You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Rough peace : understanding the avoidance of armed conflict in Bolivia Prest, Stewart 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2016_february_prest_stewart.pdf [ 3.42MB ]
JSON: 24-1.0221374.json
JSON-LD: 24-1.0221374-ld.json
RDF/XML (Pretty): 24-1.0221374-rdf.xml
RDF/JSON: 24-1.0221374-rdf.json
Turtle: 24-1.0221374-turtle.txt
N-Triples: 24-1.0221374-rdf-ntriples.txt
Original Record: 24-1.0221374-source.json
Full Text

Full Text

ROUGH PEACE: UNDERSTANDING THE AVOIDANCE OF ARMED CONFLICT IN BOLIVIA by Stewart Prest  M.A. Carleton University, 2006 B.Sc., B.A. Augustana University College; now Augustana Campus, University of Alberta (2000, 2001)  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (POLITICAL SCIENCE)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2015  © Stewart Prest, 2015 ii  Abstract  Why and how does peace persist in some contentious political contexts, but not others? I argue that certain forms of locally embedded governance institutions can play an important role in mitigating the likelihood of armed violence. Specifically, I find that inclusive communities equipped with governance institutions capable of resolving collective action problems—which I refer to as “Ostromian communities”—are, under a range of conditions, less likely to engage in armed conflict with other communities or the state. The research employs a method of process tracing on the basis of 70 participant and expert interviews, primary document collection and analysis and other archival research in the primary case of contestation over coca eradication in the Chapare region of Bolivia from 1982 until 2004. First, I illustrate how the coca growers’ federations constituted an Ostromian community. Second, I show how the federations’ inclusive political institutions encouraged and enabled high levels of coordinated and contentious non-violent political activity, but stopped short of armed resistance. This outcome resulted despite a repressive state presence in the region and regular instances of violence directed against the community occurring over a period of more than two decades. The dissertation makes several contributions to the civil conflict literature. It provides a novel explanation for why and how some countries at risk of civil conflict—such as those with unconsolidated political regimes or limited state capacities—tend to persist indefinitely in a state of rough, yet durable peace, while others experience conflict.  iii  Preface This dissertation is an original intellectual product of the author, Stewart Prest. The fieldwork reported in chapters three and four was covered by UBC Ethics Certificate number H11-02137. iv  Table of Contents  Abstract .......................................................................................................................................... ii	Preface ........................................................................................................................................... iii	Table of Contents ......................................................................................................................... iv	List of Tables .............................................................................................................................. viii	List of Figures ............................................................................................................................... ix	List of Abbreviations ..................................................................................................................... x	Glossary ...................................................................................................................................... xiv	Acknowledgements ................................................................................................................... xvii	Dedication ..................................................................................................................................... xx	Chapter 1: Introduction to the Argument ...................................................................................1	1.1	 State of the Scholarly Debate ............................................................................................. 2	1.2	 The Argument .................................................................................................................... 3	1.3	 Contributions ................................................................................................................... 12	1.4	 Case Selection and Methodology ..................................................................................... 13	1.4.1	 Evidence .................................................................................................................... 15	1.5	 The Chapare Region of Bolivia: A Case Ripe for Conflict ............................................. 18	1.5.1	 Historical and Contextual Factors ............................................................................. 23	1.5.2	 Towards an Explanation of Contention in the Chapare ............................................ 25	1.6	 Conclusion ....................................................................................................................... 27	1.6.1	 Dissertation Roadmap ............................................................................................... 28	v  Chapter 2: Ostromian Communities and Conflict: Concepts, Theory, and Rival Explanations .................................................................................................................................30	2.1	 Theoretical Specification ................................................................................................. 31	2.1.1	 Concepts: Conventional Politics, Contentious Politics, and Conflict ....................... 31	2.2	 The Meso Level ............................................................................................................... 36	2.3	 The Ostromian Community ............................................................................................. 37	2.3.1	 What OCs are…and Are Not .................................................................................... 48	2.4	 Extensions, Scope Conditions, and Caveats .................................................................... 50	2.4.1	 Rational Choices, Contingent Identities ................................................................... 50	 The Theoretical Context: Ideas in Conflict ........................................................ 51	 Ideas and OCs .................................................................................................... 53	2.4.2	 Scope Conditions: The Contingent Promise of Ostromian Communities ................ 53	2.4.3	 Extending and Integrating OC Theory ...................................................................... 56	2.5	 Alternative Explanations .................................................................................................. 57	2.6	 Conclusion ....................................................................................................................... 60	Chapter 3: The Chapare Cocalero Federations—An Ostromian Community ......................62	3.1	 Introduction ...................................................................................................................... 62	3.2	 Descriptive Theory ........................................................................................................... 63	3.3	 Operationalizing the OC .................................................................................................. 64	3.3.1	 Foundational Elements of an OC .............................................................................. 65	3.3.2	 Governance and Relational Elements of an OC ....................................................... 68	3.4	 Explaining the Emergence of the Chapare OC ................................................................ 71	3.5	 The Primary Case: The Chapare ...................................................................................... 74	vi  3.5.1	 Foundational Constituent Elements .......................................................................... 76	3.5.2	 Governance Constituent Elements ............................................................................ 95	 Collective Deliberative and Decision-making Ability ....................................... 95	 Enforcement and Oversight ............................................................................. 104	3.5.3	 Relational Constituent Elements ............................................................................. 110	 Independence from the State ............................................................................ 113	3.5.4	 Conclusion .............................................................................................................. 117	Chapter 4: Defending the Chapare—OC in Contention ........................................................118	4.1	 Introduction .................................................................................................................... 118	4.2	 Historical Background: Coca and Cocaleros in the Chapare ......................................... 122	4.2.1	 Economics of Coca in the Chapare ......................................................................... 125	4.2.2	 Symbolism of Coca ................................................................................................. 128	4.2.3	 Arc of State Involvement in Coca ........................................................................... 130	 Dictatorships 1964-1981: From Benign Neglect to Avarice ........................... 130	 The 1980s: Democratization and a Reluctant Hardening ................................ 133	 1990s and the Turn to Militarization ............................................................... 135	 Maximal Repression under Banzer and Successors, 1997-2003 ..................... 140	 Dénouement: Four Presidents and a Peace Agreement ................................... 144	4.3	 Subcases: Key Processes Defining Cocalero Response ................................................. 146	4.3.1	 The Unification of the Cocaleros ............................................................................ 148	4.3.2	 The Associations: A Challenge to Cocalero Unity ................................................. 159	4.3.3	 Cocalero Culture: Winning the Fight for the Political Instrument ......................... 163	4.3.4	 The Task of Resource Mobilization ........................................................................ 175	vii  4.3.5	 The Creation and Deployment of Self-Defence Committees ................................. 179	4.3.6	 Contested Contention: State and Cocalero Efforts to Frame the Conflict .............. 191	4.4	 Summary ........................................................................................................................ 196	4.5	 Comparing Alternative Arguments ................................................................................ 198	Chapter 5: Conclusion: Contributions, Comparisons, and Next Steps ................................201	5.1	 Scope conditions: Specifying the “Possible Cases” ....................................................... 203	5.2	 Bolivia: Democracy with Qualifiers .............................................................................. 206	5.3	 Extending the Theory I: Regional Shadow Case of Peru .............................................. 214	5.3.1	 The Role of Repression ........................................................................................... 219	5.4	 Extending the Theory II: Theoretical Complementarity ................................................ 221	5.5	 Next Steps: Further Verification and Extension of the Theory ..................................... 224	5.6	 Contributions ................................................................................................................. 226	Bibliography ...............................................................................................................................229	Appendices ..................................................................................................................................256	Appendix A Resources Used .................................................................................................. 256	A.1	 Interviews .................................................................................................................. 256	A.2	 Primary Documents Collected .................................................................................. 259	 viii  List of Tables   Table 2.1: Alternative Explanations ............................................................................................. 