To weaponise or not to weaponise: Targeting water in civil war

(1)

TO WEAPONISE OR NOT TO WEAPONISE

TARGETING WATER IN CIVIL WAR

IRENE GROOT Master's Thesis

Spring 2020

Department of Peace and Conflict Research, Uppsala University Supervisor: Johan Brosché

Word Count:

22.515

(2)

(3)

Abstract

While the increasing importance of water security has been widely recognised, little is known about why and when water is likely to become the target of rebel groups in civil war. This study aims to contribute to this knowledge gap and asks the question: Why do non-state actors sometimes weaponise water? Based on bargaining theory and insights from terrorism literature, this study contends that military necessity and political costs constitute the dilemma that a rebel group faces in deciding whether or not to weaponise water. Hence, the more a rebel group is faced with a need to pressure the government, the more likely the rebel group is to target water systems when the political costs are low. Through a qualitative case study involving process tracing and structured focused comparison, this thesis explores the suggested hypothesis for the case of the FARC in the Colombian conflict by comparing two periods of peace negotiations during 1991-1992 and 1998- 2002. The findings suggest that political costs are a critical factor while military necessity seems of less importance in the decision to weaponise or not to weaponise. Future research is warranted and this study suggests several directions.

Keywords: Water weaponisation, water security, environmental security, civil war, FARC, Colombian conflict.

(4)

Acknowledgements

Writing this thesis has been a great journey and the generous advice and support of several people has been significant and indispensable. First and foremost, I would like to thank Johan Brosché who has been an excellent supervisor and whose guidance and support has been invaluable. Thanks to Pedro Enrique Valenzuela Grueso, Juan Diego, Oliver Kaplan and Jennifer Holmes, for taking the time to share their thoughts with me and for their guidance. Thanks also to, Daniela, Maxine and Hugo that have provided excellent feedback on earlier drafts. I also thank Afaf, Annika and Daan for the many great discussions and the wonderful company throughout the process. Final thanks go to my family for their unconditional love and support.

(5)

List of Abbreviations

CGSB Simón Bolivar Guerilla Front (Spanish: Guerrilla Coordinator Simon Bolivar) DMZ Demilitarized Zone

ELN National Liberation Army (Spanish: Ejército de Liberación Nacional)

FARC The Revolutionary Armed Forces of Colombia – People‘s Army (Spanish: Fuerzas Armadas Revolucionarias de Colombia - Ejército del Pueblo)

UP Patriotic Union (Spanish: Union Patriótica)

(9)

Table of figures

Figure 1 – Causal Mechanism... 13 Figure 2 - Attacks on water systems by FARC (Source: Water Conflict Chronology) ... 17 Figure 3 - Years of Peace Negotiations FARC and Colombian Government (Source: UCDP .... 18

(10)

1

1. Introduction

As the centre of life and with increasing scarcity, water is a strategic commodity whose military utility is on the rise. For much of human history, actors have utilised the strategic value of water to harm or adverse over an opponent. Typically by poisoning, denial, diversion, and its release by dangerous force, water has been a standard weapon of conflict.¹ Although the increasing importance of water security due to rising water scarcity has been recognized in recent literature, most research focuses on water as a potential source of conflict or cooperation, where only a few scholars have placed attention on the weaponisation of water during conflict.²

Given the importance, vulnerability and the increasing military utility of water,³ and simply because all states have water systems, it seems critical to understand the mechanisms that underlie the decision to weaponise water. As such, addressing this knowledge gap is not only academically interesting but even more so, it is imperative in order to secure civilians most vital needs. After all, unless we know why and when actors decide to weaponise water, one cannot develop effective policies to prevent such attacks.

In a novel study, Charlotte Grech-Madin observes that since 1946 the vast majority of state actors have refrained from the use of water as a weapon in international conflict. She provides a norm-based explanation and argues that water has become embedded in a global moral inhibition of state actors against its use.⁴ However, most weaponisation of water in the contemporary period is done by non- state actors, and surprisingly, this is a greatly understudied subfield. In intrastate conflict, Marcus King, Mark Zeitoun, Karim Eid-Sabbagh, and Jeremy Loveless show the weaponisation of water by

1 Charlotte Grech-Madin, ‗The Water Taboo: Restraining the Weaponisation of Water in International Conflict‘ (2020), 11.

2 Grech-Madin, ‗The Water Taboo‘; Jeannie L. Sowers, Erika Weinthal, and Neda Zawahri, ‗Targeting Environmental Infrastructures, International Law, and Civilians in the New Middle Eastern Wars‘, Security Dialogue 48, no. 5 (2017): 410–

430; Marcus DuBois King, ‗The Weaponization of Water in Syria and Iraq‘, The Washington Quarterly 38, no. 4 (2015):

153–169; Mark Zeitoun, Karim Eid-Sabbagh, and Jeremy Loveless, ‗The Analytical Framework of Water and Armed Conflict: A Focus on the 2006 Summer War between Israel and Lebanon‘, Disasters 38, no. 1 (2014): 22–44.

3 Grech-Madin, ‗The Water Taboo‘; Peter H. Gleick, ‗Water and Conflict: Fresh Water Resources and International Security‘, International Security 18, no. 1 (1993): 6–42, https://doi.org/10.2307/2539033.

4 Grech-Madin, ‗The Water Taboo‘, 252.

(11)

2

jihadist groups (IS-Iraq, IS-Syria, Al-Shabaab-Somalia, and Hezbollah-Israel). Yet, it remains to be studied why non-state actors, in particular those who claim to fight for the people, weaponise water during intrastate conflict.

During conflict, water systems can be harmed intentionally or unintentionally. While collateral damage to water is likely to occur in the chaos of war, what is more puzzling is the shift by non-state actors of apparent previous restraint of intentionality, towards deliberately attacking water systems. This thesis aims to contribute to the understudied field of water weaponisation, by focusing on the shift of non-state actors from previous restraint towards deliberately attacking water systems, and asks the question: Why do non-state actors sometimes weaponise water?

Based on bargaining theory and insights from terrorism literature, I argue that targeting water systems can be an advantageous strategy for a rebel group to pressure the government but ultimately entails high political costs. Accordingly, I claim that military necessity and political costs constitute the dilemma that a rebel group faces in deciding whether or not to weaponise water.

Hence, the more a rebel group is faced with a need to pressure the government, the more likely the rebel group is to target water systems when the political costs are low.

