Ending the never-ending? : community organization and civilian protection strategies to prevent violence in urban Mexico

Ending the Never-Ending?

C

TO PREVENT VIOLENCE IN URBAN

M

A thesis submitted by

S

A

11759542

in partial fulfillment of the requirements for the degree Master of Science in Political Science, International Relations

22JUNE 2018

SUPERVISOR DR JANA KRAUSE

A

Current literature on civilian protection in civil wars emphasizes the role of civilian agency. Civilians have the ability to respond to threats posed by an armed conflict: Consequently, their preferences and strategic alignments can shape the behavior of armed groups, and the conflict dynamics.

In Mexico, the prolonged armed conflict has implied that civilians are subject to high-impact violence that is motivated by economic gain. The violent conflict that involves multiple non-state and state actors fighting for power, control and money, has implied that civilians are targeted indirectly and directly, leading to an escalated civilian victimization.

Plethora of threats that civilians have encountered in urban neighborhoods has called for collective solutions. Building on empirical evidence, this research describes in which ways civilians protect themselves, and what protection strategies they use to prevent violence. It suggests that civilians involved in a violent conflict create strategies to protect themselves and respond to challenges posed by the encompassing violence through community organization and collective action.

Communities implement merely complementary protection strategies in three ways: through collective avoidance, subtle and hostile internal strategies, and by allying with external actors. Based on interviews and participant observation in urban municipalities of Ecatepec and Nezahualcóyotl in the State of Mexico, I argue that communities inherently look for subtle and non-violent alternatives to protect themselves. However, institutional consolidation has led communities to complement them with non-violent collaborative alternatives with the state actor. On the contrary, institutional weakness and the absence of authority has contributed communities to rely on violent or illegal alternatives to confront or cooperate with armed actors.

Studying civilian protection strategies and communities’ preferences to create protection amidst a violent conflict is important for future debates on civilian protection and public security. Communities have the ability to prevent violence and bring security. However, their strategic alignments may also imply more violence, consequently shaping short-term and long-term results of conflict resolution and peacebuilding.

T

C

ACKNOWLEDGEMENTS ... 5

LIST OF ABBREVIATIONS ... 6

TABLES AND FIGURES ... 7

MAPS ... 7

CHAPTER ONE:INTRODUCTION ... 8

CHAPTER TWO:RESEARCH DESIGN ... 11

2.1. Methodology ... 11

2.2. Case selection ... 11

2.3. Fieldwork and data gathering ... 12

2.3.1. Interviews and participant observation... 13

2.4. Ethical discussion and limitations... 14

CHAPTER THREE:LITERATURE REVIEW ... 17

3.1. Civilian protection in armed conflict ... 17

3.2. Violence against civilians ... 20

3.3. Armed conflict in Mexico ... 23

3.4. Approaches to civilian protection in urban Mexico ... 24

CHAPTER FOUR:THEORETICAL FRAMEWORK AND CONCEPTUALIZATION ... 26

4.1. Civilian decision-making in civil war... 26

4.2. Civilian strategic responses in an armed conflict ... 26

4.3. Collective action and community organization ... 28

4.4. Armed groups’ sensitivities and economically motivated violence... 29

4.5. Conceptualization ... 30

4.5.1. Civil war, organized crime, and violent conflict ... 30

4.5.2. Violence ... 31

4.5.3. Civilians ... 31

CHAPTER FIVE:BACKGROUND... 33

5.1. Spiraled violence, failed protection ... 33

5.2. Military in public security ... 34

5.3. Institutional weakness in the democratic era ... 37

CHAPTER SIX:CASE SELECTION ... 42

6.1. The State of Mexico ... 42

6.2. Ecatepec ... 43

6.3. Nezahualcóyotl ... 45

CHAPTER SEVEN:ANALYSIS ... 48

7.1. Avoidance ... 48 7.1.1. Collective avoidance ... 48 7.1.2. Failed organization ... 50 7.2. Internal strategies ... 50 7.2.1. Information sharing ... 50 7.2.2. Preventive measures ... 51 7.2.3. Warning systems ... 52 7.2.4. Protest ... 54 7.2.5. (Un)armed confrontation ... 56 7.2.6. Lynching ... 57

7.3. Allying with external actors ... 60

7.3.1. Cooperation with organized crime groups ... 61

7.3.2. Paying for protection ... 62

7.3.3. Collaboration with local authorities ... 64

7.3.4. Young adults as community peace promoters ... 67

7.4. Findings... 67

7.5. Discussion ... 71

CONCLUSION ... 73

BIBLIOGRAPHY... 75

APPENDIX ... 83

Appendix 1. Table of Interviews during fieldwork ... 83

A

This master’s thesis is a result of the dedication, determination, and devotion of extraordinary people. First, I would like to express my profound gratitude to my thesis supervisor Dr Jana Krause for her guidance and unconditional support, for trusting me to take on this ambitious project, and for maintaining close contact, while giving me the liberty to explore. I am thankful for Dr Imke Harbers for her interest in my research and for accepting the role of second reader.

I appreciate the attention of Dr Wil G. Pansters, Dr José Luis Cisneros, and Dr Daniel Cunjama López who patiently listened, provided valuable comments, and encouraged me to continue toward the right direction.

I thank my family and friends – the unbreakable safety net – for their (un)imaginable assistance, and more importantly, for their comfort, encouragement, strength, and love during each step of the process: My mother, Seija Anton, for each day that you have accompanied me with love and my father, Mukram Anton, for giving me courage: you two are my underlying source of endless power. Queridos amigos Viridiana Sánchez, Allan Jarillo, Guillermo Saldaña, familias Luquin Reyes, Gasca Reyes, and their extended families; my brothers Emil and Alex Anton; confidant Timo Siukonen; fellow students and dear friends, Gabriela Greilinger, Joost Backer, Alex Carter, Larissa Haukilehto, and Francisco Chaín. I am humbled by the magnitude of kindness and sincere efforts of the following people to take this research project further, especially during my fieldwork. I am endlessly grateful to Graciela Ávalos, Carlos Quezada, Ángel Guerrero, Humberto Guerrero, Emilio Durán, Roxana Díaz, Ville Pennanen, Eduardo Pérez, Ana Paula Hernández, Anja Grahn, Saana Ihamäki, Haley Vartanian, and the students of ITESM who helped me to transcribe hours and hours of interviews.

People in Ecatepec and Nezahualcóyotl. Thank you for agreeing to be interviewed; for sharing experiences, laughs and tears. Thank you for teaching me about resilience and bravery. Thank you for caring enough to take a chance and trust me.

Lastly, Alejandro Melgoza. You are the common thread throughout each step of the process. Without you, it would not have been the same. Amigo, te agradezco eternamente.