59	Table 3.1: Foundational Constituent Elements of an OC ............................................................. 65	Table 3.2: Deliberative Decision-making Constituent Element of an OC .................................... 68	Table 3.3: Monitoring and Enforcement Constituent Elements of an OC .................................... 69	Table 3.4: Relational Constituent Elements of an OC .................................................................. 70	Table 3.5: List of Coca Federations of the Chapare ..................................................................... 74	Table 3.6: Evaluation of Foundational OC Constituent Elements ................................................ 94	Table 3.7: Evaluation of Governance OC Constituent Elements (Deliberative Decisions) ....... 103	Table 3.8: Evaluation of Governance OC Constituent Elements (Enforcement and Oversight) 108	Table 3.9: Evaluation of Relational OC Constituent Elements .................................................. 116	Table 4.1: IPSP Share of Valid Votes Cast, Municipal Elections .............................................. 195	Table 4.2: Evaluation of Alternative Explanations ..................................................................... 198	 ix  List of Figures   Figure 3.1: Map of the Six Federations of the Chapare. ............................................................... 75	Figure 4.1: Regional Deaths by Presidential Period ................................................................... 141	Figure 5.1: Percent of Respondents Agreeing Democracy is the Best form of Government ..... 208	 x  List of Abbreviations  ANAPCOCA Asociación Nacional de Productores de la Hoja de Coca (National Association of Coca Producers) ADN   Acción Democrática Nacionalista (Nationalist Democratic Action) CEDIB Centro de Documentación e Información Bolivia (Central Statistics Agency of Bolivia) CERES El Centro de Estudios de la Realidad Economica y Social (Cochabamba) (Centre for Economic and Social Studies, Cochabamba) CNCB Confederación Sindical Única de Trabajadores Campesinos de Bolivia (Unified Confederation of Rural Workers of Bolivia) COB   Central Obrera Boliviana (Bolivian Workers Central) CPR   Common Resource Pools CSCB   Confederación Sindical de Colonizadores de Bolivia (The Union Confederation of Colonizers of Bolivia or Syndical Federation of Bolivian Colonists) CSUTCB  Confederación Sindical Única de Trabajadores Campesinos de Bolivia (Syndical Confederation of Rural Workers of Bolivia) DEA   Drug Enforcement Agency (United States) DIRECO Dirección de Reducción y Control de la Coca (Office of Coca Control and Reduction)  EGTK   Ejército Guerrillero Tupac Katari (Tupac Katari Guerrilla Army) xi  ELN Ejection de Liberation Nacional de Bolivia or the 1967 Ñancahuazu Guerrilla (National Liberation Army of Bolivia) FAMA Federation Sindical Agropecuaria Mamore (Mamore Syndical Agricultural Federation) FAMBB Federación Agraria Mamoré Bulo Bulo (Agrarian Federation of Mamoré Bulo Bulo) FECCH Federación Especial de Colonizadores de Chimoré (Special Federation of Colonizers of Chimore) FECCT Federación Especial de Colonizadores de Carrasco Tropical (Federation of Colonizers of Tropical Carrasco) FETCTC Federación Especial de Trabajadores Campesinos del Trópico de Cochabamba (Special Federation of Peasant Workers of the Tropics of Cochabamba) FEYCH Federación Especial de Yungas-Chapare (Special Federation of the Yungas-Chapare) FUCU Federación Unica de Centrales Unidas Campesino (United Federation of the United Syndicates) FELCN Fuerza Especial de Lucha Contra el Narcotráfico (Special Force for the Fight against Drug Trafficking) FTC Fuerza de Tarea Conjunta (Joint Task Force) FTE   Fuerza de Tarea Expedicionaria (Bolivian Expeditionary Task Force)  xii  FZTYC Federación de Zonas Tradicionales Yungas Chapare (Federation of the Traditional Zones of Yungas Chapare IMF International Monetary Fund INCB Instituto Nacional de Colonización de Bolivia (National Institute of Colonization of Bolivia) IU   Iziquierda Unida (United Left) MAS-IPSP Movimiento al Socialismo–Instrumento Político por la Soberanía de los Pueblos (Movement for Socialism–Political Instrument for the Sovereignty of the Peoples) MAS-U  Movimiento al Socialismo – Unzaguista MNR Movimiento Nacionalista Revolucionario (Revolutionary Nationalist Movement) MIR Movimiento de la Izquierda Revolucionaria - Nueva Mayoría (Revolutionary Left Movement) MST   Movimiento Sin Tierra (Landless Peasant Movement) NEP   New Economic Policy OTB   Organizaciones Territoriales de Base (Territorial Base Organizations) OC   Ostromian Communities PRODES Proyecto de Desarrollo Chapare Yungas (Chapare-Yungas Development Project) SAP   Structural Adjustment Program SDC   Self-defense committee  xiii  SUBDESAL Subsecretaría de Desarrollo Alternativo (Subsectretariat for Alternative Development) TIPNIS Territorio Indígena y Parque Nacional Isiboro-Sécure (Isiboro Secure National Park and Indigenous Territory) UNODC  United Nations Office on Drug and Crime UMOPAR Unidad Móvil Policial para Áreas Rurales (Mobile Police Unit for Rural Areas)  xiv  Glossary  Allyu: Traditional form of community among Aymara and Quechua found throughout the Andes. Apu Mallku: Leader of the system of Ayllus, selected by a council representing parts of Bolivia, Peru, Chile, and Argentina. Current incumbent is Evo Morales. La Confederación Nacional de Mujeres Campesinas Indígenas Originarias de Bolivia “Bartolina Sisa”: National organization representing indigenous women’s groups in Bolivia named for an 18th century Aymara leader of that name. Campesino: Small scale farmer practicing a mix of subsistence farming and production for sale. Cato: Standard plot of coca, measuring 1600m2. By agreement with President Carlos Mesa in 2004, families in the Chapare were permitted to maintain a cato of coca each. Central: Organizational component of the cocalero organizational structure between federations and syndicates. Each central groups together about ten union locals, on average. Comisiones de Vigilancia (vigilance committee): Oversight committees tasked with reviewing expenditures and performance of the executive within the federations’ organizational structure. Comisiones Organica: Committees responsible for maintaining the federations’ organizational structure, including the adjudication of conflicts between component units. Coordinadora: Committee responsible for coordinating activities of all cocalero federations. Corregidor / secretario de conflicto (magistrate / conflict secretary): Executive official existing at most levels of organization within the cocalero structure responsible for adjudicating conflicts within the community. Decreto Supremo 21060 (Supreme Decree 21060): 1985 law enacting sweeping economic changes intended to end the financial crisis facing Bolivia in the early 1980s. Key components xv  included a fixed currency, the privatization or closure of state-owned enterprises, market liberalization, and an end to subsidies on numerous goods. Defensor del Pueblo: Independent ombuds officer appointed by the Bolivian state, responsible for monitoring activities of state officials, and investigating and documenting reported violations of human and civil rights. Ecological Police: Bolivian anti-drug police force focused on seed and seedling eradication. Federación (Federation): Groupings of multiple centrals together as a single political unit. Ley 1008 / Law 1008: the Regulation of Coca and Controlled Substances. This law declared that the Chapare was to be a “zone of transition” in which coca production would eventually be eliminated. Pacto Militar-Campesino (Military-Campesino Pact): Arrangement through which rural Bolivians supported the military against leftist subversion during the first decade of dictatorship in exchange for state subsidies and an agreement to uphold gains achieved during the land reform that followed the 1952 social revolution. Pijcheo: The traditional practice of chewing coca leaves. Plan Dignidad: Coca reduction strategy introduced by the government of President Hugo Banzer in 1997. Included an emphasis on forced eradication, with reimbursements for producers declining over time. Seis Federaciones del Trópico de Cochabamba (Six Federations of the Tropics of Cochabamba): The campesino organizations representing cocaleros and other producers in the Chapare region of Bolivia. xvi  Seminarios / seminars: Regular syndicate meetings in which presentations are made to group members. During the years of contention, these were often used to reinforce different points regarding the history and strategy of resistance. Sindicato (syndicate / union): basic organizational unit of the Chapare cocalero federations, consisting of about 50 families on average. Many initially founded as colonies during the initial period of regional settlement.   xvii  Acknowledgements  Thanks first go to my doctoral committee co-supervisors, Brian Job and Antje Ellermann. Each step of the way—from vague idea, to evolving field research project, to finished dissertation—I benefited from their guidance and advice, and their kind yet implacable questioning. Both Brian and Antje showed unwavering belief in my ability to do the work despite all the bumps in the road. Max Cameron, the third member of the committee was also a treasured resource, providing the right balance of supportive comments and useful critique. All three have gone out of their way to help me from my first months as a PhD student; I am and always will be grateful for their support. Lisa Sundstrom, Gastón Gordillo, and Erica Chenoweth all gave their expertise and time to read and evaluate the finished work as external examiners. I thank them for this, and for the valuable comments made during the examination process. They have given me new insight into my own work, and have pointed me towards intriguing new potential avenues of research. In La Paz, I thank Luz Pacheco for opening her home to me during my first few months of fieldwork, and giving me valuable information regarding the right way to get research done in La Paz. I am grateful as well for all the support provided by María Renée Bejarano, who was a good friend to our family long before the dissertation began. I owe a debt of thanks to both Lola Parades and the library at the Universidad de la Cordillera for all sorts of help with my archival research, and the staff at the Archivo de La Paz for access to the carefully curated newspaper archives there.  In Cochabamba, Godofredo Reinecke and Kathryn Ledebur were among my earliest points of contact in the area, and both were very generous with their time. I benefited greatly from their insight. Fernando Mayorga and Fernando Salazar Ortuño of the Universidad Mayor de xviii  San Simon were tremendously welcoming and helpful throughout my time in the region. Fernando S. in particular was indefatigable in his support for my research, and much of what I was able to accomplish here depended on his own prior research. Jose Luis Mamani acted as guide and occasional interpreter in the Chapare, no easy task. The dissertation would not have been possible in this form without his assistance. My two research assistants, Erika Higueras Mendieta and Saúl Sejas Escobar were indefatigable in their archival and, in Erika's case, her transcription work. I also thank Huascar Valverde as well for his Quecha translation and transcription. I received much in the way of research support from Roberto Laserna and the staff at CERES' Cowork Café, without a doubt one of the best-kept secrets (if secret it still remains) in Cochabamba. Thanks to CERES as well for kind permission to use its unique data on social conflict. I am grateful to Cecilia Illanes for her help in making use of the CEDIB archives in Cochabamba as well; CEDIB remains an unparalleled resource for anyone planning to study subjects related to coca. In the Chapare, my thanks go to the many producers there who shared their time and thoughts with me, and to federation leaders who helped me gain access to the region. I was truly fortunate to encounter such open and welcoming people so eager to tell their story — a story of triumph over long odds, at considerable cost. I hope I have done justice to their conversations with me, and the trust that those conversations represented. Several Canadian and American scholars and experts lent me their time and expertise as well, for which I am grateful. John Cameron, Linda Farthing and the late Ben Kohl all helped me while in Bolivia, as did Anne Catherine Bajard back in Canada.  I wish to thank the Social Sciences and Humanities Research Council, the International Development Research Centre, and the Security and Defence Forum, and the Liu Institute for xix  contributions to this research endeavour. Additional thanks to the University of British Columbia, the Faculty of Arts, and most particularly the Department of Political Science, its faculty, staff, and students for providing a wonderful environment in which to study, and supporting my work in ways to numerous to mention.  