The Revolutionary Armed Forces of Colombia – People‘s Army (Spanish: Fuerzas Armadas Revolucionarias de Colombia - Ejército del Pueblo , FARC-EP, from now FARC) provides an interesting case for this study. In part, because the FARC claims to fight for the people, and because the conflict had been on-going for several decades (starting in 1964) but it was only in 2002 that the first attack on a water system by the FARC has been reported. Taken into account that the Colombian conflict has received a great deal of attention by scholars, it is interesting that the attacks on water systems by the FARC have not been the focus study. In fact, the novelty of this thesis becomes even more evident given that prominent scholars of the Colombian conflict reacted surprised by the very occurrence of these attacks.⁵ Compared to the limited previous research of water weaponisation, the case of the FARC is also arguably fairly different, as it is neither comparable to jihadist groups, nor follows the trend of state actors of increasing restraint, or a

‗water taboo‘, as observed by the study from Grech-Madin. In fact, the opposite seems to have occurred.

5 Personal communication Oliver Kaplan and Pedro Enrique Valenzuela Grueso

(12)

3

I empirically examine the proposed theoretical argument by conducting within- and between-time analyses of two peace negotiation attempts during 1991-1992 and 1998-2002, through process-tracing and a structured focused comparison.

The results indicate that partial support can be found for the hypothesis. The main variation has been found in the component of political costs, while variation in military necessity has been less clear and seems of less importance as a push factor for the weaponisation of water. Hence, the role of the military necessity component seems to have been of less importance in the dilemma whether or not to weaponise water. The findings tentatively suggest that decreasing or non-existent concerns of reputation loss are a critical factor for explaining a rebel group‘s shift towards deliberately targeting water systems.

The thesis proceeds as follows. Chapter 2 provides a conceptualisation of water weaponisation and reviews the previous research. In Chapter 3, the theoretical framework of the study is presented. Thereafter, I will turn to the research design of the study in Chapter 4. To situate the empirical analyses a brief introduction to the FARC and the Colombian conflict is provided in Chapter 5 and this is followed by the empirical analyses of the two peace negotiation processes in Chapters 6 and 7. Chapter 8 is devoted to the comparison of the two time periods, and here, I will also propose and extension of the argument, outline some alternative explanations and the limitations of the study. Finally, Chapter 9 provides the main conclusion of the study and outlines directions for future research.

(13)

4

2. Previous literature on water weaponisation

This chapter presents the previous literature on water weaponisation during conflict. It starts by outlining the conceptualisation of water weaponisation and follows by reviewing previous research.

Defining water weaponisation

Water weaponisation has only received limited scholarly attention in recent years and remains significantly understudied in the social science. Only a few scholars have sought to define water weaponisation, of which most recently Charlotte Grech-Madin⁶ building on the insights of other scholars, has provided the most elaborate efforts. I will follow her definition in this study. She refers to the weaponisation of water as the use of water as physical arms to harm and/or gain leverage over and adversary. This definition of water weaponisation acknowledges water as a distinct type of arms and follows the definition of the UCDP that denotes arms as any material means, e.g. manufactured weapons but also sticks, stones, fire, water, etc.⁷

Forms of water weaponisation

Water can be used as a weapon in different ways and for different objectives. Mark Zeitoun, Karim Eid-Sabbagh, and Jeremy Loveless⁸ develop a legal analytical framework and categorise the different forms of water subject to attack, including water infrastructure, water resources, water sector institutions and hydro political relations. Grech-Madin classifies the use of water weaponisation based on two distinct outcomes, namely water deprivation and water inundation. Where water deprivation refers to the reduction or complete denial of water quantity, or degradation of its quality (potability), needed for basic subsistence. Water inundation, on the other hand, denotes the rapid release of a large mass of water.⁹ An interesting remark made by Grech-Madin is that deprivation of water is not directly physical violence. Although fatalities would be the extreme outcome, this type of violence results in great suffering for the civilian population deprived of water through the prevention of the fulfilment of its

7 UCDP, ‗Definitions - Department of Peace and Conflict Research - Uppsala University, Sweden‘, 2020, https://www.pcr.uu.se/research/ucdp/definitions/.

8 Zeitoun, Eid-Sabbagh, and Loveless, ‗The Analytical Framework of Water and Armed Conflict‘.

(14)

5

most vital needs. For simplicity, and because this is not the variation of interest, this study takes the physical targets subject to water weaponisation together, and will from now on refer to these as water systems.

Objectives of water weaponisation

Turning to the objective of water as a weapon, Marcus King¹⁰ distinguished five categories of use, namely: strategic weaponisation, tactical weaponisation, psychological terrorism, extortion or incentivisation, and unintentional weaponisation. The provided descriptions of these categories, however, are fairly specific to the case of study and incidents have been assigned to multiple categories. Arguably, as Grech-Madin remarks, some of these categories could be merged. Following the wider distinction of strategic and tactical weaponry applied in security literature, Grech-Madin refers to strategic weaponisation of water as directed to fulfilling higher-order goals in conflict such as victory and unconditional surrender of the adversary. Water here is used to destroy important areas of land and population centres, typically on a large scale and with the expectation of decisive results.¹¹ Tactical weaponisation, on the other hand, aims to fulfil low-order goals in conflict, namely to degrade the fighting capacity of the adversary by damaging objects of primarily military value and disrupting military movement and supply lines.¹²

The distinction of strategic and tactical use is advantageous in terms of the fulfilment of high-order or low-order goals. However, the relevance of the level of intenseness in relation to the objective of water weaponisation in a context of intrastate conflict is questionable. In part because the weaponisation of water on ‗home territory‘ (although in varying degrees) inherently brings forth the risk of self-harm.¹³ The logic is as follows, rebel groups do not want to destroy critical water systems which they (will) need themselves, nor placing additional costs for reconstruction or social costs, and risk undermining the post-war legitimacy.¹⁴ Additionally, the level of harm done might be a consequence of the level of success of the attack and as such can lead to false inference regarding its objective. For example, an attack on a water aqueduct might have had the objective to destroy the

10 King, ‗The Weaponization of Water in Syria and Iraq‘.

12 Grech-Madin, 51.

13Grech-Madin provides an interesting discussion on self-harm (p.76) as a rationalist explanation for restrained, where she argues that water can be weaponised in a controlled manner in a way that minimises self-harm, deeming it an insufficient explanation for restraint for state-actors in international conflict.