L

A

CNDH National Human Rights Commission

Comisión Nacional de los Derechos Humanos

CSP Civilian Self-Protection

ENVIPE National Survey of Victimization and Perception of Public Security Encuesta Nacional de Victimización y Percepción sobre la Seguridad Pública INEGI National Institute for Statistics and Geography

Instituto Nacional de Estadística y Geografía

IO International Organization

LRA Lord's Resistance Army

NAFTA North American Free Trade Agreement

NGO Non-governmental organization

PAN National Action Party

Partido Acción Nacional

PGR The Attorney General of Mexico

Procurador General de la República

PRD Party of the Democratic Revolution

Partido de la Revolución Democrática

PRI Institutional Revolutionary Party

Partido Revolucionario Institucional

PRONAPRED National Program of Social Prevention of Violence Programa Nacional de Prevención del Delito

SEDENA Secretariat of Defense Secretaría de la Defensa Nacional

SESNSP Executive Secretariat National System of Public Security Secretariado Ejecutivo del Sistema Nacional de Seguridad Pública

SSP Secretariat of Public Security

Secretaría de Seguridad Pública

ZMCM Valley of Mexico Metropolitan Area

T

F

Figure 1. Homicide rate in Mexico from 1997 to 2017 36

Table 1. Perception of Insecurity and Public Security in Ecatepec and Nezahualcóyotl 45 Table 2. Communities’ protection strategies in Ecatepec and Nezahualcóyotl 68

M

Image 1. Municipality of Nezahualcóyotl, the State of Mexico, Mexico (Wikimedia Commons 2011).

Image 2. Municipality of Ecatepec, the State of Mexico, Mexico (Wikimedia Commons 2011).

I

“The last ones were killed here.” In a neglected and dusty road, I stand on what five years before was an ample puddle of blood. A community activist describes the bloodbath of 13 people in a neighborhood in Ecatepec, Mexico: tortured and shot bodies were brought to be dumped here, and the “last ones” – a father and son – happened to witness the moment when everything took place. This violent killing in 2013 was a continuation of confrontations between organized crime groups that had started to frequent the neighborhood a year before, increasingly impacting civilians’ lives.

A few days later in the neighboring municipality of Nezahualcóyotl, I stand in front of a gated house, in the same spot where the resident was forced to surrender his car at gunpoint. The armed robbery in 2012 led to an armed confrontation a few blocks away between the local police and the delinquent, barely an adult, who was detained. The police officer involved in the shooting pleaded with the resident to report the crime because of fear of being killed due to what happened. He complied. When returning home after filing a formal report, a group of people was waiting for him outside the house. They explained calmly: “You have a wife, a daughter, and a son,” and requested him to withdraw the report. Again, he complied.

These testimonies illustrate the prolonged violent conflict in Mexico during its transition to democracy since the beginning of the 2000s, which encompasses high-impact violence with economic motivations, multiple actors fighting for control, impunity, and escalated (threat of) civilian victimization; in sum, persisting insecurity and fear have infiltrated into people’s everyday lives. However, more remarkable is what occurred after these violent events: In both municipalities, residents started to organize and collectively create strategies to protect themselves as a response to the violent events.

The violent conflict in Mexico submits civilians – understood here as non-combatants – to a daily (in)direct victimization: in addition to violence by organized crime groups, local gangs, and state security forces, the poor urban neighborhoods face unemployment, youth recruitment into organized crime groups, political manipulation, and limited access to basic services such as electricity and water, and most importantly1_{, to security. 93 percent of the people in Ecatepec and 80 percent of the people}

in Nezahualcóyotl do not feel safe (INEGI 2018b). With an average of 2100 homicides per month, 2017 was described as the most violent year in two decades2 _{(SESNSP, 2018). The phrase La gente está}

harta describes the frustration of the people; they are tired.

Nonetheless, civilians are not powerless bystanders caught in the midst of an armed conflict; they have agency. I base this assumption on the recent civil war literature that emphasizes the role of civilians and their ability to respond to threats posed by an armed conflict (e.g. Arjona 2016; Baines & Paddon 2012; Kalyvas 2006; Kaplan 2017; Masullo 2017): Instead of receiving protection, civilians have the capacity to provide it by making decisions and creating strategic alignments to prevent violence. These choices shape armed groups’ actions and, consequently, the conflict dynamics (Arjona 2014; Kaplan 2013; Kaplan 2017).

To understand the mechanisms used by civilians, first it is necessary to understand why they are targeted: “Civilians’ actions cannot be divorced from armed groups, whose motives and incentives, preferences and choices are key for explaining violence (…)” (Kaplan 2017: 11). In Mexico, the violence that takes place in urban neighborhoods, both opportunistic and strategic3_{, is predominantly}

motivated by economic gain. Pansters (2018) describes how within the past two decades, a transition from low-impact to high-impact violence has occurred, which calls for alternative strategies: not only do civilians have to protect themselves from assaults and armed robberies, but also from extortion and kidnapping.

This leads to the research question:

How do civilians protect themselves in urban neighborhoods in Mexico?

Additionally, the following questions are asked:

What collective protection strategies do civilians use to prevent violence?

How do state actors and organized crime groups shape communities’ attempts to prevent violence?

By using empirical evidence from two urban municipalities in Mexico, this research aims to describe what protection strategies civilians use in order to prevent violence, focusing predominantly on alternatives that have emerged as a result of collective decision-making. Additionally, it will shed light on how non-state and state actors use violence against civilians and how their actions shape the strategies that organized communities adapt to prevent it in a violent conflict.

Based on participant observation and a total of 60 informal and in-depth interviews, I argue that communities inherently look for subtle and non-violent alternatives to protect themselves. However, if the local authority is perceived present (understood here: legitimate), these strategies are complemented with non-violent collaborative alternatives with the state actor. On the contrary, if the local authority is perceived absent (understood here: illegitimate), communities complement their strategies with violent or illegal alternatives to confront or cooperate with armed actors.

Encouraging examples of communities that have avoided violence and ensured protection amidst civil war exist around the world (e.g. Anderson & Wallace 2013). This research will bring insight to strategies that people use in conditions of a violent conflict, where armed groups do not fight over state power but for economic interests. Beyond contributing to academic discussion, this research is important to the political debate on public security and, more importantly, to the people who face violence as part of their everyday lives.

The study proceeds as follows. First, I present the research design and methodological approach, including a description of the fieldwork in Mexico and ethical considerations. Second, I discuss civilian protection and violence against civilians in civil war literature, summarizing what strategies are implemented and what are the armed groups’ motivations to target civilians. Third, the theoretical framework of civilian decision-making in civil war and economically motivated violence is presented. I highlight when communities are more likely to implement subtle vis-à-vis hostile alternatives of protection. Fifth, the background of Mexico is contextualized on national, state, and municipal levels, summarizing the political, economic, and social conditions that have led to the persistence of the violent conflict. Sixth, the results of the empirical evidence are presented. The protection strategies are structured according to the realms of organization: avoidance, internal strategies, and allying with external actors; which is then followed by a presentation of the key findings and a discussion of alternative explanations in the literature. Lastly, the research question will be answered, including a discussion of the research’s limitations and concluding remarks.