The last and largest acknowledgements go to my family. Thanks to my parents for being so supportive in absolutely everything. All that I am I owe to you. Special thanks go to my Dad, for his just-in-time proofreading. Thanks to Wendy and Marty for their cheerful willingness to help out with childcare. Above all, thank you to my wife Manuela, for the innumerable ways you have supported me in getting this dissertation out the door. Finally, thanks to Sidney and Benjamin for giving me such a good reason to work hard, and an even better reason to put the work aside at the end of each day.  Notwithstanding the invaluable contributions of all named above, the work and conclusions that follow are my own, as are all the errors.  Vancouver December 2015  xx  Dedication  To Manuela, Sidney, and Benjamin.  Smushed it!1  Chapter 1: Introduction to the Argument  On 27 June 1988, a confrontation between Bolivian cocaleros—coca farmers—and the country’s anti-drug police erupted in violence in the town of Villa Tunari. About a dozen campesinos lost their lives, with more than twice that number either injured or arrested, or both (Lee 1988: 95, Salazar Ortuño et al. 2008: 142-143). Some drowned, jumping off a cliff into a nearby river to escape the violence. The event marked the effective beginning of two decades of highly contested confrontational politics between the state and the region’s cocaleros and their allies. Given the presence of so many factors identified as conducive to insurgency within the literature, ranging from material factors such as rugged terrain, a local economy significantly shaped by the production of an easily looted and highly profitable natural resource, limited government capacity in general and specifically in the region and the recurrent use of violence by state forces in suppression of political protest against the policy of eradication to ideational factors including sharp class and ethnic divisions between state policymakers and the region’s campesinos, and recently rediscovered and reinvigorated cultural tradition of indigenous campesino activism—including strains of resistance to the state—there was good reason to expect the emergence of some attempt at violent resistance or insurgency. Yet this is not what happened. The cocaleros remained strongly united in pursuit of dual strategy of conventional political competition and non-conventional political protest. What explains such cohesion and commitment to non-violence in the face of continuing state repression?   2  1.1 State of the Scholarly Debate Before answering that question in greater detail, I take a step back to provide some theoretical context. In recent years we have seen a “social turn” in the study of civil conflict and civil war. Emphasis has shifted from consideration of the state, to the choices and behaviours of non-state actors in explaining conflict dynamics and outcomes in a given context. The move has produced innovative research, from Elisabeth Wood’s (2003) study of evolving norms and social networks among Salvadorian peasants and insurgents to Paul Staniland’s (2012) analysis of how community organization influences the dynamics of violence in the context of civil conflict. More generally it has produced significant new insights regarding the interrelationships between internal rebel organization and behaviour, as well as the pre-existing and evolving social contexts within which they are embedded (e.g. Weinstein 2007, Mampilly 2011, Parkinson 2013).  However, the full potential of the move has not yet been realized, as the social turn has perpetuated old silences alongside its new insights. In particular, studies to date have replicated a larger tendency in the study of conflict in recent years to privilege the study of conflict occurrence over its avoidance. That is, researchers study why and how conflict breaks out, along with its ensuing dynamics and effects. Questions regarding where, how, and why peace persists—particularly in seemingly unstable contexts—remain understudied, notwithstanding the fact that globally, peace is by far the more common outcome. It may not always be pretty, yet most countries muddle through, most of the time. Intuitively, one would expect some relationship to exist between community organization and the decision to rebel, just as Staniland (2012) and others have identified a logic linking community organization to rebel behaviour once conflict has broken out. For instance, one would anticipate that atomized and fragmented communities behave very differently than those 3  with strong social networks or inclusive local institutions of governance when faced with violent external pressure. Indeed, studies of conflict consistently highlight factors such as political grievance, social fragmentation, and alienation in accounting for individuals’ and groups’ decision to engage and/or participate in rebellion (e.g. Walter 2004, Humphreys and Weinstein 2008, Cunningham et al. 2012, Cunningham 2013). Despite this, we lack both the theory and evidence to make and rigorously defend the converse case, namely that strong local institutions can, under certain circumstances, mitigate or even prevent conflict through their effects on communities’ internal decision-making and decision enforcement processes. It is precisely this that I argue accounts for the observed outcome in the Chapare. 1.2 The Argument In this dissertation, I advance a set of interrelated claims arguing that the ways in which communities resolve (or fail to resolve) the significant social dilemmas that they face—collective action problems of various kinds—shape their interactions with extracommunal actors. As a specific example of this effect, I advance the concept of what I refer to as “Ostromian communities” (OCs) so called in recognition of Elinor Ostrom’s (1990) work articulating the characteristics of an important subclass of such social communities: those associated with the shared administration of common resource pools. As I use it, the term refers not only to instances of resource management, but more broadly to all communities featuring locally embedded and inclusive institutional solutions to significant social dilemmas that include mechanisms to enforce group decisions. I argue that OCs (1) support rational group-level decision-making in the pursuit of shared community interests. Given their ability to effectively mobilize resources in the service of those interests they (2) engage in high levels of conventional and nonconventional political 4  contestation across a range of social arenas with both the state and other communities. Likewise they (3) avoid engagement in violent conflict in defending those interests provided there remain other, more utile options. Only in rare situations, those characterized by lack of less costly alternative strategies of resistance, do OCs resort to violence.  As such, I offer a “meso-level” explanation for conflict, one that provides a detailed mapping of a particular set of social phenomena that shape relations between actors encountered at the “micro” (individual and small groups) and “macro” (country and state) levels of analysis. Accordingly, as with other recent research on civil conflict, this dissertation offers a potential bridge between the poorly integrated research on civil conflict undertaken at the micro and macro levels. While I offered more detailed discussion of the macro/meso/micro trichotomy in Chapter Two, for the moment one may understand the meso level as the set of formal and informal social and political structures, processes, mechanisms, and relationships linking the small groups and individuals populating the “micro” level of interaction with state level political institutions and processes. As I use it here, the OC constitutes a meso-level institution that serves to organize cocaleros’ political and social lives, representing them to the state. As Straus (2012: 347) puts it, “at the meso level, an important game-theoretic insight is that when groups interact with other groups repeatedly over time, it is in their interest to develop formal and informal institutions to facilitate inter-communal, mutually-beneficial cooperation.” One may achieve similar such insights via alternative theoretical approaches as well; the concept of stable social actors—including those other than the state at one extreme, and the individual or small group on the other—constitute a core concept within constructivist theory as well, employing concepts of habit and a “logic of appropriateness” to explain how such stable entities emerge and shape the interactions of individuals with one another, with other social actors, and with the state (March 5  and Olson 1995, 1998). As such, it provides a bridge between individual motivations of grievance, anomie, need for participation, and so forth, and the national political institutions and processes that decisively shape politics owing to the state’s unique role in political life. As such, it shapes the interactions between state and social actors in significant and, as I argue below, predictable ways. More concretely in the context of the literature on civil conflict, one might think of meso level entities and process as providing a potential frame through which to organize our disparate understandings of a range of phenomena in which social actors emerge and come to significantly shape human action, providing a stable frame in which to theorize everything from ethnic and religious contention and conflict (Fearon and Laitin 1996) to the involvement of various NGOs, militias, and so on within the context of conflict. As such, much of the most interesting research currently ongoing in the field of civil conflict is concerned with understanding entities and interactions taking place at the meso level, ranging from Cunningham et al.’s (2012) study of fragmentation in rebel groups, to Staniland’s (2012) study of social networks in Kashmir and elsewhere in South Asia (2012), to Shesterinina’s (2014) study of normatively constituted mobilization networks in Abkhazia, to studies of the differential effects of different forms of rebel organization on conflict outcomes by Weinstein (2007), Cohen (2013), and others. OCs, I argue, constitute one such kind of meso-level entity. As such, they shape intercommunal interactions in important ways in the contexts in which they are found. This implies a certain set of scope conditions—the possible cases of conflict, to use the terminology of the “possibility principle” developed by Mahoney and Goertz (2004)—over which a theory of 6  local institutions and conflict may be expected to have an effect.1 Simply put, cases of overwhelming repressive state capacity, those in which the state can (and perhaps does) overwhelm any local organization through the deployment of security resources, are beyond the scope of this study. Likewise, such a theory has little to offer in the explanation of peace in consolidated democracies, where conflict is managed through a multiplicity of formal and informal processes and institutions, where robust civil peace is over-determined.  Rather, the theory of OCs provides leverage in explaining outcomes in the set of “possible cases” existing between those over-determined poles, where the state is not too powerful to deter or repress all would-be insurgents, and democratic institutions are not so deeply entrenched as to render violent resistance unthinkable. One way to conceive of this is to say I am interested in cases at and around the inflection point of what others in both political science and sociology have referred to as the “inverted U” (Hegre 2001, Davenport 2007, Fein 1995, Muller 1985). Imagine a curve illustrating the relationship between conflict and regime type. According to the inverted U hypothesis, the “anocracies,” the unconsolidated or otherwise underdeveloped states at the centre of the graph, experience far more conflict than the consolidated democracies and autocracies at either end of the spectrum.2 Though a meso-level entity, I argue the OCs have notable effects at both the macro and micro levels. With regard to the former, while a significant portion of the literature focuses on the effects of democracy on state actions—notably in constraining the executive—I widen the                                                 1 As Mahoney and Goertz put it, “the Possibility Principle holds that only cases where the outcome of interest is possible should be included in the set of negative cases; cases where the outcome is impossible should be relegated to a set of uninformative and hence irrelevant observations.” (2004: 653). 