14 Grech-Madin, ‗The Water Taboo‘.

(15)

6

aqueduct and cause the rapid release of water to flood an area. In case the attack turned out to be

‗unsuccessful‘ in this regard, and merely damaged the aqueduct, the empirical observation would be different, rather a paucity of water. The level of intenseness thus need not be connected to the objective of the attack on the water system in civil wars.

To summarise, this study will draw on previous research to recognise water weaponisation as a means to an end; a rational instrument to fulfil high or low-order goals. This view requires further elaboration on when harm to water during conflict is considered weaponisation.

Intentional and unintentional weaponisation of water

In the study of Zeitoun et al.,¹⁵ special attention is given to the distinction between discriminate and indiscriminate use. This typology is a little confusing, as indiscriminate violence in the context of civil war is often understood as the indiscriminate selection of victims as opposed to selective victimisation based on individual‘s actions.¹⁶ What is meant with indiscriminate water weaponisation by Zeitoun et al., reflects what Grech-Madin and King refer to as collateral damage, and denotes the distinction between intentional and unintentional water weaponisation posed by King. Although unintentional weaponisation might be seen as collateral damage, Grech-Madin argues that viewing water as a victim bypasses the military utility that is built into it.¹⁷ As such, what is to be considered weaponisation of water need not be contingent on intentionality, but requires reasonable foresight.¹⁸ Thus unintentional unforeseeable weaponisation is essentially a pure accident, and unintentional but obviously foreseeable damage to water is considered to be weaponisation as well.¹⁹

Whilst the discussion of what is considered weaponisation of water is a useful one, the variation of interest in this study concerns a shift towards the deliberately attacking of water systems.

As such the focus is on water becoming the direct target, rather than looking at water weaponisation based on the outcome. In other words, the variation of interest is the deliberately direct attacks on water systems, as opposed to indirect water weaponisation, where water is harmed through an attack on a non-water target.

15 Zeitoun, Eid-Sabbagh, and Loveless, ‗The Analytical Framework of Water and Armed Conflict‘.

16 Stathis N. Kalyvas, ‗The Paradox of Terrorism in Civil War‘, The Journal of Ethics 8, no. 1 (2004): 97.

18 Grech-Madin, 97.

19 Grech-Madin, 53.

(16)

7

Findings of previous research on water weaponisation and the apparent research gap

Previous research on water weaponisation has either focused on an apparent trend of restrained in the international community or on the use of water by jihadists groups in intrastate conflict. Grech- Madin observes that since 1946 the vast majority of state actors have refrained from the use of water as a weapon in international conflict, and offers a norm-based explanation of a moral inhibition of state actors against its use; a water taboo. While she does not seek to explain why actors weaponise water, the discussion of the waxing and waning of the water taboo‘s influence is a useful one.²⁰ She finds that when the taboo is not fully internalised, competing priorities in conflict place limits on its influence. Derived from the deviant case under study, the US in the Gulf War, she finds 3 key conditional variables that may limit the taboo‘s influence, namely: military necessity, field independence of the military vis-à-vis the political leadership, and the level of embeddedness in the international norm community.²¹

In intrastate conflict, King observes the use of water as weapon as the critical enabler for a strategy of territorial acquisition for the Islamic State. The findings of this study are, however, specific to the actor under study which limits the applicability to other groups.

To conclude, the literature on water weaponisation is scarce and mostly concerned with documenting the occurrence or explaining its absence. Hence, it remains to be studied why non- state actors sometimes weaponise water during internal conflict. In the following chapter a theoretical argument will be developed based on the provided conceptualisation of water weaponisation and will draw from neighbouring theories that seek to explain violence in conflict research.

20 Grech-Madin, 225.

(17)

8

3. The theoretical framework

This chapter presents the theoretical framework of the study. It starts with a discussion of bargaining theory in civil war and how it has been applied to explain violence against civilians and terrorism. Based on bargaining theory and the complementary insights from terrorism literature, I develop an argument to explore why water is weaponised by non-state actors.

Bargaining theory in civil war

One of the main theories in conflict research holds that an armed conflict is essentially a bargaining game to determine both sides‘ relative bargaining position and consequently the outcome of the conflict.²² All interaction, both fighting and negotiations, serve the objective to communicate respective capabilities and willingness to fight in order to influence the parties‘ expectations about the outcome of the conflict.

In internal conflicts, rebel groups that challenge the central power are often the weak contenders of the game. Some rebel groups can, however, pose a real threat to the state power which makes it reasonable to assume that although bargaining theory was originally developed to explain interstate relations, it also plays an important role in internal conflicts.²³ Rebel group is as such here defined as a non-state group that challenges the government with military means. Internal conflicts thus may be viewed as a bargaining game in which a rebel group challenges the government by making demands, and the government is to respond by either giving in and make concessions or deny this and fight to defeat the rebels.²⁴ In this view, knowing the relative capacity and willingness to fight is of most relevance for the government, as it will determine what the government is willing to settle for in the bargaining process. For the rebels, the key objective is consequently to signal the capability and willingness to fight; by imposing high costs on the government in order to convince

22 Thomas C. Schelling, ‗The Strategy of Conflict. Prospectus for a Reorientation of Game Theory‘, Journal of Conflict Resolution 2, no. 3 (1958): 203–264; James D. Fearon, ‗Separatist Wars, Partition, and World Order‘, Security Studies 13, no.

4 (1 July 2004): 394, https://doi.org/10.1080/09636410490945965.

23 Lisa Hultman, ‗Battle Losses and Rebel Violence: Raising the Costs for Fighting‘, Terrorism and Political Violence 19, no.

2 (6 April 2007): 205–22, https://doi.org/10.1080/09546550701246866.

24 Hultman.

(18)

9

the government that it is better off by settling the conflict through making concessions as the costs of continued fighting are too high.

Being the weak contender, rebel groups typically try to minimise direct confrontation, and often rather engage in sabotage and hit and run operations.²⁵ In the context of civil war, the resort to violence is as such often seen as synonymous to terrorism.²⁶ These tactics are fairly cost-efficient and presumed effective,²⁷ both in order to avoid losses and to maximise the rebel groups‘

advantage.²⁸ Attacking water systems is one such strategy that is relatively cheap and capable of imposing high costs and sending a strong signal to the government.