CHAPTER TWO

R

D

This chapter explains how this research was conducted. First, the qualitative research methodology and the selected methodological approach of comparative case study is presented. Second, the case selection is briefly described, followed by a detailed explanation of the fieldwork as the primary element of the data gathering process. Lastly, the ethical imperatives and limitations are discussed.

2.1. METHODOLOGY

The research approaches the topic of civilian protection by conducting a comparative case study using qualitative methods. This type of social inquiry enables an in-depth descriptive analysis by using a small sample size, and to obtain a variety of interpretations and explanations of how people perceive their reality (Holloway 1997). The comparative case-oriented approach aims to provide “rich descriptions of a few instances of a certain phenomenon” (Porta 2008: 198). This research draws on empirical evidence from the field to detect the similarities and differences in protection strategies in neighboring municipalities, and how these shape civilian preferences and collective action in the situation of a violent conflict (Bryman 2012).

2.2. CASE SELECTION

The case selection is based on the approach of paired comparison between two municipalities. The selected cases share similarities, for example in terms of levels of violence, urbanization, and social and economic development, but the research focuses on comparing differences that are produced in both contexts (Schmitter 2008).

Through the case selection of two Mexican municipalities, the research establishes a dialogue between the theoretical framework and empirical material. These municipalities, Ecatepec and Nezahualcóyotl, will be comprehensively presented in chapter six. In more practical terms, and within the context of countries facing armed violence, Mexico, and the urban area of the State of Mexico, is relatively safe for conducting research. A significant consideration in the case selection was whether or not conditions would be suitable to obtain first-hand empirical evidence in the field and to be able to successfully execute it in a limited amount of time; for example, towns of the states of Guerrero and Michoacán are also intriguing for research but would require months of preparation due to accessibility, location, and high-risk conditions (Personal Communication 3 April 2018).

Three years of previous professional and personal experience in Mexico were used for the benefit of the research. The prior knowledge facilitated interaction with the locals and efficient data collection. The possibility to conduct interviews in the local language facilitated data gathering and collecting empirical evidence in the field. No translators were needed and the interviews were carried out without the intervention of others. As the interviewees were able to express themselves using slang, swear words and local expressions, the descriptions were precise and direct, consequently benefitting the specificity of evidence obtained from the field.

2.3. FIELDWORK AND DATA GATHERING

This research draws on 60 interviews conducted during four weeks of fieldwork in Mexico – Mexico City and the State of Mexico – in April 2018. The aim was to gather the primary first-hand empirical data on a local level in two municipalities. Additionally, an extensive amount of other primary and secondary sources was used for the purposes of the research. Academic literature, governmental and non-governmental reports, statistics and surveys, new articles, and personal communication with experts have had a significant role prior to, during, and after the data collection period in Mexico.

A commonly used procedure for a data gathering in qualitative research, called snowball or chain sampling, was used throughout the process. This procedure takes advantage of existing social networks in a natural way in order to “generate a unique type of social knowledge – knowledge which is emergent, political and interactional” (Noy 2008: 327). The benefit is that the researcher is able to increase the amount of connections while conducting interviews and meeting new people. The approach was particularly effective in the Mexican context where much of the connections in general are established through existing networks; people are more approachable when there is a reference of a mutual contact or when there is a possibility to do a ‘favor’ that could benefit or help to expand the network in the future.

The procedure is dynamic and accumulative: In February 2018, a small group of people who the researcher knew previously in Mexico were contacted. A month later, a standardized email was sent to a dozen people and organizations selected due to either a recommendation or the researcher’s own investigation. Nine initial contacts established during prior fieldwork, in addition to other people met later in Mexico, provided the access to interview civilians in both municipalities, state-level authorities, local government representatives, security, and armed forces both officially and unofficially in addition to a variety of civil society and (non)governmental institution representatives. Written or recorded data exists from 70 interviewees, not including dozens of people with whom there

was an opportunity to talk to in diverse circumstances during observation periods in Ecatepec and Nezahualcóyotl (Appendix 1).

Trust played a significant role, and it was evident particularly when interviewing police officers. An informal community leader – a native resident from Ecatepec whose actions for the locals have implied that he enjoys exceptional trust within a large network of people – organized unofficial interviews with police of different levels from both municipalities. Due to mutual support and respect – for example, he provides police officers with low-cost ammunitions and equipment, and had a doctor come vaccinate the interviewees; he in return is saluted as comandante, commander – people confided delicate details related to their work and everyday struggles. He, on the other hand, was accessible to support the research due to a mutual contact: “Because of [X], I trust you. I trust him so I trust you. This is how it works, we are building a cycle of good” (Personal Communication 25 April 2018).

2.3.1. INTERVIEWS AND PARTICIPANT OBSERVATION

During fieldwork, empirical evidence was obtained in three ways: by conducting in-depth interviews, informal interviews, and through participant observation.

In-depth interviews are described as “conversations with purpose” where structure is combined with flexibility (Webb & Webb cited in Legard, Keegan & Ward 2003: 138). During the interviews, a predefined set of key topics is covered while keeping the structure of the interview conversational. The approach allows the interviewee to talk freely after the initial question while letting the researcher react spontaneously to relevant issues (Legard, Keegan & Ward 2003). During the interviews, I strived to cover the following themes: authority, control, corruption, impunity, justice, (community) organization, security, and violence. Various techniques – for example, long pauses, follow-up questions, and repetition – were used to encourage depth in answers.

The informal interviews were used or to gain more descriptive information in the field and “are an essential part of gaining an understanding of a setting and its members' ways of seeing” (Cohen & Crabtree, 2006). These interviews were not planned nor followed any particular guidelines. The way of interview resembled more of a conversation with neighbors where participant observation took place.

In-depth and informal interviews allowed me to respond flexibly to the changing conditions in the field. When the time and space allowed me to conduct an in-depth interview, it was carried out. In other situations, I relied on informal, conversational interviews to gather testimonies and experiences of community organization, violence, and protection strategies. Respondents in the two

case study municipalities included both state and non-state actors: civilians such as community leaders and activists, students, local business owners, professors, and local residents of all ages in the neighborhoods. State actors were primarily municipality and state police officers, directors, and government officials from violence prevention programs4_.

Since the foundation of using qualitative methods is to understand human experience, participant observation was used to make sense of the dynamics and interpretations occurring in specific surroundings (Cohen & Crabtree 2006). I engaged and participated in the social setting, gathering and documenting my own observations from the field in both municipalities. The observations are used as chief notions to describe, contextualize, and to compare, allowing for explanations of existing conditions and variations in both neighborhoods. The observation periods I spent both with civilians and the municipality police; I participated in Nezahualcóyotl municipality police’s daily security committee and training session. Additionally, I witnessed the block networks’ meetings with the local police and information sessions. With civilians, I spent time in number of neighborhoods is both municipalities, visited local schools, and concurrently carried out informal interviews.