2 While debate continues as to the precise explanation for the phenomenon (Hegre 2014), and some question the extent to which the “inverted U” hypothesis is capturing an artifact of regime coding in the Polity IV and Freedom House datasets (the most widely used global measures of regime type), repeated studies support the conclusion that there is some “there” there (Vreeland 2008).  7  discussion to examine how the presence of democracy affects the decisions of social actors in opposition to the state.3 I draw attention to the normative effects of democracy on civil peace, drawing on Fearon (1995) in doing so, albeit with a novel mechanism. Specifically, as effective unitary rational actors, OCs will tend to search for preferred outcomes; given the costliness of conflict, they are as keen to avoid violence as any of Fearon’s realist billiard ball states, their lack of state-ness notwithstanding. Contrary to Fearon (1994 and 1995), Hegre 2014, Lacina 2006, and others writing on the subject however, it is not the presence of effective commitment mechanisms providing OCs with confidence in present agreements reached that prevents conflict. On the contrary, the Chapare case is littered with instances of negotiation and agreement between the state and OC representatives, only to see the terms violated by one side or the other (more often than not, the more powerful state) within days or weeks. To cite just one example, in the months preceding passage of the “Ley 1008” or Law 1008 in 1988—a significant piece of legislation that, among other things, rendered coca production illegal in much of the Chapare—the state and representatives of coca growers negotiated a halt in the legislative passage of the bill pending a report from a “Tripartite Commission” charged with outlining a consensus-based path forward with regard to coca regulation. The commission never fulfilled that role however, and the state passed the new drug regime in June 1988, setting the stage for the nearly two decades of contention over the right to grow coca that was to come.                                                  3 Partial exceptions include Humphreys and Weinstein (2008) who focus on micro-motivations for agents’ participation in violence; Walter (2004), who presents a theory linking political participation to conflict avoidance, albeit one grounded solely in macro-level data; and the long-recognized, but recently revitalized understanding that democratic institutions act solutions to commitment problems among state and would-be social combatants (Acemoglu and Robinson 2005, North et al. 2009, Lacina 2006, and World Development Report 2011). 8  The salutary effects of democracy in the case were normative, diffuse, and future-oriented, rather than specific, institutional, and related to present agreements, and not contingent on any particular sense of trust in the word of the government on any particular issue. Rather, it was the presence of strong (i.e. credible) commitments to democracy within the larger society and state representatives, and to the idea of democratic processes as potentially viable mechanisms for representation and conflict resolution in the future—in the face of deep flaws in those institutions at the time of conflict significantly impairing their representativeness of indigenous and rural Bolivians including campesinos—that encouraged cocaleros to consider, adopt, and subsequently adhere to a strategy of formal democratic competition. Simply put, the presence of a local OC made possible cocalero acceptance of, and participation in, Bolivia’s imperfect democratic institutions, enabling a continued commitment to non-violent processes of conflict resolution. On the basis of a rational analysis of the problem that concluded that democratic channels offered the greatest likelihood of success over the long term. Though marginalized by the system as it was then constituted (Lucero 2008, Yashar 2005, Van Cott 2003) and faced with an interlocutor they considered untrustworthy, they continued to trust in the “idea” of democratic participation, and the long term prospects it offered. They subsequently developed and adhered to a strategy along those lines, even at the risk of significant short-term costs, and despite the presence of alternative approaches including violent ones. At the same time, the meso-level explanation of the OC has several significant micro-level effects, also tending to lessen the likelihood of conflict under conditions such as those described above. Once again, though grounded in rational analysis and action on the part of the cocaleros as a group, the mechanisms were in part normative. Specifically, I argue that the OC offers a powerful opportunity to participate in politics in a way that members find meaningful, 9  even as they remain marginalized in the country’s larger political processes. Accordingly, comparatively few seek out other, possibly more radical and violent methods of participation. Further, as we see in the case of the Chapare at least, the OC proved sufficiently robust to control or deter any so motivated once settled on a particular dual strategy of conventional democratic participation and non-conventional political resistance. This finding is not unique. Several other studies have highlighted the importance of participation in shaping individuals’ approach to conflict. Wood (2003) for instance observes the importance of the “pleasure of agency” in shaping Salvadorans’ decisions to participate in, or support insurgents fighting the repressive, reactionary state during that country’s civil war. Likewise, Walter (2004) employs country-level data to describe the importance of “political participation” in shaping the likelihood of resumption of conflict in post-war cases. Humphreys and Weinstein (2008) in turn provide powerful support of a similar hypothesis on the basis of within-case survey data of former combatants in Sierra Leone. For them, the decision to fight is often marked by isolation from political organizations in the country. In channelling that participatory drive, OCs significantly reduce the space for other political entrepreneurs to operate and win followers for more radical and potentially violent projects of resistance. In sum, I argue that the presence of an inclusive OC in a possible case of civil conflict (i.e. in a region in which armed resistance is feasible and potentially justifiable due to the perception of grievance, or subject to what one might refer to as conflict pressures) serves to reduce —though not eliminate—the possibility of armed violence by several mechanisms acting at macro, meso, and micro levels–terms that I define more precisely in the following chapter. In the context of the specific case of the Chapare, I find that coca growers, when faced with that concerted government campaign of eradication, consistently resisted any urge—10  occasionally expressed to me, and often, according to written accounts and previous statements by leaders to the media, to one another—to respond to government violence with violent escalatory actions of their own. They did so out of an interest-based, and institutionally enforced commitment to a long-term strategy of nonviolent resistance combined with electoral competition. Cocalero leaders and community representatives, coordinated through an inclusive umbrella labour organization known as the Seis Federaciones del Trópico de Cochabamba, focused in particular upon the presence of flawed, but from the cocaleros’ perspective ultimately salvageable, national democratic institutions. Though Bolivian democracy was procedurally biased to favour the interests of the country’s urban mestizos, the rural indigenous population in the Chapare and elsewhere in the country concluded after much deliberation that they could wrest political control from traditional elites via the ballot box. The did so in the face of longstanding cultural norms and formal institutions militating against the direct involvement of indigenous campesino Bolivians in the country’s political life.4,5 This commitment persisted through intense periods of state repression, as the democratic state regime maintained a military presence in the region and engaged in periodic violent campaigns of forced coca eradication.  Eventually the dual strategy of electoral activism and nonviolent resistance proved successful, with cocaleros regaining the right to grow coca legally. My research traces how the structure of the community’s representative institutions—previously established to fulfil the                                                 4 As discussed in Chapters Three and Four, in addition to formal political organization the cocaleros’ repertoire included marches, blockades, strikes and other similar actions relying on a capacity for periodic large-scale mobilization (Tilly 2003), along with local level “self-defence” committees capable of resisting specific incursions by state forces to undertake projects such as coca eradication. 5 Explaining that confidence is one of the tasks taken up in Chapter Four. In brief, the country’s previous experience with military rule, including in particular the economically and politically disastrous way that era ended, combined with the decisive role played by civil society in bringing about that end, gave most Bolivians both severe distaste for nondemocratic governance (a fact borne out in opinion surveys throughout era in question) and confidence that the civilian population could and would collectively and successfully rise in opposition to any subsequent attempt to implement a nondemocratic power transition. 11  basic functions of governance in a region almost completely neglected by the central government—played a pivotal role in shaping the pattern of campesino response to the eradication program. It did so in a variety of ways; some were purely interest-based, while others, more normatively constituted, emerged over time. As an example of the latter, campesinos over the course of the confrontation came to include “nonviolent” as an element of their own political identities, something that clearly distinguished them from the violent state agents they opposed. In effect, a particular form of polarized identity emerged that actually served to reduce the likelihood of violent escalation; at the very least, it would have taken some shock, a development sufficiently strong to shake the shared socialized cocalero identity of peaceful resister opposing a capricious and violent state dominated by imperialist “Yanquis”.  The same institutional mechanism, conditional on the surrounding context, accounts for all these outcomes. As Ostrom (1990) and others (e.g. Wade 1994, Agrawal 2001) document, effective shared solutions to significant collective action problems imply and require the presence of a certain set of local institutions with capacities to, among other things, reach shared decisions and mobilize the resources necessary to achieve them. By bringing together members of the community, giving them a role in organizational decision-making processes, and ensuring subsequent support for communal decisions through some form of “contingent compliance” (Levi 1989)—including mechanisms for policing both leaders and members—OCs are well equipped to act strategically in pursuit of shared interests; as rational agents, more often than not that includes the avoidance of violent conflict and the ex post costs it generates (Fearon 1995).6                                                 6 One area for future research concerns whether and to what extent the concept of “Ostromian community” as articulated here is complementary with North’s concept of an “open access regime” (North, Wallis and Weingast 2009). 