Water as an advantageous strategy

When a rebel group attacks water systems they impose costs on the government in several ways.

Firstly, the resulting deprivation (too little) or inundation (too much water) inflicts great suffering to the population with fatalities as the extreme outcome. Fatalities will most likely be the case for inundation, considering the consequences when entire areas are flooded. However, the suffering inflicted on the civilian population should not be underestimated. Since the government is ultimately responsible for the safety of the population and the provision of its vital needs, it will have to respond by regaining control, reconstructing the water systems, and providing immediate relief of the suffering of its population. Additionally, the suffering and fear of recurrence that these attacks impose will be detrimental to the patience of the population.²⁹ As a result, it will become more difficult for the government to govern the country and uphold control.³⁰ Moreover, it will also require the military protection of other potential water systems subject to attack. Since water systems

25 M.L.R. Smith, ‗―‗Strategy in an Age of ―Low-Intensity‖ Warfare: Why Clausewitz Is Still More Relevant Than His Critics,‘‖ in Isabelle Duyvesteyn and Jan Angstrom, Eds., Rethinking the Nature of War (London: Frank Cass, 2005), 49.‘, n.d.

26 Kalyvas, ‗The Paradox of Terrorism in Civil War‘, 99 Kalyvas refers to violence in order to achieve compliance.

However, strategic use of violence against civilians is an indirect strike against the state. .

27 Kalyvas, ‗The Paradox of Terrorism in Civil War‘.

28 Smith, ‗―‗Strategy in an Age of ―Low-Intensity‖ Warfare: Why Clausewitz Is Still More Relevant Than His Critics,‘‖ in Isabelle Duyvesteyn and Jan Angstrom, Eds., Rethinking the Nature of War (London: Frank Cass, 2005), 49.‘

29 Michael Horowitz and Dan Reiter, ‗When Does Aerial Bombing Work? Quantitative Empirical Tests, 1917-1999‘, Journal of Conflict Resolution 45, no. 2 (2001): 147–173.

30 Hultman, ‗Battle Losses and Rebel Violence‘, 209.

(19)

10

are vulnerable targets and found throughout the country, it is virtually impossible for a government to military secure all the water systems. As such, attacks on water systems disrupt and discredit the processes of the government,³¹ undermining its legitimacy. After all, a government that cannot provide for the security of its own people and their most vital needs fails in two of its fundamental tasks.

In the terrorism literature, scholars have placed special emphasis on terrorism as a strategy of signalling,³² where the symbolic nature of terrorism often aims to send a political message to a wider audience.³³ This is especially true for symbolic charged targets which may even primarily serve to signal to the opponent, compared to its military advantage.³⁴ Since water is at the centre of life, there is something more essential about water and damaging it, than is the case for other forms of civilian infrastructure (King qtd. in Grech-Madin).³⁵ The vitality of water makes it more sanctified, or as Grech-Madin says it well, the water taboo is not simply the sum of its parts, but holds a force of its own.³⁶There are only few resources as essential to human existence as water, crosscutting many civilian needs and societal functions. The direct link to life makes water a symbolic charged target, that, although it is a low-risk target and relatively cheap for a rebel group to attack, it sends a very strong signal.³⁷ An attack on water systems is thus not only military and politically costly for the government, it also conveys the rebel groups‘ commitment to the conflict; communicating a message of how far the rebel group is willing to go. The mere ability of the rebel group attacking water is by itself a signal that the group is capable and deserving of the desired concessions,³⁸ and likely to attract a lot of attention, both domestically and internationally.

31 Martha Crenshaw, ‗The Causes of Terrorism‘, Comparative Politics 13, no. 4 (1981): 379–399.

32 Andrew Kydd and Barbara F. Walter, ‗Sabotaging the Peace: The Politics of Extremist Violence‘, International Organization 56, no. 2 (2002): 263–296; Crenshaw, ‗The Causes of Terrorism‘.

33 Virginia Page Fortna, ‗Do Terrorists Win? Rebels‘ Use of Terrorism and Civil War Outcomes‘, International Organization 69, no. 3 (2015): 519–556; Crenshaw, ‗The Causes of Terrorism‘.

34 Johan Brosché et al., ‗Heritage under Attack: Motives for Targeting Cultural Property during Armed Conflict‘, International Journal of Heritage Studies 23, no. 3 (16 March 2017): 248–60,

https://doi.org/10.1080/13527258.2016.1261918.

37 Brosché et al., ‗Heritage under Attack‘.

38 Brosché et al.

(20)

11 Water: The weapon of the weak?

In principle, all rebel groups can weaponise water,³⁹ yet they do not do so at all times. This implies the existence of a window of opportunity and/or cost structure in place which influences a rebel group‘s decision on whether and when to attack water systems.⁴⁰ While I have argued that attacking water systems can be an advantageous strategy for a rebel group to raise the costs of fighting for the government and to send a strong signal of commitment, it is also likely to backfire.

For a rebel group that challenges state power in civil war, winning the hearts and minds of the population can be of great importance. Like other violence against civilians, the suffering inflicted on the population as a result of attacking water systems will not make the rebel group popular nor attract support from the population. Hence, just like it discredits and undermines the legitimacy of the government, so does it undermine the rebel groups‘ image of the ‗fighters and protectors of the people‘, and will be especially costly for the post-war legitimacy. In additions to high political costs, attacking water systems is also likely to inflict self-harm by the destruction of water systems that the group (will) need themselves and consequently places additional costs of reconstruction for once the war is won.

Since water weaponisation is here conceptualised as a means to an end; a rational instrument, it will only be employed if the rebel group expects to benefit from it. Accordingly, considering that targeting water systems is a double-edged sword which ultimately entails high political costs, it should not be the strategy of preference but more likely a strategy of need. More specifically, the decision to target water systems is likely driven by a need that outweighs the potential costs of the action.

A rebel group may experience an increased need to pressure the government when it perceives that challenging the state cannot be left to the future.⁴¹ A most basic reason for necessity is then when the rebel group perceives it is losing. Scholars of violence against civilians, although often looking at fatalities, have argued in the same vein that civilian victimisation is driven by perceived strategic necessity,⁴² or as a ‗weapon of the weak.‘⁴³ The increased pressure to prevail right

41 Crenshaw, ‗The Causes of Terrorism‘, 395.

42 Alexander B. Downes, ‗Desperate Times, Desperate Measures: The Causes of Civilian Victimization in War‘, International Security 30, no. 4 (2006): 170.