An inductive approach was used for data analysis. First, primary data was gathered and recorded. Second, it was transcribed, labeled, and categorized to find patterns.5 _{Lastly, the findings}

were placed into a theoretical context and the final argument was defined in order to draw the conclusion of the research.

2.4. ETHICAL DISCUSSION AND LIMITATIONS

Ethical considerations were taken into account from the beginning of this research. The leading imperative of research is to do no harm (Wood 2006), which is “intensified in conflict zones by political polarization, the presence of armed actors, the precarious security of most residents, the general unpredictability of events, and the traumatization through violence of combatants and civilians alike” (Wood 2006: 373).

During fieldwork, I followed the norm of informed consent (Wood 2006). All the interviewees received a comprehensive explanation of the objectives and purpose of the research prior to the interview. The people I interviewed understood the risks and benefits that our interaction might imply and got to decide whether to talk with me or not. However, not all risks can be controlled. Participants’ evaluation of risks is subjective. Also, an interview can cause psychological or physical harm or

retraumatize the interviewee (Fujii 2012). During interviews, I aimed to build trust and a space of confidence. Most of the immediate feedback was positive; the interviewees were responsive and grateful ‘that someone finally listens’.

For the purposes of the research, it was important to hear people’s perceptions and testimonies from their everyday lives. Thus, all interviews were conducted under anonymity. Precautions were taken when recording the interviews; no names, dates or places were mentioned. Interviewees were given instructions to interrupt or stop the recording or the interview at any time.

The fieldwork required that several detailed safety measures were implemented. I relied on safety recommendations suggested by the locals and the people accompanying me in both municipalities when entering areas with which I had no previous experience. I did not attempt to interview active members of organized crime groups due to the risk it might have posed for those interviewed. Due to the same reasons, no reference to specific neighborhoods is made in this research as it could reveal delicate information. Also, I aimed to remain as apolitical possible: Thus, no neighborhood delegates were interviewed due to their strong affiliation with governing political parties. If political preference of an interviewee is considered noteworthy further in this research, it will be clearly indicated.

As Schmitter (2008) posits, case selection always implies trade-offs. By selecting various cases the quality of the data and the amount of observations are not as diverse as they would be if focusing only on one case. “Inversely, the more narrowly you have defined and operationalized those variables – that is, the lower they are on the ladder of abstraction – the less likely they are to be relevant in a wide range of cases” (p. 274). In spite of similar protection strategies or dynamics between civilian and state authorities that might exist in other places, this research focuses on the variation of conditions between these two municipalities, without generalizing the results across social settings. As Bryman (2012) maintains, “qualitative findings tend to be oriented to the contextual uniqueness and significance of the aspect of the social world being studied” (p. 392). The variations within one municipality are remarkable, and as will be mentioned in the analysis, even on the same street the conditions might change. The perceptions that are obtained from the field are subjective and context-related, and thus, unlikely to occur exactly alike; if they do, it is an empirical matter (Bryman 2012). Four weeks of constant interviews and observation in the municipalities were barely enough to have a representative sample for the purposes of this research. The municipalities are highly populated and in order to gather a sample that would thoroughly describe the nuances between different neighborhoods, the intensity and variation of experienced violence, and the socio-economic and political factors and their influence on protection strategies, more time would be required.

information they share with strangers. The use of existing networks as an approach to obtain interviewees was applied to reduce factors that might cause distrust or lead people to omit information.

CHAPTER THREE

L

R

This chapter discusses current literature on civilian protection and violence against civilians in civil war. Also, it gives an overview of the contributions and limitations specific to literature on the armed conflict in Mexico. First, civilian protection in civil war is approached through the emerging literature on civilian agency and civilian self-protection, referring to the existing conceptualizations and implications on community organization. Second, the use of strategic and non-strategic violence is discussed, referring to the armed groups’ preferences and strategies to use or restrain from violence to gain support, control, or economic benefit. Lastly, the current literature on Mexican armed conflict is presented, highlighting the role of organized crime groups and ambiguities related to the conceptualization of the conflict. Also, some recent approaches to civilian protection in urban Mexico are considered.

3.1. CIVILIAN PROTECTION IN ARMED CONFLICT

Predominantly, states, international organizations (IOs), and non-governmental organizations (NGOs) have been perceived as the traditional actors who offer protection, and civilians as its recipients. Although state-centric theories6 _{have emphasized the top-down approach to civilian}

protection, civilians have also succeeded in protecting themselves when the external actors have been absent, insufficient, or the ones perpetrating violent acts (Jose & Medie 2015).

Recent civilian protection and civil war literature has increasingly emphasized the importance of civilian agency in armed conflicts and referred to civilians as participative actors rather than passive subjects or powerless victims (Arjona 2016; Baines & Paddon 2012; Bonwick 2006; Godoy 2006; Jose & Medie 2015; Kalyvas 2006; Kaplan 2013; Kaplan 2017; Krause 2017; Masullo 2017; Valentino 2014). Scholars assert that civilians are chief actors in creating strategies to protect themselves when facing a conflict. While “[i]t is not obvious that unarmed civilians in civil war can protect themselves against heavily armed combatants, (…) some civilians do” (Bonwick 2006; Jose & Medie 2015; Kaplan 2013; Kaplan 2017: 3; Masullo 2017).

Jose and Medie (2015) suggest that protection strategies implemented by individuals and communities vary due to geographical setting, the interests and ideals of warring factions, and the economic opportunities available to civilians (p. 528). Kaplan (2017) explains that “multiple kinds of violent threats civilians face call for multiple solutions. Subject to constraints of available ideas and

imperfect information about levels of danger, civilians may thus select different kinds of autonomy strategies in different places as they organize and adapt responses to different types of violence” (p. 10).

The emerging self-protection literature has offered some typologies of civilian self-protection (CSP) strategies. Jose and Medie (2015) divide protection strategies into three categories: non-engagement, nonviolent non-engagement, and violent engagement. The first category of non-engagement implies no interaction with combatants nor ‘traditional’ protection actors that are perceived as a threat. The strategies include alternatives of flight, shelter, and silence. The second category, nonviolent engagement, suggests that some interaction with one or many parties exists in order for the civilians to maintain safety. This can be ensured by maintaining neutrality, paying for protection, and providing information or labor. Lastly, violent engagement consists of organized or unorganized acts of violence committed by individuals or collective groups to ensure protection from physical harm by supporting or joining armed groups.

