 2 Network of the ISI and the Arab groups (Source: Tomsen 2011, p. 250) . . . . . 10 3 Most populated cities and the international environment of Afghanistan and Pakistan ...... 12 4 Battle-related Deaths in Pakistan (including Kashmir) and Afghanistan (1989- 2012) (Data: UCDP) ...... 19 5 Civil war between non-state actors (l) and violence against civilians (r) (Data: UCDP) ...... 21 6 Hotspots of violence in Afghanistan and Pakistan (Data: ACLED, 2008-2009) . . 30 7 Drone Attacks between 2008 and 2009 ...... 31 8 Violent events in and around major cities ...... 33 9 Violent events in and around major cities ...... 34 10 Violent events in and around major cities ...... 39 11 Number of Violent Events (y-axis) per Class measured from the Durand Line (x-axis; 1 class = Distance of 20 km) ...... 44 12 Violent events of political actors between 2007 and 2008 ...... 44 13 Mountain ranges ...... 46 14 Scatterplots of distance (x) and the number of deaths per event (y) ...... 53 15 Distance to the Durand Line and average number of deaths ...... 54 16 Violence in mountainous areas ...... 55