(21)

12

now might lure the group to consider alternative strategies which they might not consider otherwise.⁴⁴ The logic is that a rebel group can perceive it has no choice as it equates the decision of whether to attack water systems, to the choice between action to secure survival, or inaction leading to ‗the death of the resistance‘.⁴⁵

On a similar note, a rebel group might perceive that the conflict has become forgotten. As argued before, targeting water sends a strong signal of commitment to the conflict, and is likely to gain a lot of attention domestically and internationally. Publicity might thus be the goal to raise the salience of their struggle and to improve their relative bargaining position. Especially, when international attention is deemed required, the audience has become larger, more diverse and perhaps more accustomed to terrorism.⁴⁶ To attract the necessary attention, the rebel group might perceive it must go to extremes to shock its intended audience.⁴⁷ In this way, the strategy is deemed imperative as the group must signal its commitment by extreme measures.

Military necessity has also been identified in previous research on water weaponisation, as a conditional variable for the water taboo‘s influence on state-actors to refrain from water weaponisation. Meaning that actors who perceive targeting water as taboo might break this commitment in a situation of military necessity. Although Grech-Madin does not seek to explain why actors weaponise water and focuses on restraint, the discussion of the waxing and waning of the water taboo‘s influence is a useful one as it provides the other side of the coin.⁴⁸ However, two remarks should be made. Firstly, the component of internal beliefs for the acceptability of the strategy is relevant for explaining restraint but has become self-evident when investigating its occurrence. Second, the influence of norms on rebel groups differs fundamentally from how norms influence state-actors. States ratify treaties, where rebel groups do not. And while international law has been found to have restraining effects in international conflicts,⁴⁹ the effects of norms on

43 Hultman, ‗Battle Losses and Rebel Violence‘.

44 Downes, ‗Desperate Times, Desperate Measures‘.

45 Crenshaw, ‗The Causes of Terrorism‘, 388.

46 Crenshaw, ‗The Causes of Terrorism‘.

47 Crenshaw.

49 James D. Morrow, ‗When Do States Follow the Laws of War?‘, American Political Science Review 101, no. 3 (August 2007): 559–72, https://doi.org/10.1017/S000305540707027X.

(22)

13

shaping violence in civil conflicts among non-state actors is less clear.⁵⁰ Although there is potential for norms to restrain the behaviour of rebel groups, the external effect of norms is captured in the logic of consequence in terms of political costs as an instrumental reason related to concerns of legitimacy or reputation loss that may restrain a rebel group‘s behaviour.⁵¹

Consequently, we find that military necessity and the political costs of attacking water systems - as a strategy to pressure the government - may counter each other‘s effect. Simply put, since attacking water systems is a politically costly strategy, it may only be considered advantageous when there is an increased need to pressure the government and when the benefits of the strategy are believed to outweigh the costs. Thus, on the one hand, this means that military necessity might lead a group to value present outcomes over future ones, providing a push factor to deliberately attack water systems. On the other hand, this means that when the political costs at stake are low, there is little left that may restrain an actor from using the strategy. As such, these factors – military necessity and political costs - constitute the dilemma that a rebel group faces in deciding whether or not to weaponise water.

Based on this, we can formulate the hypothesis:

The more a rebel group is faced with a need to pressure the government, the more likely the rebel group is to target water systems when the political costs are low

50 Oliver Kaplan, ‗Nudging Armed Groups: How Civilians Transmit Norms of Protection‘, Stability: International Journal of Security and Development 2, no. 3 (2013).

51 Kaplan.

Military Necessity

Low Political Costs

Deliberate attacks on water systems

Figure 1 – Causal Mechanism

(23)

14

4. Research Design

In the previous chapter, it has been proposed that the need for a rebel group to pressure the government and the political costs at stake provides a starting point to examine why rebel groups attack water systems during internal conflicts. This chapter presents the research design used to explore this argument empirically. An open approach, by conducting a comparative case study is considered most fruitful in connecting the empirical data to the research question in light of the glaring research gap.⁵² For this study, the FARC has been selected for a within- and between-time analyses of two periods of official peace negotiation attempts between the FARC and the Colombian government. First, a within-case analysis to process trace and scrutinise the causal mechanism will be conducted. After, a cross-period comparison will be carried out to seek to identify key variables of covariance with the dependent variable. Together, the within- and between- case analyses will enable to enhance the understanding of water weaponisation by non-state actors towards the development of a causal story. This chapter proceeds as follows. First, the methodological approach will be outlined, after which the case selection strategy will be presented.

This is followed by an overview of how the empirical analysis is structured and ends with a discussion on material and sources.

Methodological approach

Given the limited knowledge of water weaponisation by non-state actors, this study is explorative in nature and seeks to enhance the understanding of water weaponisation towards the development of a theoretical argument. In chapter 2, a gap in the existing research on water weaponisation revealed the apparent puzzle of a shift for non-state actors towards the deliberately attacking of water systems during internal conflict. Based on bargaining theory, one of the key theories in conflict research, a hypothesis has been developed to explore whether viewing deliberate attacks on water systems as a strategy that rebel groups use to pressure the government can contribute to the understanding of water weaponisation by non-state actors towards theory development. Hence, the approach of this study is less concerned with testing existing theory, but rather can be seen as a starting point for the understanding of the phenomena and for the development of a theoretical argument. For theory development, it is important to identify relevant variables, as well as a causal

52 Robert K. Yin, Case Study Research, vol. 5, Applied Social Research Methods Series (SAGE, 2009).

(24)

15

pathway to the outcome variable of interest. This study takes a qualitative approach for its opportune ability to expose causal mechanisms.