Masullo (2017) posits that strategies that are encountered in different conflict settings are lacking solid conceptual framework or adherence to theory within civilian protection research. For example, Jose and Medie (2015) state that Ugandan civilians used a combination of strategies to escape the Lord’s Resistance Army (LRA); Baines and Paddon (2012) claim that the civilians’ strategy was to avoid the armed group. Anderson and Wallace (2013) stress the difference of conflict prevention and conflict avoidance, highlighting civilians’ decision to “set themselves apart from the agendas of a war” in order to avoid the violent conflict, not the conflict as a whole (p. 10–11).

The conflict conditions and their characteristics vary, and the nuances of applied protection strategies may be disregarded in search of a broader classification. Measures for civilians to assure their survival often combine a set of multiple strategies (Baines and Paddon 2012; Kaplan 2017; Masullo 2017) or include motives not only limited to protection. Baines and Paddon (2012) posit that in Uganda, civilians received money for working as informants for the LRA. Testimonies describe that “there were three primary reasons why a civilian would cooperate or collaborate with an armed group: financial gain, family or fear” (p. 239). The notion of collaborating due to fear raises an important consideration: How to approach these ‘strategies’ that civilians apply, when they are only a result of armed groups’ coercive measures? For example, Jose and Medie’s (2015) typology includes tax/toll paying as a strategy of nonviolent engagement. Bonwick describes a similar case (2006): “A colleague in Congo several years ago told of proudly using her humanitarian status to negotiate her way through a checkpoint on the outskirts of the city without paying – until begged by the driver to do so. The people from the checkpoint came to his house every night to demand the ‘tax’ from him”

Scholars coincide by arguing that available alternatives for the civilians to protect themselves are to flee (Kaplan 2017; Steele 2009), to seek protection from one of the armed groups (Baines & Paddon 2012; Kaplan 2017), and not to participate (Kaplan 2017; Masullo 2017). Masullo (2017) proposes a typology of civilian responses to explain existing alternatives in an armed conflict: whether civilians leave or stay, and, if they stay, whether or not they decide to cooperate.

When facing violence, many civilians flee from it. This has happened, for example, in the Northern states of Sinaloa and Sonora, where towns were abandoned due to the drug-related violence (BBC 2013; Fregoso 2017). However, as Kaplan (2017) describes in the context of the Colombian civil war, displacement is not always perceived as an attractive alternative and can pose even higher risks than staying in the community.

Civilian autonomy, community-level organization, and strategies in civil war settings can bring security and respond to different types of violence perpetrated against civilians. Collective organization allows strategies that “may have general benefits for reducing violence above and beyond what individuals are capable of, regardless of conflict conditions” (Kaplan 2017: 15).

Kaplan (2017) asserts that armed conflicts do not necessarily damage social relations but can actually promote them. Organized civilian agency and collective action can promote pacific norms and collaboration within the community to inhibit civilians’ involvement with armed groups. Civilian capacity to prevent violence is linked to communities’ social cohesion and cooperation in war settings, ultimately to achieve social capital. As he suggests, “[w]hile civilian strategies for autonomy can and do arise specifically as a response to deal with armed conflict, preexisting bases for social cooperation are a helpful catalyst” (p. 4).

However, while united and strongly organized communities are able to protect civilians in certain conditions, their capacity to recover is challenged when they are subjected to the actions of armed groups. Kaplan (2017) argues that social cooperation and organization to collectively seek protection is a way for the people to control their communities and manage the armed groups’ intents to gain control. He claims that through cohesion and solidarity mechanisms, civilians take advantage of the previously established social harmony, allowing communities to respond spontaneously but coherently to aggressions of an armed group. Though less enduring and more informally sustained, the collaboration is based on pre-existing ‘collective agreement’ owing to well-founded community dynamics.

Scholars suggest that self-interest and self-protection are key motivations for non-violent prevention efforts (Anderson & Wallace 2013; Chenoweth & Cunningham 2013; Krause 2017). Communities predominantly follow norms that respect human life and prefer pacific measures to

instability and protect their livelihoods” (Kaplan 2013: 5). By not getting involved, communities transmit a peaceful agenda and norms of acceptable behavior to armed groups: “[i]n the face of abuse and transgressions by rebel (or even government) troops whose norms are either weak or eroding, civilians have been known to collectively protest or engage in non-cooperation to voice their grievances and call for adherence to norms against civilian victimization” (Kaplan 2013: 5).

Community organization can also lead to violent outbursts, such as lynchings in the Guatemalan civil war. Still, the driving grievances are not self-explanatory. As Godoy (2006) argues, “communities seek not only to punish and deter criminal activity, but, perhaps more important, to collectively reassert themselves as agents rather than victims” (p. 102).

Civilian protection strategies can result in an increase in the risks of violence against civilians, especially when violent strategies are applied. Jose and Medie (2015) underscore that peaceful resistance is not always an alternative for people caught in violent conflicts. However, if civilians turn to violent self-protection strategies, the violence might escalate even further: The protection strategies can become insufficient and make civilians even more vulnerable if armed groups perceive civilians as a threat to their survival (Kaplan 2017).

Masullo (2017) argues that violent and nonviolent should be treated separately when referring to civilian protection strategies. Even though cooperation or collaboration with armed groups might be due to the civilians’ primary objective to ensure protection, “civilian participation with armed groups shape war dynamics in ways that can directly affect the prospects of civilian survival” (p. 23). For example, in Southern Thailand, the increase in firearms and civilian militias influenced the conflict dynamics, leading to deteriorating communal violence (Jose & Medie 2015). Arjona (2016) underlines that recruiting and controlling civilians imply that armed groups enjoy increased territorial control, while Kalyvas (2006) argues that control and collaboration between armed groups and civilians alter levels of violence in conditions of irregular war.

3.2. VIOLENCE AGAINST CIVILIANS

The cause-centric approach to civil war has disregarded understanding violence, conceivably a central feature of civil wars, and its variation in each conflict. Until the beginning of the 2000s, the political science research on civil war still largely focused on understanding the causes of war in able to prevent wars in the future (Kalyvas 2006; Valentino 2014). By treating violent conflicts alike, an important notion remained neglected: to understand why some conflicts are so much more violent than others, especially when it comes to civilian casualties (Valentino 2014).

The prevalent literature emphasizes a rationalist approach to comprehend violence against civilians: instead of being “wanton and senseless”, scholars argue that strategic reasons motivate armed actors to use violence against civilians and link it to a strong intentionality of action. (Kalyvas 1999 in Hoover Green 2016: 620; Valentino, 2014). However, Hoover Green (2016) posits that in armed conflict, the armed groups are not unitary and violence can occur due to a group strategy, material necessity, unfavorable combatant selection, individual opportunism, habituation, or panic – or due to a combination of some or all of them applied simultaneously (p. 621). Still, the armed actors tend to use violence to achieve short-term objectives: “the long-term consequences of severe violence are often difficult to predict” (Arendt, 1970 in Wood 2010: 648).