12  1.3 Contributions Beyond the immediate value that arises from the careful explanation of an important, yet understudied case of non-violent contention, this dissertation makes several discrete contributions to the study of civil conflict and related scholarly traditions. Through the development of the novel general concept of the Ostromian community, the argument described in detail below furthers the study of conflict in a number of ways: First, in expanding the focus beyond cases of conflict to include the close reading of a case in which violent conflict was eminently possible yet did not occur—a negative case, or a dog that did not bark—the research contributes to a better understanding of the full universe of relevant cases, correcting an ongoing selection bias in the study of conflict (Mahoney and Goertz 2004). Second, the comparative study of Ostromian communities and other forms of collective action furthers the social turn described above by providing a causally significant, widely occurring, and easily recognizable ontological point of reference around which one may construct a general “meso level” theory of peace and conflict. As such, it has the potential to subsume a variety of other relevant factors—such as ethnicity, religion, language groupings, income and class divisions, and so on—that have been used for this purpose to limited effect. There is some theoretical precedent for this already. Fearon and Laitin (1996) for instance provide a formal model detailing how internal communal structure can act as a damper on intercommunal conflict. While subsequent studies have identified mechanism by which, for example, ethnicity informs practices of specific and general reciprocity (Habyarimana et al. 2009), few have taken up the challenge of empirically linking communal-level practice to the onset or avoidance of civil conflict. Ostromian communities, and the study of social organization 13  more generally, thus provide a new point of leverage from which to integrate findings from a variety of research traditions. An institutional theory of communities in conflict can go beyond studying whether things like ethnicity and social fragmentation matter, and begin to uncover how and under what circumstances they do. In doing so, it informs the articulation of an alternative “meso” level approach to the study of peace and conflict. Finally, by establishing new theoretical connections between positive local development outcomes and security outcomes—specifically, that enhancements in communities’ internal capacity to carry out collective action may be linked to reductions in the likelihood of extra-communal violent conflict—the study speaks to questions of conflict prevention and management, and does so in a way that highlights the agency of local actors. As such, its findings will have relevance to both security and development policy communities. Events of the past decade have driven home just how problematic and limited international intervention in civil conflict can be. Accordingly, the prevention of such conflict takes on even greater normative weight, and any work that meaningfully adds to our understanding of how such conflicts may be avoided constitutes a significant and worthwhile endeavour. In the balance of this chapter, I provide a brief overview of the dissertation’s primary case, and an overview of the argument that I develop further in subsequent chapters.  1.4 Case Selection and Methodology The primary purpose of this dissertation is the identification of a novel theoretical account for a certain class of outcomes: namely, contentious, yet consistently nonviolent politics. Its use of the Chapare region of Bolivia as a primary case for investigation conforms to what George and Bennett (2005) refer to as “heuristic” case study, which they describe as being used to “inductively identify new variables, hypotheses, causal mechanism, and causal paths. ‘Deviant’ 14  or ‘outlier’ cases may be particularly useful for heuristic purposes, as by definition their outcomes are not what traditional theories would anticipate” (2005: 75).7 That is precisely the role played by the primary case of the Chapare here. As an outlier, it is a hard case for existing theory to explain, but by examining how contention emerged and evolved throughout the period of confrontation one may gain a better understanding of alternative explanatory approaches previously missed. More specifically, the case of the Chapare illustrates how inclusive and enforceable intracommunal institutions provide significant latent resources in the management of extracommunal conflict. In short, the well-organized cocalero community actively and ultimately successfully pursued its shared interests, but did so in ways that avoided the costs of violent conflict. They were able to do so, I argue, in large part thanks to the latent institutional capacities embodied in their federations’ syndical structure, which I characterize in Chapter Three as an example of a federated OC. As a dissertation reliant on qualitative data in the form of archival records, contemporary news reports, and in person interviews with participants and experts—supplemented where possible by quantitative data—it employs a primary research methodology of in case process tracing, one heavily dependent on, as Bennett and Checkel (2014) and George and Bennett (2005) describe, theory-laden analytical narrative. The present study employs a method of process tracing within the primary case. Though a buzzy term that has generated some misunderstandings in recent years, when properly applied the method is a powerful way to undertake small-n research in a way that lends itself to the cumulation of knowledge in cooperation with other literature on a given subject. In many ways it is reminiscent of the                                                 7 In Van Evera’s terms, it is a “hard case” for existing theory to explain (1997). 15  historical method (e.g. Trachtenberg 2009), only with a more formal, explicit, and rigorous relationship with competing theoretical explanations. In doing so, I follow Bennett and Checkel (2014: 7) in defining process tracing as The analysis of evidence on processes, sequences, and conjunctures of events within a case for the purposes of either developing or testing hypotheses about causal mechanisms that might causally explain the case.   One way to think of it is as a form of theory-laden analytical narrative. Its utility stems from the ability to capture specific within-case variation over time, and use the successive developments of the case, and the decisions of the actors within them, to provide a series of within case tests of competing theoretical explanations. The result is a more comprehensive account not just of why an outcome occurred, but how, and how the conclusions of that account can inform wider theoretical discussions regarding the social phenomena being explained. Given my focus on a particular heuristic case (using George and Bennett’s term) and particular interest in establishing the complementary effects of a set of social phenomena distinctly resistant to quantitative analysis, the context-sensitive approach of process tracing is a natural fit for the primary case study. It allows me to preserve much of the salient richness of the case while doing so in a manner that makes possible comparison against specific predictions generated by competing theories. Likewise, process tracing allows for a good fit between conceptual and empirical scope. In any comparative context—whether cross-sectional or longitudinal—some violence is done to concepts used in the name of generalizability.  1.4.1 Evidence  I have assembled an extensive set of materials for use in the primary case of the Chapare. They range from expert and participant interviews focusing on a complete account of a particular set of events, rather than on the creation of a representative sample of some population. 16  Accordingly, my participant identification method sought to incorporate diverse perspectives, and I continued to seek out interviews effectively until I reached a point of repetition, where successive interviews with other participants yielded variations on responses I had already heard. In the end the study included more than 70 interview participants, several of whom I spoke with on more than one occasion. The interviews themselves were semi-structured, and conducted in such a way as to allow participants some freedom to focus on what they themselves felt was most important to talk about while still ensuring the discussions remained focused on the general set of topics I was trying to explore and explain.  Interview recruitment proceeded using a modified form of snowball sampling. As much as possible I followed the procedure advocated by Scacco (n.d.), seeking out multiple snowball “seed” contacts in an effort to develop a variety of contact chains, resulting in a diversity of viewpoints. In practice this proved a challenge owing to the apparent denseness of intra-cocalero networks, particularly among those in current and former leadership positions. Many leaders knew, or at least knew of, one another. Most would not talk to me unless they had evidence that I had talked to someone else they trusted previously—effectively, respondents within the cocalero community drew confidence from being part of the same “snowball.” Even so, the diversity of viewpoints obtained ensured that the research captured a diverse set of perspectives from a range of epistemic communities connected with the region. I fsought and obtained multiple viewpoints across various meaningful social distinctions in the region. These included interviews with rural cocaleros, relatively rare non-coca farmers, town inhabitants sympathetic to the cocalero cause, former miners, longer-term inhabitants, and a small number of (confidentially conducted) interviews with regional inhabitants in some ways ambivalent towards, or opposed to the cocaleros’ project or methods; many of these were 17  mestizo urban Bolivians. Within the six federations, interviews include local, mid-level, and senior cocalero leaders, both past and present representatives, with some attention to gender balance (though female cocalero representatives often proved much more reluctant to speak, and more guarded in their responses). Finally, I sought representation across the Chapare including respondents from all six of the cocalero federations in the region. On the side of the state, I spoke with a limited number of officials still active in the region who had first-hand knowledge of events during the period of contention. This proved a far more challenging task, as the turnover since the period in question is almost complete. Few remain who were party to the conflict. I secured an interview with a coca assessor, a current police officer who at the time worked with an eradication team, and a former civil servant working in the area of coca regulation. Finally, I spoke with a range of more neutral actors, including foreign aid providers, researchers with Bolivian universities and not-for-profit research centres, and North American researchers with expertise in the region as well. Beyond interviews, the research builds on primary archival documents, including federation constitutions and biennial congress summaries, city council reports from the municipality of Villa Tunari, maps, and a limited number of artifacts, including campaign materials distributed by cocalero organizations during elections. I secured two different newspaper archives covering part or all of the period in question, and rendered both searchable using optical character recognition software. While coverage of rural events is far from perfect, the reports nonetheless capture most significant events, and provide valuable evidence regarding how the cocalero movement was perceived outside of the region at the time. In addition, I obtained a coded dataset of all conflictual events reported in national newspapers over the period 1970-2013, maintained by the not-for-profit research organization 18  CERES (El Centro de Estudios de la Realidad Económica y Social). The research also relies on corroborating evidence located in secondary literature on the subject produced by both Bolivian and western researchers. Most notably, Kawsachun Coca (translated as “Long Live Coca”) is a two volume accounting of the deaths and violence that occurred in the Chapare region over the course of the more than 20 years of confrontations between state and society. Based upon a combination of published accounts, interviews, and the records of the region’s Defensor del Pueblo8, the Asamblea Permanente de los Derechos Humanos de Cochabamba9, the Andean Information Network, and the Washington Office for Latin America (Salar Ortuño et al. 2008). 1.5 The Chapare Region of Bolivia: A Case Ripe for Conflict   Having asserted in general terms that Bolivia—and most particularly the Chapare region during the struggle over coca—is a challenging case for many leading theories to account for, I now develop that argument in more detail. Bolivia features an extremely diverse and rugged physical geography, ranging from the Andean mountains and the Altiplano in the west, to the Amazon basin in the north, to the great Bolivian Salt Flats in the south, to the Chaco scrub and Pantanal wetlands in the east, with the steep valleys of the Chapare and the lowland plains around Santa Cruz in between. It is sparsely populated as a result, with travel between regions challenging. Such forbidding terrain is linked with enhanced risk of conflict onset. Fearon and Laitin report an increased likelihood of conflict in countries with rough terrain (2003), while Kalyvas’ (2006) close study of the Greek civil war highlighted the extent to which obstacles to                                                 8 An arms-length ombuds-type office created by the country’s Department of Justice in order to undertake reporting, mediation, and other forms of good offices within the region, which also documented much of the violence that occurred during the course of events. 