Combining Within- and Between-case Analysis

Combining a within- and between-case analysis is favourable to the overarching aim of theory development, as it allows exploring how the proposed theoretical argument plays out in each of the periods and facilitates theory building within the case study. This thesis seeks to expose the causal leverage of the proposed theoretical argument and to suggest improvements to the theory. In order to do this, a first step is to seek to scrutinise the causal mechanism for which his study will conduct process tracing to identify a causal story for why non-state actors sometime weaponise water. The next step is to establish a general evaluation of the covariance and the structural differences with regards to when water is weaponised. It aims to expose relevant variables with regard to attacks on water systems. A suitable way to do this, especially given that little is known about the phenomena, is by comparing cases that are as similar as possible regarding the background conditions but show variation for the dependent variable.⁵³ This approach is generally referred to as a most-similar design. Consequently, a between time comparison of periods that reflect variation in the dependent variable are examined is beneficial as a most-similar case design. This makes it possible to control for the influence of several factors in the comparative part of this study which can be logically excluded as possible causes.⁵⁴ To compare the time periods, a structured and focused comparison method will be used. This is an advantageous method to evaluate the differences between the time periods. The method is structured by asking the same questions that reflect the purpose of the study for the time periods under study, which enables a systematic comparison of the cases and allows for the accumulation of findings. The method is focused in that only deals with certain aspects of the cases studied.⁵⁵

53 John Gerring, Case Study Research Principles and Practices (Leiden: Cambridge University Press, 2006).

54 Leanne C. Powner, Empirical Research and Writing: A Political Science Student’s Practical Guide (1 Oliver‘s Yard, 55 City Road London EC1Y 1SP: SAGE Publications, Inc., 2015), https://doi.org/10.4135/9781483395906.

55 Alexander L. George et al., Case Studies and Theory Development in the Social Sciences (mit Press, 2005).

(25)

16 Case Selection

The universe of cases concerns rebel groups, defined as a non-state group that challenges the government with military means in internal armed conflicts.⁵⁶ The selection of cases has been strategically; based on an observed variation on the dependent variable over time. To identify which rebel group to study, I have used data on water weaponisation (Water Conflict Chronology)⁵⁷ and armed conflict (UCDP/PRIO Armed Conflict Dataset).⁵⁸ The Pacific Institute‘s ―Water Conflict Chronology‖ has tracked and categorised events related to water and conflict from the beginning of humankind. The UCDP/PRIO Armed Conflict Dataset includes armed conflict in the period 1946 to the present. Although the threshold of 25 battle-related deaths for armed conflict is of less relevance for this study, the focus has been on rebel groups in civil war.

For this study, the FARC in the Colombian conflict has been selected as this case is particularly interesting to study. In part, because the FARC is very different from the Islamic State (the focus of previous studies), and also, the conflict had been on-going for several decades (starting in 1964) but it was only in 2002 that the first attacks on water systems by the FARC had been reported. While this shift is puzzling in itself, recent research⁵⁹ exposed an increasing trend of restraint concerning the weaponisation of water by state actors, hence in the case of the FARC the opposite seems to have occurred. The shift of the FARC towards deliberately attacking water systems in the contemporary period is thus an interesting case to study in a theoretical, as well as, a policy point of view.

To identify which periods to study, data on water weaponisation⁶⁰ (Water Conflict Chronology) and peace negotiations (UCDP Dyad Archive)⁶¹ have been used.

56 I use the terms internal conflict, intra-state conflict and civil war interchangeably referring to an armed conflict where a rebel group challenges the state power.

57 Pacific Institute, ‗Wolrd Water Conflict Chronology‘, World’s Water (blog), accessed 25 March 2020, https://www.worldwater.org/water-conflict/.

58 Peace Pesearch Institute Oslo (PRIO), ‗UCDP/PRIO Armed Conflict Dataset - PRIO‘, accessed 27 March 2020, https://www.prio.org/Data/Armed-Conflict/UCDP-PRIO/.

59 Grech-Madin, ‗The Water Taboo‘.

60 The Water Conflict Chronology categorises entries along the use of water based on trigger, casualty and weapon.

However, these definitions remain vague and entries classified as casualty were found to include direct attacks on water systems as well. The data on water weaponisation used here thus includes entries based on both casualty and weapon that reflect direct attacks on water systems.

(26)

17

Figure 2 - Attacks on water systems by FARC (Source: Water Conflict Chronology)

As shown in figure 1, attacks on water systems by the FARC in the Colombian conflict have been reported for the first time, and most often, in 2002.⁶² The starting point of this study has been that all interactions of the rebel group serve the objective to influence government response, of either making concessions or deny this and fight to defeat the rebels. As such, to identify relevant periods of comparison, an overview of yearly peace negotiations has been considered useful. As can be seen in the following figure, four official peace negotiation attempts between the FARC and the government have occurred according to the UCDP dyad archive.⁶³

61 ‗UCDP - Uppsala Conflict Data Program‘, accessed 2 April 2020, https://ucdp.uu.se/additionalinfo/623/4#1983dyadarchive.

62 The years prior to 1983 have been excluded for readability but show no reported events of water weaponisation.

63 ‗UCDP - Uppsala Conflict Data Program‘.

(27)

18

Figure 3 - Years of Peace Negotiations FARC and Colombian Government (Source: UCDP

As shown in these figures, attacks on water systems by the FARC mostly took place at times of peace negotiations (8 out of 11), and the first and nearly half of all attacks reported took place in 2002 (5 out of 11). Based on these observations, this study will compare the third period of peace negotiations that is commonly delineated from 1998 to 2002 and which includes the shift towards deliberately attacking water systems by the FARC, with the previous peace negotiations period of 1991 to 1992 for which attacks on water systems had been absent. Hence, both periods play out in the post-cold war era which is favourable for comparison. Although the later peace negotiation process from 2012 to 2016, as well includes deliberate attacks on water systems, these events are relatively rare (including 3 attacks) and concern a reoccurrence that is arguably largely a consequence of what has happened in the 2000s. For this reason, the later period has been excluded and a comparison of the second and the third attempt of peace negotiations are considered most fruitful to expose the key variables regarding the shift towards deliberately attacking water systems.

Structure of the Empirical Analysis

The first empirical chapter presents a general overview of the FARC and how the rebel group is positioned in the Colombian conflict. This is followed by two empirical chapters that each focus on a period of peace negotiations and will be structured as follows.

Each chapter will start with outlining the FARC's behaviour regarding water, which is followed by an overview of the peace negotiations to situate the rebel group‘s position, where after the same questions will be asked to both cases in order to grasp the empirics and to present them in a consistent manner. To steer the empirical analysis in the direction of the theoretical argument, the questions will be specific enough to capture the key variables of the theoretical argument and to allow for comparison, while at the same time, remain open enough to enable capturing different nuances of the time periods. A focused but open approach is considered beneficial with regard to

(28)

19

the novelty of this thesis and hence the limited existing knowledge on why rebel groups weaponise water. To recap the theoretical argument posed: when a rebel group is faced with an increased need to pressure the government, a rebel group is more likely to attack water systems.