Stanton (2016) posits that even though governments and non-state groups benefit from using violence against civilians strategically, strategic and non-strategic violence should be considered as something that coexists in armed conflicts. Hoover Green (2016) states that referring to armed groups as rational decision-makers who strategically use violence against non-combatants is unsatisfying: “[I]n reality violence against civilians is frequently committed, without, or even in defiance of, explicit orders from above” (p. 620). She argues in the context of organizational control that ‘optimal’ agreement on levels, targets, and types of violence are constantly contested; even though that agreement would exist, combatants can make the decision not to follow the given instructions (Hoover Green 2016).

Non-strategic violence plays an important role in the context of armed conflicts. Scholars argue that violence against civilians is likely to occur when an armed group is not able to control its soldiers and their use of violence. The lack of organizational structure and control translates into lack of discipline especially among the non-committed combatants who are more prone to commit violent acts towards civilians (Stanton 2016).

Hoover Green (2016) asserts that effective behavioral control within armed group institutions can imply that more restraint towards civilians is exercised. The opportunistic approach to violence against civilians is related to organizational dynamics: In order to control it, the organizational strategy requires a solid hierarchy and line of command that exercises authority and control. Thus, “leaders must use discipline to reward desired behavior and punish transgressions as well as implement political education to align the preferences of armed group with those of the group leadership” (Hoover Green 2011 in Stanton 2016: 57; Manekin 2013; Stanton 2016).

Armed groups and the government use violence as a means to control civilians or territory. Kalyvas (2006) argues that in conditions of conflict, civilians lean on any group that is able to provide them more security. In the case of state terror, civilians often lack organized alternatives to fight back:

by political actors. Political actors try to shape popular support (or collaboration) and deter collaboration with their rival (or defection)” (Kalyvas, 2006: 12; Wood 2010). Stanton (2016) asserts that the use of violence comes with a strategic consideration of the political objectives that will be achieved. For example, the more civilians favor a specific group in a contested territory, the higher threat the population poses to other groups. Consequently, the benefit of gaining control over these civilians is high – as is violence as a mean to achieve it.

Stanton (2016) states that when armed groups depend on civilians, restraint towards them is practiced. This is due to the high domestic costs of violence: the group might lose supporters that it desperately needs if violence against civilians is used. These armed groups are often ones whose structures are strongly linked to the communities where they operate: Valentino (2014) describes that armed groups rely on civilians for supplies, intelligence, shelter, and recruits (p. 94). If social cohesion is solid, less violence should occur. Accordingly, killing becomes costlier for armed groups and organizational abuses less tolerated by commanders: armed groups are more likely to coerce or abuse when there are little consequences (Kaplan 2017: 11; Stanton 2016; Valentino 2014; Wood 2010).

State or non-state armed actors restrain from violence against civilians also, when they are in total control of a specific area (Kalyvas 2006). Similarly, violence is not used in territories where they do not enjoy any control. Consequently, violence against civilians occurs when the territory is contested and where the control is fragmented among state and non-state actors. In these situations, civilians are coerced to cooperate in order for the actors to gain more power in the territory. Wood (2010) stresses that the use of collective violence can be temporarily beneficial in that it interrupts opponents’ control over territory – to limit resources, for example.

Kalyvas (2006) states that large-scale violence against civilians is inefficient and suggests that armed groups engage in selective violence to attack specific civilians who support the ’enemies’ in some way. As Schwartz and Straus (2018) relate, in the Guatemalan civil war “state actors used violence to deter civilians from future collaboration with the insurgency. Violence here served a communicative function; it sought to ‘teach a lesson’. (…) Civilian groups were ‘winnable’, and violence was a means of persuading them to abandon the guerillas” (p. 223). Stanton (2016) refers to low-casualty control that implies limited violence which is directed at those with undesired behavior: “such violence seeks to punish or eliminate those individuals who are aiding the opponent, as well as to deter the wider civilian population from future collaboration with the opponent” (p. 45). Thus, not all civilians are targeted. Rather, selective violence is exercised against those considered to support the opponent (Kalyvas 2006; Stanton 2016; Valentino. 2014).

3.3. ARMED CONFLICT IN MEXICO

The Mexican violent conflict has been overlooked, if not disregarded, in civil war literature. Kalyvas (2015) has made an important contribution in studying organized crime in the context of civil war to find alternative approaches to understand the occurring conflict in Mexico. Approaching the conflict is complex due to many ambiguities relating to definitions of the organized crime itself. To understand it, the type of actors, their motivations, and other underlying factors have to be studied. This requires a well-constructed conceptualization that thus far has been lacking in the case of armed conflict in Mexico (Carpenter 2013; Kalyvas 2015).

Until now, descriptions of ‘narcoterrorism’, ‘narcoviolence’, ‘drug war’, ‘civil war’, ‘criminal insurgency’, and ‘the new Mexican civil war’ have all been used in literature and predominantly drawn from the context of criminal justice to describe the armed conflict in Mexico (Carpenter 2013; Kalyvas 2015). Kalyvas (2015) maintains that two dominating factors have contributed to this ambiguity: the empirical – the perceived intensity of violence and criminal activities by the armed groups, and the conceptual – it is not an interstate war and the levels of violence are the same as a civil war; thus, it is a civil war.

The casualties incurred by the organized crime have surpassed the threshold of civil war. However, as Fearon and Laitin (2003) and Kalyvas (2015) stress, for a conflict to be a civil war, the armed groups must have political objectives and aim for state power. This does not occur in the Mexican case, since the organized crime groups are primarily led by profit. As Kalyvas (2015) posits, “the new Mexican civil war is not a classical civil war in which ideological insurgencies strive to topple state power. It is a prototypical ‘new’ civil war, fought for material gain not for social justice” (p. 1520). Accordingly, Carpenter (2013) challenges the notion of the Mexican conflict as a drug war, but questions that armed groups are merely economically oriented criminal organizations, disregarding the tight interconnections with political state institutions and access to political power: “[P]olitical and criminal worlds are interpenetrated (…). Treating Mexico’s conflict as a struggle between criminal “drug lords” and the “government” overlooks the fact that governmental actors were complicit for decades (…)” (p. 146). She suggests Mexico has a factional-economic conflict: the political and criminal world collide, spaces become parallelly governed, and armed groups provide protection to local communities. In this context, the state is ‘at best one of the many competitors for authority’ (Clunan & Trinkunas cited in Carpenter 2013: 148).

Carpenter (2013) posits that the vague denominations of the conflict have resulted in state responses that have mainly focused on confronting violence inflicted by the conflict by diverse military and law enforcement strategies, emphasizing “drugs and ‘narcos’ as the subject or cause of war and

violence” and considering the organized crime and smaller (youth) gangs, pandillas, as “the threat to Mexico in terms of criminal actors” (p. 144). While the United States has also reinforced these measures by funding Mexican antinarcotic and public security strategies through foreign assistance, these measures have been inadequate to reduce violence. Instead, building social capital between communities and law enforcement could yield more significant results in terms of tackling violence prevention (Carpenter 2013).