9 The Permanent Assembly for Human Rights of Cochabamba, a human rights observatory established during the era of dictatorship. 19  government control in certain areas were associated with high levels of violence.10 The Chapare itself is located within the country’s central valley region, a tropical and area that saw significant settlement beginning only in the 1950s following the state’s sweeping agrarian reform (Spedding 2004: 91-95) and opening of the region with new roads (Cole 1958, García Linera et al. 2005: 272).11 The main population centres huddle along the main road running from Cochabamba to Santa Cruz, but the majority of the population is rurally located.  In terms of its human geography, the country is also extremely diverse, with sharp (and often reinforcing) cleavages between rich and poor, mestizo and indigenous, urban and rural, even east and west (Webber 2010). The country boasts a strong and varied indigenous heritage, with more than half the country identifying as such. The indigenous populations are also divided, ranging from the Aymara- and Quechua-dominated highlands in the west of the country, to the distinct and diverse cultures found throughout the Amazon basin. The eastern half of Bolivia is comparatively rich due in part to both agribusiness and abundant oil and natural gas reserves; its inhabitants also often try to differentiate themselves from the indigenous cultural symbols associated with the western altiplano. In the broader literature various authors have identified a positive relationship between either polarization or horizontal inequality (i.e. inequality between groups) and civil conflict (e.g. Montalvo and Reynal-Querol (2005), Stewart 2002).  Authors such as Lake and Rothchild (1998) have highlighted the link between conflict in neighbouring countries, and conflict onset. Though it is less direct, Paul Collier has drawn attention to the multiple links between isolation (i.e. landlocked), poor governance, and poverty                                                 10 Specifically, Kalyvas (2006) argues that violence in conflict is highest in areas in which either government or rebel forces exercises partial control, with lower levels occurring in areas of either state or rebel dominance. 11 It is one of the complexities of Bolivian politics that many indigenous or originarios are actually from different areas of the country. The majority of the population in the Chapare for instance is indigenous, but very few are originally from the region (Grisaffi 2010). 20  and conflict (2008). Bolivia is also relatively isolated from its geographical neighbours and the global economy. It is landlocked, and large portions of its frontier are dominated by forbidding geographical obstacles. Moreover, its regional neighbours have experienced significant levels of political violence. Peru, its neighbour to the west, is very similar in numerous respects, having experienced a prolonged insurgency during the period under discussion. Some of the violence in Peru took place in the department of Puno, on the Bolivian border, as Sendero Luminiso operated in the region from late 1981 through the rest of the conflict (Kent 1993: 450-451). The populations of Puno in Peru and the bordering regions of the Bolivian department of La Paz are in many ways closely intertwined, rendering possible the spread of political violence both through direct contact and indirect emulation or diffusion (Iqbal and Starr 2008: 319) across the border during the Peruvian war.  Bolivia is a large producer and exporter of both coca leaf and processed cocaine (UNODC 2005). That production is centred in two areas: the Chapare region in the department of Cochabamba, and the Yungas region of the department of La Paz. There is a tremendous literature now on the relationship between commodity extraction and conflict.12 Indeed, Bolivia is arguably unique now globally as a country that is a major producer and exporter of narcotics and/or narcotic precursors, yet has not experienced significant civil conflict.13 Within the region, Colombia has endured a long-running conflict increasingly fuelled by the production of coca,                                                 12 Exemplars include Collier et al. (2009), Le Billon (2001), Humphreys (2005), and Ross (2004). 13 Conversely, the list of countries with significant levels of narcotic-producing plant cultivation that have also experienced significant levels of violence is long. Coca: Peru and Colombia; opium poppies: Afghanistan, Myanmar, and (to a more limited extent) Mexico; cannabis resin: Morocco and Afghanistan. All have experienced extensive political violence or outright civil war. Laos provides an interesting test case in this regard. Though poppy production there remains low in absolute terms, it has recently experienced rapid levels of opium poppy production (UNODC 2011: 58), It will be important for both scholarly and policy purposes to watch in the coming years for any linkage between increased illicit drug production and heightened risk conflict onset.  21  and attendant trafficking of cocaine (Angrist and Kugler 2008). Similarly, while coca production was not involved in the onset of Sendero Luminoso insurgency in neighbouring Peru, its role as revenue source quickly made it central to the violence in the Upper Huallaga valley, one of that country’s chief areas of coca production, with effects resonating throughout the wider conflict (Weinstein 2007).  Within the Chapare itself, the ethnic divisions were and are complex, yet during the time in question they were thrown into stark relief. Predominantly Aymara and Quechua farmers faced off against a mestizo-dominated government increasingly determined to eliminate all coca production from the region.14 (The government in turn was motivated in large part by external pressure from the United States). Given the centrality of coca production to both the economic livelihoods and self-identity of the region’s farmers15 in the region, a policy of eradication was interpreted as an existential threat to their way of life. Fearon and Laitin (2003) have also argued that a lack of state capacity can be a predictor of conflict insofar as it tends to enhance the opportunity for rebellion. Given that the Bolivian state has limited capabilities in certain geographical areas and has, moreover, developed distinct areas of sectoral competence and incompetence, this would constitute another factor predisposing the country to conflict. This is particularly so in the rural areas of the country such as the Chapare, the Aymara highlands, or the more sparsely populated Amazon basin, where                                                 14 At the same time, remarkable amity persisted and even grew between members of different indigenous ethnicities over time, a fact that itself is worthy of comment and explanation. Accordingly I return to it in Chapter Three during the discussion of the inclusiveness of the cocalero federations. Also noteworthy—though not a topic I explore directly in this dissertation—is the extent to which the cocalero federations have engaged in vigorous competition with other indigenous actors outside federation territory, most notably over the proposed construction of a highway linking northern Bolivia with Brazil. The cocaleros are staunch backers of the project, which is opposed by other indigenous groups, particularly those living in a protected ecological preserve (Territorio Indígena y Parque Nacional Isiboro Secure, TIPNIS) through which the road must pass. 15 Indeed, they are often called cocaleros in reference to the close relationship between plant and the farmers’ shared identity. 22  indigenous populations have traditionally enjoyed considerable autonomy from the central state, though that freedom comes at the expense of limited development and a lack of meaningful direct representation within the government (Yashar 2005, Lucero 2008).  For instance, the United States made provision of aid contingent on development of the armed forces during the 1950s, a factor that Kohl and Farthing argue contributed to the onset of the subsequent period of military rule, essentially constituting the domination of the government by the military branch (2006: 49-50). Throughout the period in question, moreover, the US continued to provide significant aid, including both financial transfers, expertise, and military resource, predicated on Bolivia’s adherence to the US’ securitized counter-narcotics strategy with much of the funding going to Bolivian security forces. At the same time, the introduction of neoliberal reforms in the mid 1980s led to the retrenchment of other parts of the state, with tens of thousands of civil servants dismissed, nationalized industries closed down, and social services curtailed (Kohl and Farthing 2006: 65-73; Gill 2000: 11-18). Indeed, even within the security sector, the Bolivian state has historically had—and continues to have—significant difficulties in enforcing border control, with the movement of untaxed and/or illicit commodities a recurring problem at various times in the country’s history (Kohl and Farthing 2006: 73-74).  In the Chapare region specifically the unevenness of state capacity comes into sharp relief. While capable of maintaining a substantial military presence in the region during the years of attempted coca eradication, other forms of state presence were virtually nonexistent until the last two decades. Accordingly, coca producers of the region were forced to come together and collectively arrange for the delivery of important collective goods including clear roads, health and education services, maintenance of local order and other functions normally considered the purview and responsibility of the modern state (García Linera et al. 2005: 389-390). Indeed, to 23  foreshadow the argument to come, it was this unusual degree of self-reliance that produced the high level of communal institutionalization in the region that I use to account for the emerging pattern of highly contentious yet largely nonviolent conflictual behaviour on the part of the cocaleros.  Taken together, Bolivia is marked by a wide variety of structural conditions associated with civil conflict in the literature, yet has not experienced significant conflict, let alone civil war. This constitutes one way in which to characterize the country as a whole, and the Chaparean confrontation in particular, as an “outlier case” in need of novel explanation. Structural factors and other such macro-level arguments alone do not suffice.  1.5.1 Historical and Contextual Factors Bolivia’s political history is tumultuous. It has the “dubious distinction” of undergoing more than 150 coups d’état since 1825 (Kohl and Farthing 2006: 37). Some of the changes have been particularly dramatic. In “modern” Bolivian history, there has been a national revolution of 1952 led by the Movimiento Nacionalista Revolucionario (MNR), a series of military dictatorships between 1964 and 1982, a slow transition to democracy between 1980 and 1982, and most recently a democratically achieved transformation of the state with the election of Morales as the first indigenous leader, whose Movimiento a Socialismo (MAS) party has since decisively altered the state’s relationship with Bolivia’s majority indigenous, rural population. Both between and during those periods of rapid change, Bolivia experienced significant periods of political resistance (Dunkerley 1984).16 While the activity was often relatively peaceful, consisting of general strikes, protests, road blockades and other such actions, there                                                 16 It is indicative of the level of political contestation in Bolivia that Dunkerley’s (1984) authoritative text on the country’s history is entitled Rebellion in the Veins. 24  were at the same time a variety of small guerrilla movements operating at various times, for instance during the initial military dictatorship of General René Barrientos (Klein 2011: 225). Other guerrilla groups were active during the 1980s, for instance those associated with the Ejército Guerrillero Tupac Katari (Tupac Katari Guerrilla Army) (EGTK; Klein 2011: 242-243). Social protest—whether itself violent or non-violent—has in turn provoked varied and at times brutal forms of government repression (Gill 2000).  