To empirically grasp military necessity and the political costs of this strategy the following questions will steer the empirical analysis of each period of peace negotiations:

 Military necessity: Did the rebel group perceive it was losing? Did the group lose in terms of military capacity? Was the government particular strong? Did the group lose in terms of relative bargaining power? Did there seem to be a situation characterised by haste?

Measuring perceptions of military necessity or ‗losing‘, requires further elaboration as this is understandably difficult to empirically obtain, if possible at all. In an attempt to understand the FARC‘s perceived military necessity, the military position of the group will be investigated along the trajectory of its long-term strategic plan. This is considered fruitful since perceptions of losing are related to what a group desires to ‗win‘.

 Political costs: Did the group seem to have a clear political strategy? What was the image of the rebel group? Did the rebel group enjoy public support? Did the rebel group seem to prioritise its domestic image? Did the rebel group seem to prioritise its international image?

The political costs refer to concerns of legitimacy and a loss of reputation. To identify the political costs of attacking water systems it is considered useful to assess whether the group displays concerns about the political costs in terms of reputation and legitimacy, as well as, the current political costs at stake in terms of the rebel groups image.

Data collection

Data is gathered from a wide spectrum of primary, secondary and tertiary sources. Primary sources, such as interviews and official documents from the rebel group are of great value in understanding the rationales of the group for decision-making around the use of strategies, as well as the perception of military necessity. Complementing these primary sources with a wide spectrum of secondary and tertiary sources, such as academic material, reports and books devoted to the time

(29)

20

periods of study of the Colombian conflict, has enabled to scrutinise the developments on the ground, the position and image of the group during the conflict, as well as in relation to the government. These triangulation efforts have helped to achieve as high reliability as possible. As limited data is available regarding the attacks on water systems, this study has also drawn upon media sources and databases such as the Pacific Water Chronology and Factiva News Service.

Combining the primary sources with secondary and territory data provided a comprehensive insight to explore why non-state actors sometimes weaponise water.

(30)

21

5. The armed conflict between the FARC and the Colombian government

To situate chapters 6 and 7, which constitute the empirical foundation of this study, this chapter outlines some core characteristics of the Colombian conflict, the FARC and developments up until the first period of peace negotiations under comparison.

The Colombian conflict: an overview

The Colombian conflict has been one of the world‘s most protracted and violent conflicts. For more than half a century, Colombia has experienced a long-lasting armed conflict between the government and several guerrilla organisations,⁶⁴ resulting in over 200.000 fatalities, thousands of forced disappearances and kidnappings, and almost 7 million people displaced.⁶⁵ While some rebel groups disarmed through negotiated settlements in the 1990s, the FARC who has been the largest group and causing the most fatalities remained active.⁶⁶ A final peace agreement between the Colombian government and the FARC has been reached in late 2016.

The topography of Colombia and the history of fragmented state presence lead to unequal development in the country. Since independence from Spanish rule in 1810, Colombians have struggled with many forms of inequality including land, income, state services, access to political representation, and inequality before the law.⁶⁷ The ultimate cause of the deep-rooted violence has

64Other guerrilla organizations include the ELN (Ejército de Liberación Nacional; National Liberation Army), the EPL (Ejército Popular de Liberación, Popular Liberation Army), M19 (Movemiento 19 de Abril, 19th of April Movement).

Besides the government, fighting also occurred with the paramilitary organizations AUC (Autodefensas Unidas de Colombia, United Self-defence Forces of Colombia). In this study the focus is on the interaction between the FARC and the Colombian government.

65 ‗Innovations in the Colombian Peace Process | Conciliation Resources‘, accessed 31 March 2020, https://www.c- r.org/resource/innovations-colombian-peace-process.

66 ‗UCDP - Uppsala Conflict Data Program‘, accessed 31 March 2020, https://ucdp.uu.se/conflict/289; ‗Innovations in the Colombian Peace Process | Conciliation Resources‘.

67 Ana Arjona, Rebelocracy: Social Order in the Colombian Civil War, Cambridge Studies in Comparative Politics (New York, NY: Cambridge University Press, 2016), 88.

(31)

22

been pointed out by scholars to have its origins in political exclusion.⁶⁸ Against this background, several guerrilla groups with a communist agenda emerged, including the FARC.

The origins of the FARC

The origins of the FARC go back to rural self-defence groups that had grown out of the tradition of the struggle for land. While initially defensive, these organisations became more offensive, and developed close ties with the Communist Party, during a decade of large-scale political violence known as La Violencia (1948 – 1958).⁶⁹ During this period, the Conservatives and the Liberals; the two dominant political parties in Colombia fought each other. The violence commenced after the assassination of the Liberal Party leader, Jorge Eliécer Gaitián on April 9 in 1948, and intensified after a Conservative victory in 1949.⁷⁰ At this time, the Communist Party urged Colombians to defend themselves.⁷¹

In the years that followed, the Conservative government attempted to exclude the Liberals from state power and used the army and police to repress any opposition.⁷² In 1964, the U.S.–

backed Conservative government then launched an offensive against rural self-defence groups, motivated as a campaign against a communist threat,⁷³ to curb the rise of ‗independent republics‘ in some of Colombia‘s most remote areas. Instead, this sparked the unification of the insurgents who later became the FARC.⁷⁴ The rebels regrouped as the Southern Bloc Guerilla (Bloque Guillero Sur) and held its first Guerrilla Conference in July 1965. At the second conference, a year later, the group reorganised as the FARC and established an outspoken agenda to seize power from capitalists and

68 Ana Bejarano, ‗Protracted Conflict, Multiple Protagonists, and Staggered Negotiations: Colombia, 1982-2002‘, Canadian Journal of Latin American and Caribbean Studies 28, no. 55/6 (2003): 225.

69 Claire Metelits, Inside Insurgency: Violence, Civilians, and Revolutionary Group Behavior, 2010; Harvey F. Kline, Chronicle of a Failure Foretold: The Peace Process of Colombian President Andrés Pastrana (Tuscaloosa: University of Alabama Press, 2007).

70 Angel Rabasa and Peter Chalk, Colombian Labyrinth: The Synergy of Drugs and Insurgency and Its Implications for Regional Stability (Rand Corporation, 2001).