3.4. APPROACHES TO CIVILIAN PROTECTION IN URBAN MEXICO

Müller (2016) highlights that research on civilian self-protection in Mexico and other ‘violent democracies’ in Latin America is neglected, even though people are increasingly finding means of self-help to respond to (urban) insecurity through various measures. In the context of Mexico City, he has identified the categories of self-policing, which is “the regulation and limitation of one’s own conduct and activities in public and private spaces due to security concerns”; commodified security, an example of which is the privatization of public security; and community justice (vigilantism and lynching) (p. 120).

Müller argues (2016) that individual protection measures that aim to inhibit crime consequently influence social relationships, fueling distrust among people. This securitization ‘from below’ is giving preference to security over social interaction, as people in urban neighborhoods in Mexico City are distancing themselves from public spaces. “In social contexts where, in principle, everyone unknown is expected to be a potential security threat, many people not only ‘bury themselves in their homes’ (…) but also tend to reduce their contact with other people, and carefully evaluate the risks involved in such interactions” (p. 121). The increased self-policing has fueled the fragmentation of urban space, while the environment of fear gives room for the government to create mano dura policies, as people feel increasingly insecure (Müller 2016).

However, non-violent and violent collective efforts also occur, particularly in marginalized neighborhoods (Godoy 2006; Müller 2016; Pansters 2018). These measures include, for example, exchanging telephone numbers with neighbors and taking mutual care of the businesses among shopkeepers. In these conditions, the community itself is providing security, founded on strong social cohesion. Still, that perception of ‘permanent informal surveillance’ in the neighborhood may have counterproductive effects: conflicts within the community are internally resolved, decreasing the assistance sought from the authority. As Müller (2016) maintains, “the search for protection ‘beyond the state’ through forms of collective self-policing (…) mirrors the social constraints placed on local

residents in neighborhoods with very dense levels of internal social cohesion and social control to seek assistance of public institutions (…)” (p. 124).

Pansters and Castillo Berthier (2007a) posit that the persisting violence and the absent state security forces have increased the different civilian-led actions to bring security, shifting from public to private: civilians increasingly use public security for private means to assure protection. Since the late 1990s, the use of private police and street guards in residential middle-class neighborhoods has increased in big cities not only in Mexico but in Latin America in general (Müller 2016; Pansters & Castillo Berthier 2007a).

However, the participation of authoritiesin neighborhoods’ protection strategies has remained merely informal or illegal. This ‘privatization of public security’, according to Müller (2016), has led to the police becoming the “single most attractive actor within the local protection business” (p. 130). Koonings and Kruijt (2004) posit that where security forces and private police forces intertwine, also the ‘legal violence’ has expanded towards ‘extra-legal violence’ (p. 9). Pansters and Castillo Berthier (2007a) argue that while in wealthier neighborhoods the iron gates, high fences, and guards protect private spaces from the violent reality, “poor neighborhoods also suffer from different forms of crime but their inhabitants lack the means to protect themselves in this manner” (p. 51).

CHAPTER FOUR

T

C

This chapter presents the theoretical framework for civilian protection and violence against civilians in an armed conflict. First, I explain Kaplan’s (2017) theory of civilian decision-making in civil war, emphasizing the role of civilian agency to prevent violence by community organization, strategic alignments, and collective action. Second, the theoretical approach of violence against civilians for the purpose of economic gain is presented, distinguishing the profit-oriented armed groups’ motivations to use or refrain from using strategic and non-strategic violence against civilians. Lastly, central concepts of war and conflict are discussed, in addition to definitions relevant to this research.

4.1. CIVILIAN DECISION-MAKING IN CIVIL WAR

Different social conditions influence how communities respond to and survive an armed conflict. Kaplan (2017) refers to social landscapes instead of vacuums; there may be notable variations from one town to the next that influence civilians’ ways to manage the conflict and, successively, the way they are treated by armed actors.

Civilians are neither simply collaborators or actors. Kaplan (2017) highlights that civilians have agency; they can apply efficient protection strategies and prevent violence in conditions where violence would normally occur. This requires collaboration within the community: “Lone civilians, like lone soldiers, are unlikely to be effective in complex and changing environments and against military organizations” (p. 36). In instances where civilians would easily submit to threats and pressures posed by armed groups, groups of civilians have the possibility for stable responses even though the situations are complex and face constant changes. Organized communities are able to make collective decisions on how to respond to the threats posed by the conflict. Arguably, civilians do not have the same military capacity or power as armed groups. However, it does not imply that they would not be capable to protect themselves from more ‘powerful’ actors in an armed conflict (Kaplan 2017).

4.2. CIVILIAN STRATEGIC RESPONSES IN AN ARMED CONFLICT

Kaplan (2017) posits that civilians evaluate both internally and externally before deciding on the strategy they will apply: “Civilian organizations select tactics (…) after consideration of both the

where organization fails due to a lack of collectively defined strategies, lack of incentives, or posed threat, a common response from communities is to do nothing. Exposed without a strategy, the civilians may rely on individual displacement, which Kaplan (2017) describes as happening gradually, one by one, to different directions. The communities that are constantly left with fewer residents are challenged to collectively organize in the future against armed actors.

Civilians may ally with armed actors for protection. Where civilians provide material support and information, they receive protection and continuation of public order. However, civilians may also be forced to comply with armed groups’ conditions to offer protection. For example, civilians can be forced to pay “protection taxes” in exchange for protection. Kaplan (2017) posits that “[w]here the presence of the army is strong, most communities side with the government for protection from insurgent groups” (p. 45).

The strategies of autonomy7 _{that civilians reinforce in order to secure the absence of violence}

may or may not include neutrality. Kaplan (2017) defines strategies between weapons of the weak – subtle and primarily internal strategies that focus on community-building – and weapons of the not-so-weak – strategies that are directed against armed actors and impose costs to the community. Commonly, nonviolent measures of protection are implemented in order to reduce violence: these can be spontaneously coordinated and applied or involve a more profound coordination and underlying organization. Thus, actions can respond to a direct, immediate threat, or be preventive and concurrent actions, such as sharing information about conflict conditions and persuading young community members not to take part in armed groups (Kaplan 2017).

As Kaplan (2017) describes, the variety of the strategies that are implemented in communities depends on the existing alternatives; successful protection requires creativity and exchange of ideas within the community. Alternatively, strategies “may come from knowledge about which strategies have been used in other places, what their requirements are, and how they have fared” (p. 52). While some alternatives might be complementary, others may be exclusive: the implementation of strategies is an evaluation of the benefits that can be achieved relative to possible outcomes. Kaplan (2017) argues that communities implement subtle strategies first; they protest or arm themselves as a last alternative. The theory suggests that communities that are strongly organized, have external support, or are under moderate threat of armed groups adopt the most hostile strategies. In conditions of high threat by the armed groups, more subtle strategies might be more successful and sustainable to apply.