Within the context of the Chapare, the long-running confrontation was by Bolivian standards quite violent at times. In 1988 the MNR government passed Ley 1008: Ley del Régimen de la Coca y Sustancias Controladas [Law 1008: the Regulation of Coca and Controlled Substances], declaring the Chapare a “zone of transition” in which coca production would eventually be eliminated. From that moment until an eventual agreement—effectively a peace settlement—signed by Evo Morales and then President Carlos Mesa in 2004 that provided limits on production in exchange for an end to eradication efforts, the state placed varied but continual pressure on the cocaleros to abandon the crop in favour of alternatives (or leave the region entirely). The violence, moreover, was not strictly one-sided, with more than two dozen security officers killed between 1998 and 2004 (Ledebur 2002: 1; Salazar Ortuño et al. 2008: 9-10). The stakes were significant, with control of a highly lucrative and marketable resource under contestation. There were sharp identity markers dividing cocaleros and a state that at the time tended to view its interests as being closely aligned with the country’s relatively small, urban, and mestizo middle and upper class populations (Lucero 2008: 42). The cocaleros, moreover, were well organized and emphasized a shared identity of indigenous producer over other possible identities, including more specific ethnic ones that might have served to divide the community. 25  In doing so they exhibited the capacity to coordinate a variety of large-scale political activities. Yet, despite the stakes, the organization and determination on both sides, and the recurring flare-ups that could have escalated, the conflict continued for nearly two decades without breaking out into open armed violence. Thus, the case defies easy explanation using conventional theoretical explanations. For instance, Kalyvas’ (2006) control-based narrative cannot fully account for the observed peace, given that control was often contested and unequally divided between the state, which maintained its hold on major towns in the region; and the peasant, indigenous, and cocaleros organizations that maintained control in the rural areas, where many of the confrontations took place (García Linera et al. 2005, subject interviews). This situation of contested control is precisely the condition that Kalyvas associates with heightened levels of violence in conflict; while some violent interactions did take place during confrontations in the countryside, they never rose to the levels of sustained armed conflict. Moreover, individuals and groups at times encountered a variety of strong incentives to engage in violence, often for deeply personal reasons such as revenge, perceived injustices, and indiscriminate repression of the growers’ livelihoods (Kohl and Farthing 1997). More prosaically, it cannot account for cocaleros’ restraint in using violence to protect their share of coca revenues, or likewise their continued reliance on constrained forms of political resistance to achieve their desired end of legitimate production (Harnecker and Fuentes 2008; Salazar Ortuño et al. 2008).  1.5.2 Towards an Explanation of Contention in the Chapare So, why were the cocaleros so well organized, unified, and disciplined in their response? Drawing on the above accounts of the general theory and primary case, it is now possible to sketch out in brief the argument. Settlement of the Chapare region actually began in earnest only 26  in the 1960s, and occurred with very little in the way of official government support. Accordingly, the new colonizers looked to themselves to provide core services, employing a “syndical” or union-based structure to organize themselves within the new settlements. In turn the settlements organized themselves into the large nested structure of the six federations. The cocaleros, in short, constituted themselves into a federated OC in response to pressing local collective action problems. Accordingly, within that organizational structure, the norms of burden sharing and active engagement in the realization of communal interests remained prominent, operating at both local and community-wide levels. When faced with government policies of crop substitution and coca eradication, including the use of violence in pursuit of those goals, cocaleros possessed the means to organize themselves in response and articulate a common strategy built around the advancement of the “political instrument”—that is, their own political party, eventually known as MAS—using both conventional and non-conventional political strategies. The former involved advancing candidates for office in local and national elections; the latter encompassed a broad repertoire including marches, demonstrations, blockades, and hunger strikes. The use of violence by cocaleros remained sharply limited and coordinated through the cocalero unions and their officials; indeed, union officials rigidly enforced those limits. As a result, there was virtually no social space for marginalized community members to develop alternative strategies in response to government policy, including armed resistance. Once undertaken, the community of cocaleros remained committed to the dual strategy of non-violent, non-conventional resistance combined with conventional political competition, collectively eschewing more violent alternatives despite at times significant provocations to do just that in order to respond to what they viewed as an existential threat to their way of life (subject interviews; Congreso Ordinario 27  de la Federación Especial de Trabajadores del Trópico Cochabamba 2000). That campaign ultimately concluded in victory, as the state in 2004 formally conceded cocaleros’ right to produce a limited amount of the coca plant. 1.6 Conclusion In this chapter I advance three interrelated sets of claims linking specific patterns of social organization to a reduced likelihood of violent conflict. First, I advance the concept of the Ostromian communities, which refers to communities featuring locally embedded and inclusive institutional solutions to significant social dilemmas. I argue that the members of Ostromian communities tend to 1) engage in high levels of conventional and nonconventional political contestation across a range of social arenas with both the state and other communities in service of shared communal interests. At the same time, they 2) avoid engagement in violent conflict in defending those interests, excepting 3) the presence of a conjunction of perceived existential threat combined with a lack of feasible alternative strategies of resistance. Second, I argue that once members of the community took the decision to avoid conflict on the basis of rational analysis of shared interests, it put in motion processes of socialization that tend to shape identities in ways that reconstitute those same actors’ interests. While taken contingently, the choice to remain nonviolent under certain conditions can become a component of communities’ social identity and institutional structure. More violent alternatives become progressively less likely as such identities harden. Third, I argue that it is precisely in the context of unconsolidated states—where democratic institutions function imperfectly, and the state is not sufficiently powerful to unilaterally suppress all conflict—that they will have the greatest independent effect on pattern of contestation. Among other things, members of such communities are likely to see value in pursuing democratic reform over other, higher risk 28  strategies so long as an underlying commitment to democratic norms remains in place within the state and civil society at the national level; that is, they will avoid conflict so long as other less costly options remain available. Taken together, I argue that under a wide range of conditions, OCs are less likely to engage in armed violence than communities exhibiting other patterns of social organization. In the context of the Chapare, I argue that the cocalero unions constituted an example of such an OC. Accordingly, once faced with the significant extracommunal threat represented by state efforts at coca eradication, the cocaleros were able to collectively articulate, decide upon, coordinate, supply, and enforce community-wide compliance with an ultimately successful strategy of conventional political activity and non-conventional yet nonviolent resistance. In subsequent chapters I develop further the empirical basis on which to justify these claims. 1.6.1 Dissertation Roadmap The rest of the dissertation unfolds as follows. In Chapter Two, I develop the theoretical explanation connecting local intracommunal organization to extracommunal violence. In doing so I define and explain key concepts such as contention and violent conflict, along with the constitutive elements of an Ostromian community. The chapter also develops the leading theoretical alternatives to my hypothesized explanation, and generates testable predictions flowing from each to enable the kind of theoretically rich comparative analysis of the case that characterizes the process tracing approach. Chapter Three first provides an operationalization of the constituent elements of an OC, before presenting and analyzing the specific case of the cocalero federations. It provides a variety of empirical evidence establishing how the federations emerged, and how they constitute OCs; it also offers a set of hypothesized explanations of why they emerged in the form that they 29  did, though testing of these different arguments must await additional comparative case analyses. Chapter Four in many ways constitutes the heart of the dissertation. It provides an overview of the history of contention between the state and cocaleros before entering into a series of subcase studies describing and accounting for six distinct processes that occurred over the course of contention in the Chapare. These range from the unification of the federations themselves and the ways in which they dealt with subsequent challenges to that unity, to the process of socialization that constituted an important aspect of internal cocalero culture, to the processes of local and federation-wide mobilization in support of distinct strategies of resistance, to a struggle with the state over just how to frame the ongoing confrontations in the region. Taken together, these cases use the evidence assembled to illustrate how the OC structure of the federations shaped contention with the state in important ways, leading to energetic and often unconventional resistance of the state, but in the context of carefully delineated limitations on the use of violence. Finally, in Chapter Five I turn to questions of generalization, analyzing the scope conditions bounding the applicability of meso-level institutional explanations. In doing so, it also begins the work of comparative analysis both to specify relevant scope conditions more clearly and to begin the work of verifying the external validity of the account rendered here. In Chapter Five I also summarize the dissertation, provide a review the its central theoretical contributions, and identify several steps in a longer-term research project necessary to corroborate and extend those findings. The chapter also provides a summation of results and contributions. 30  Chapter 2: Ostromian Communities and Conflict: Concepts, Theory, and Rival Explanations  Having provided an overview of the dissertation project in the previous chapter, I now focus on theoretical development of the “Ostromian community” along with other significant concepts employed in this dissertation. In the first section, I review central concepts related to the factor to be explained, including both violent conflict, and the more general phenomenon of contentious politics. Second, I situate the analysis to come as a “meso-level” theory—effectively, a study of institutions linking together micro-level incentives and actions with macro level actors and phenomena such as the state, an insurgent group, as well as what Kalyvas (2006) refers to as the “master cleavage” in situations of contention and conflict. In section three I develop the concept of the Ostromian community in detail as an example of such a meso-level explanation. I also specify its hypothesized relationship to patterns of contentious politics. In this section I also spend some time examining how identities change in the context of political contention; specifically, I consider the possibility that identities may under certain circumstances harden in conflict in ways that make violent conflict less likely, in contrast with the more conventional understanding of identity hardening as constituting a precursor to violent conflict. The chapter then concludes with a discussion of the scope conditions that limit the effectiveness of meso level explanations such as OCs, notably including state regime type and degree of consolidation, and an outline of alternative explanations for the primary case of Bolivia and other comparable cases.  