71 Kline, Chronicle of a Failure Foretold.

72 ‗UCDP - Uppsala Conflict Data Program‘, accessed 27 March 2020, https://ucdp.uu.se/actor/743; Metelits, Inside Insurgency; ‗Innovations in the Colombian Peace Process | Conciliation Resources‘.

73 ‗UCDP - Uppsala Conflict Data Program‘; Kline, Chronicle of a Failure Foretold.

74 Kline, Chronicle of a Failure Foretold; I. William Zartman, ed., Elusive Peace: Negotiating an End to Civil Wars (Washington, D.C: Brookings Institution, 1995), 270.

(32)

23

direct the country according to Marxist theories, with the backing of the Communist Party and Manuel Marulanda and Jacobo Arenas as key leaders.⁷⁵

Earlier developments

In addition to the support of the national Communist Party, Cuba and the Soviet Union also played important roles in the early support of the FARC during the 1960s and 1970s. Towards the 1980s, the FARC had found other ways of income, mainly from extortion and kidnapping ransoms,⁷⁶ and extended their influence both on the countryside and in urban areas. While the FARC had been pushed by the military into the jungles of Guaviara, Caquetá and Putumayo, these were the regions of coca production and would serve as a substantial base of funding to the FARC.⁷⁷ Although the fall of the Soviet Union would be challenging for the FARC as it lost part of its funding and questioned the validity of the group‘s ideology, the group‘s new funding activities increasingly empowered the group and enabled them to recruit and arm more fighters on their own.

In the 1980s, local opposition to the guerrillas would develop in form of illegal self-defence forces known as the paramilitaries, which worked with the Colombian armed forces to fight against the guerrillas and in particular the FARC.⁷⁸ The impact of the paramilitaries would continue to grow in the 1980s, and even more so in 1990s, and has been extensive on the Colombian conflict. The paramilitaries have reportedly killed more than 2 times as many civilians in attacks than the guerrilla groups. Scholars have emphasised the role of the paramilitaries for the transformation of the FARC from a peasant-based rebel group that became extremely violent to the population it previously had claimed to protect.⁷⁹

75 Rabasa and Chalk, Colombian Labyrinth.

76 Kline, Chronicle of a Failure Foretold, 172.

77 Kline, 187.

78 Arjona, Rebelocracy, 90; Metelits, Inside Insurgency, 102.

79 Metelits, Inside Insurgency.

(33)

24

6. The times of the negotiation process of 1991 – 1992

At the times of the peace negotiation process of 1991-1992, the FARC had experienced continuous growth since the early 1980s and was militarily thriving. Following the developments in the mid- 1980s, where the FARC had explored the political path in the midst of a cease-fire that would lead the group to pause its infamous kidnappings, the group enjoyed a remarkable surge of national popularity resulting in tremendous electoral success in 1986. This trend of increasing popularity would continue into the early years of the 1990s. In this context, the group had been in a process of using less, not more, extreme tactics. However, developments at the end of the decade would compromise the viability of a political solution for the FARC, and the group had participated in the negotiations of 1991-1992 as a means to serve the group‘s long-term military plan along which it believed it was able to win the war. In the following sections, these developments are further discussed, starting with the FARC's earlier activities that have affected water, after which the times of negotiation process will be outlined, and followed by a discussion of the FARC's military and political developments.

6.1 FARC and harm to water

At the times of the negotiation process of 1991-1992, there are no reported attacks on water systems by the FARC. However, during this time, the group has not been an active protector of water either.

During the early 1990s, although it has mainly been a strategy of the ELN, the FARC has attacked oil pipelines. The resulting oil spill of these attacks has severely contaminated soil, forest and rivers.⁸⁰ In addition, the FARC has also attacked energy infrastructure causing power outages that indirectly inflict a paucity of drinking water, as water pumping activities become halted.⁸¹ Hence, although the FARC did not directly attack water systems, it did engage in activities to undermine the ‗capitalist‘

government that indirectly affected water systems.

80 Stéphanie Lavaux, ‗Natural Resources and Conflict in Colombia: Complex Dynamics, Narrow Relationships‘, International Journal 62, no. 1 (2006): 19–30, https://doi.org/10.2307/40204242.

81 Zartman, Elusive Peace, 286.

(34)

25 6.2 The Negotiation Process of 1991 - 1992

In the late 1980s, the level of violence and the intensity of the armed conflict increased unprecedented. While the total number of attacks by armed groups, including homicides, kidnappings, sabotage, hijacking, and attacks to infrastructure grew steadily, clashes between the guerrillas, paramilitaries and the Colombian armed forces occurred more often and became more deadly.⁸² Violence escalated even more, as, at the same time, the incumbent President Barco (1986- 1990) stepped up the war against the drug traffickers.⁸³ The violence dissolved Barco‘s effort to address the Colombian crisis, and lead the government into new bargaining with several guerrilla groups in the last two years of his administration.⁸⁴ Prior to the elections of 1990, César Gaviria called for national reforms including a new constitutional mandate to open up the political system and he later won the presidency.⁸⁵ As such, Gaviria offered the guerrillas inclusion in drafting the new constitution in exchange for their disarmament. Accordingly, Gaviria invoked the National Constituent Assembly that was to assist the creation of the constitutional change, and its members were to be chosen in a national election in December 1990.

Meanwhile, the FARC had explored the political path in the mid-1980s and gained tremendous electoral success. However, this success was followed by numerous assassinations of the FRAC‘s party members for which it held the government accountable. After the event, the FARC no longer considered national politics no longer a viable option. In September 1987, to address the asymmetry with the government, the FARC had joined the rebel groups of the National Guerrilla Coordination Front, which together created the Simón Bolivar Guerilla Front (CGSB).⁸⁶ While the government pressed on negotiations with the CGSB, the group had no interest in a negotiated solution. Arguably to curb potential political cost by a rejection of a peace process, the group participated but attached unacceptable conditions for the government to the negotiations.⁸⁷ Although the negotiations in 1990 did lead several guerrilla groups⁸⁸ to demobilize, the FARC emerged as the strongest guerrilla group.

82 Arjona, Rebelocracy, 91.

83 Zartman, Elusive Peace, 277.

84 Zartman, 278.

85 Zartman, 280–83.

86 Zartman, 277.

87 Zartman, 279.

88 These include EPL and the M-19

To weaponise or not to weaponise: Targeting water in civil war