7 _{Kaplan (2017) makes a conceptual clarification between autonomy, neutrality and nonalignment: “I see autonomy as a}

broader concept since it does not require positional decisions and can be an ideal that is sought even in the face of a single abusive group (whereas neutrality is often a relational concept vis-à-vis other actors). Autonomy can be conceived

4.3. COLLECTIVE ACTION AND COMMUNITY ORGANIZATION

The varying social and demographic conditions, in addition to support from external actors, enable some communities to organize more easily than others. A majority of civilians prefer staying put instead of fleeing, even in conditions of violent conflict. Similarly, a majority of civilians benefit from bringing an end to violence. However, not all have the disposition to actively go against the armed groups to protect themselves or their communities; the odds of success are small and can result in more violent acts by the armed groups. Kaplan (2017) states that “fear can be both pervasive and devastating to cooperation” (p. 36). This shapes community dynamics: people stop communicating with each other out of fear, thinking that shared information may later be used against them.

Communities with homogenous and divergent interests are both able to create collective strategies to inhibit violence. The organization of communities whose members share common interests to stay and to guarantee their livelihoods is considerably effortless: “civilians will participate even with little knowledge about the preferences of their neighbors” (p. 37). Their similar objectives facilitate agreeing upon strategies that will be implemented. In communities where interests differ, community members’ choices can be shaped by others and by the community leadership. Kaplan (2017) refers to ‘first movers’ who have strong preferences to protect the community but cannot succeed as a minority. These community members are likely to mobilize neighbors by persuading or compensating others to join, or to trust that their exemplary actions will convince people to participate.

Social capital – understood as “networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit” – is considered to be a central force for community organization because communities with homogenous interests are not frequently encountered (Putnam cited in Godoy 2006: 119). Kaplan (2017) refers to the importance of horizontal relationships among community members: “Social ties help residents to decide what is best for their community, to launch organizations that survive over time, and to directly affect levels of violence by preventing armed groups from seeding dissent among them” (p. 38). Additionally, factors such as geography and demography – location, ethnicity, population size, and density – may impact the fomentation of social capital. For example, the presence of large cities allows for an easy exit. Consequently, this might result in shallow interaction with other community members, leading to weak organization and low costs of displacement.

Communities may seek support from local organizations and, in this manner, reinforce community cooperation. The presence of external actors can facilitate collective action among

and recommendations on how to cope with the occurring conflict. Kaplan (2017) argues that external actors’ capacity to support communities relies on, first, whether they have previous knowledge in decision-making and founded authority, as well as access to information. Second, well-established institutions that enjoy legitimacy tend to be more effective among communities. Third, in comparison to state actors, organizations without political interests are able to achieve better results in terms of protection since they are not perceived as a threat or as an enemy by armed groups.

4.4. ARMED GROUPS’ SENSITIVITIES AND ECONOMICALLY MOTIVATED VIOLENCE

For the armed groups to be influenced by the organized communities’ strategies, they first must have motives to use violence and, second, must be flexible in modifying these motives in case they are influenced by civilians (Kaplan 2017). These sensitivities describe the armed groups’ underlying motivations to use or not use violence in changing conditions. Three characteristics can be observed: first, groups’ institutional foundations, ideology, and behavioral norms that influence their use of violence; second, dependence on civilians as a resource and its implications on discipline and reputation; and third, factors related to territorial control and tendency to cooperate with civilians. However, armed groups that are solely interested in increasing their economic benefits from civilians do not respond to protection strategies implemented by the communities: “Groups of this nature will mostly be influenced when they are met with force” (Kaplan 2017: 55).

Strategic and non-strategic violence coexist in armed conflicts, and also define armed groups that primarily operate due to motives of economic gain. While Kaplan (2017) describes these groups as cartels, Kalyvas (2015) alludes to organized crime groups, as they are strategically oriented, “hierarchically organized groups of criminals with the ability to use violence, or the threat of it, for acquiring or defending the control of illegal markets in order to extract economic benefits from them” (p. 1518). On the contrary, opportunistic violence occurs when individuals attempt something that would usually be out of reach or prohibited: violent interaction such as kidnapping, hostage taking, or looting might occur “during or in the immediate aftermath of major conflicts” (Tilly 2003: 131). These actions might be due to low-capacity regimes that “remain susceptible to unauthorized sequestering of resources” (p. 134). However, in conditions of civil war – where low-capacity regimes coincide with undemocratic governments – the civilians are more likely to suffer opportunistic violence (Tilly 2003).

The more armed groups involve civilians in their operation, the more their survival depends on them. Armed groups with a focus on economic gain are less dependent on civilian support and often

extortion. Thus, they are less perceptive to civilian strategies. Kaplan (2017) suggests that the most effective strategies to confront these groups are with subtle or extremely hostile alternatives.

Discipline and internal norms defined by the armed group may influence restraint and the use of violence against civilians (Kaplan 2017). For example, even though members would participate in opportunistic violence, they still might stay loyal to their family and community. In contrast, weak internal control implies more violence (Manekin 2013; Stanton 2016).

Armed groups with these characteristics may not aim to harm civilians, and most of their activities do not selectively target civilians. However, as criminal organizations, they do not have anything that impedes them from doing so (Kaplan 2017). Pansters (2012) categorizes violence that is motivated by economic gain into high and low levels of organization. The determining factor in both is the relationship with the state: fragile, corrupt and inefficient providers of protection, weak and delegitimate institutions, infiltration between (organized) armed groups and state actors (p. 20). High-level organization involves kidnapping, organized (drug) gang violence, and crime. These are translated into armed confrontations between armed groups and state security forces, extortion, coercion, executions, and torture. Low-level organization involves armed robberies, homicides, gang violence, violent conflicts over resources, increased occurrence of car thefts, mugging, ‘narcomenudeo’, drug dealing, and gang disputes over territory.

4.5. CONCEPTUALIZATION8

4.5.1. CIVIL WAR, ORGANIZED CRIME, AND VIOLENT CONFLICT

Descriptions of the Mexican conflict are founded on empirical and conceptual foundations. References to the ‘drug war’, ‘war against narcos’ and ‘civil war’ can be found in media, political discourse, and academic research (e.g. Kalyvas 2015). Fearon and Laitin (2003) posit that conflict is a civil war when both non-state and state armed groups have organized military forces and substantial fatalities occur; at least 1000 killings in total and 100 on a yearly average (Kalyvas 2006; Sarkees & Wayman 2010). In Mexico, between 2006 and 2011, on a yearly average over 10 000 people lost their lives, largely due to organized crime. As Kalyvas (2015) posits, the number of homicides exceeds the used fatality threshold and “elevate[s] organized crime in Mexico to the level of a civil war” (p. 1520).