31  2.1 Theoretical Specification 2.1.1 Concepts: Conventional Politics, Contentious Politics, and Conflict  All societies are to some extent conflictual. Societies inevitably exhibit a diversity of interests and values; moreover, preferences can and do change over time, such that a harmony of interests, even if achieved at one point in time, is unlikely to last. Likewise, all states experience organized violence. Weber’s idealized definition of the state places the means of violence under government monopoly (2004 [1919]). Contemporary societies, however, also include numerous non-state-based purveyors of organized violence, from criminal gangs, to local militias, to private security companies. The question, “why has country X not experienced civil war?” therefore implies and requires a number of conceptual specifications. Drawing on Tilly (2003), Chenoweth and Cunningham (2013) and others I make a three-fold distinction between 1) conventional politics, marked by the use of formal institutions to resolve conflicts; 2) nonconventional but still nonviolent resistance, marked by the use of diverse repertoires such as marches, protests, blockades, hunger strikes and other forms of civil disobedience, as well as innovative and controversial strategies within existing institutions to press political claims; and 3) violent conflict, marked by the organized use of violence in pursuit of political ends. In the third category I place both civil conflict and civil war as defined by Gleditsch et al. (2002) among others. Following Tilly, (2003: 30) I refer to the latter two categories collectively as “contentious politics” and “contentious behaviour,” to distinguish both from the “politics as usual” that characterizes interaction falling into the first category. It likewise includes Kalyvas’ definition of civil war (which he argues is better specified as “internal war”): “armed combat within the boundaries of a recognized sovereign entity between parties subject to a common authority at the outset of the hostilities” (2006: 17).  32  Contentious politics, whether nonconventional nonviolent or armed conflict, is marked by a “contested incompatibility” (Gledtisch et al. 2002), in which one or both sides of an issue make claims that cannot be resolved (or alternately, that one or both sides refuse to resolve) via existing political institutions alone. It incorporates forms of political confrontation that at least partly fall outside the conventional institutions of politics in the given country. I also refer to this as unconventional politics. In this, my approach is consistent with studies of nonviolent resistance as characterized by Chenoweth and Cunningham (2013: 271): We define nonviolent resistance as the application of unarmed civilian power using nonviolent methods such as protests, strikes, boycotts, and demonstrations, without using or threatening physical harm against the opponent. Civilians challenging the state through nonviolent struggle employ irregular political tactics, working outside the defined and accepted channels for political participation defined by the state. Both civil conflict and unconventional political conflict are thus contrasted against a more stable form of consolidated peace, of which there are numerous variants. Fein (1995), Hegre 2001, and others for instance isolate two distinct, stable equilibria in the form of autocratic and democratic peace. This conforms well with established findings with regard to regime outcomes found by Przeworski et al. (2000) for example, who identified robust stability among relatively affluent democracies and autocracies, but much greater variation in outcomes among less prosperous states. These stable equilibria themselves can and do break down, but the paths from consolidated autocracy or democracy to political instability are distinct from the various paths from instability to the various instantiations of violent conflict.  In making this move to define three distinct forms or modes of politics, I am attempting to push back against what remains an all-too-common set of assumptions within contemporary conflict studies. In essence, I resist the assumption built into much of the relevant literature that 33  our outcome of interest takes only two values: peace and conflict.17 While research undertaken with this assumption has been fruitful, at the same time it has suffered from what might be called a "positivity bias,” resulting in important silences in the literature. For instance, the landmark work by Kalyvas (2003, 2006) heralded a shift in focus from explanation of conflict onset and resolution to providing an account of the dynamics of conflict. Political dynamics in the context of violent conflict are fundamentally different from those that occur in peacetime, a point articulated most explicitly but by no means only by Kalyvas (2006: 76-83; see also Auteserre 2010, Cohen 2013). Thanks to the work of Kalyvas and other conflict scholars, along with the renewed focus on counter-insurgency in response to developments in the Middle East and Central Asia (e.g. Lyall et al. 2013, Berman et al. 2011, and Fair et al. 2013), the study of the dynamics of civil war has become dominant in recent years. Researchers have striven to uncover, with some notable successes, the particular dynamics that emerge once war has begun. Likewise, policy-oriented research likewise has focused increasingly on the types of intervention likely to be effective in the midst of sustained armed violence. That disciplinary choice has led to increasing neglect of the previously central question—also crucial in terms of policy applicability—of what causes conflict onset, and by extension how to either mitigate or avoid it. To be sure, that agenda represented an exciting and productive vein of exploration that opened up just as the returns to the first generation large-n dataset approach, with its reliance on a limited number of panel datasets, had begun to diminish. We have important findings from the largely quantitative “first generation” of research on the subject, with important results on subjects ranging from the structural conditions most commonly associated with civil conflict                                                 17 To the extent that the issue has been addressed, it is to identify where the line between peace and conflict ought to be drawn; the more fundamental assertion that there are but two possible values has been more rarely contested, though there are some exceptions such as Tilly (2003), and Chenoweth and Cunningham (2013). 34  such as limited state capacity (Fearon and Laitin 2003), to the material feasibility of resistance (Collier et al. 2009), the role of specific types of fungible resources in sustaining conflict (Ross 2004, Humphreys 2005), the relationship between democracy and conflict (Hegre 2014), the effects of horizontal inequality (Stewart 2002, Østby 2008), the effects of refugees on conflict dispersal (Salehyan and Gleditsch 2006), the variation and causes of sexual violence in conflict (Cohen 2013), and fractionalization at both elite and group levels (Goldstone et al. 2010, Montalvo and Reynal Querol 2005, Esteban, Mayoral and Ray 2012).  Notwithstanding such progress, or perhaps rather because of it, the work in this vein increasingly has turned to the sort of “gap filling” associated with what Kuhn (1962) describes as “mature science”: ensuring the proper specification of models, the competitive evaluation of available datasets, and so on. These are all important tasks, crucial to the cumulation of knowledge on the subject of conflict, but limited in their ability to generate new insights, or new approaches to existing research problems. Indeed, much of the most interesting work has taken place in allied areas of study. Straus (2012) stands as an intriguing exception in a related field, focusing on the avoidance of genocide, examining micro, macro, and meso level factors associated with robust societies able to “retreat from the brink” of genocidal violence. Chenoweth and Cunningham’s edited special issue on the logic (2012), determinants, and effects of nonviolent resistance is likewise a valuable contribution, bridging as it does the large field of conflict studies with the important but (until recently) less systematic study of unconventional yet nonviolent political resistance and contention. 35  As alluded to above, in elucidating his broad concept of “contentious politics” Tilly (2003: 30)18 provides one basis on which to grapple with the idea of a spectrum of ways of experiencing social conflict, some of which are violent and some of which are not. In doing so he provides tractable, if causally indeterminate, conceptual tools with which to articulate and explore the widely held intuition that a tremendous variety of different kinds of conflictual outcomes—from conventional electoral politics, to protest, civil disobedience, non-violent resistance, through to unorganized scattered attacks, covert violence and even civil war—may result from the interaction of contingent intervening factors with a relatively small set of fundamentally similar structural preconditions of inequality, arbitrary governance, in conjunction with instances of perceived grievance, individual self-interest, and political opportunity. While Kalyvas is undoubtedly correct that the dynamics of violent conflict, once joined, are distinct from socio-political dynamics in non-violent context, it is just as correct that those situations were in some ways commensurate, and therefore comparable, prior to initial onset. What we require, therefore, is a theory specifically articulating the internal variations that distinguish the stably unstable cases from those in which order can and does break down at some point. It is a central goal of this project to provide some theoretical and empirical work in pursuit of that larger research agenda.  In the following sections I suggest one strategy to move towards that goal: focusing on the description and explanation of various types of meso-level interaction. I first give some sense of what a meso-level theory looks like, and then develop one particular example that I refer to as                                                 18 Tilly and his collaborators in McAdam et al. (2001) outline an alternative approach. They define contentious politics as “episodic, public, collective interaction among makers of claims and their objects when (a) at least one government is a claimant, an object of claims, or a party to the claims and (b) the claims would, if realized, affect the interests of at least one of the claimants (1998: 5). 36  an “Ostromian community”, the presence of which I argue influences patterns of extracommunal contention and conflict in distinct and predictable ways. In developing this argument, I identify a number of constituent elements of an OC, and describe in general terms how its presence in an episode of contention would shape members’ decisions, actions, and even—over time—their identities.  2.2 The Meso Level To start with, Kalyvas (2006) argues that to gain a better understanding of civil war violence one needs to integrate the three levels of analysis. He characterizes the micro level as the realm of “intracommunal dynamics and individual behaviour;” the macro level as that of “elites, ideologies, and grand politics;” and the meso as “the institutional context within which interactions between political actors and civilians take place” (2006: 10-11, 106). He finds that civil war violence has a joint character defined by these interactions, and his theory of irregular war accordingly aims to “bridge… the meso- and micro-levels [to] predict… the likelihood of violence as a function of control” (2006: 12). At the same time, towards the end of his life Tilly offered a conception of meso-level study in his framework of “relational” theorizing focusing on bundled mechanisms and processes. These, he argued, ought to be the central arena of investigation for the social sciences (Tilly 2001, 2004). The same guiding principle holds in this dissertation—the identification of recurrent patterns of social interaction that, while varying in form from place to place (and indeed from time to time), nonetheless produce identical or similar relations among social entities. They thus represent a stable form of generalizable theorizing that nonetheless remains more sensitive to the vast variety of potential human interaction, and the need for sensitivities to context and historical contingency than is possible under the relatively inflexible “covering law” 37  approaches to social and political theorizing that dominated “first generation” studies of conflict. There are various ways this might be conceptualized, from Kalyvas’ transactional model to Tilly’s relational frameworks.19 To state my own conception concretely, I understand the meso level as the arena that links together actors and processes occurring at the micro and macro levels. Specifically, I focus on an institutional model of that space here. 2.3 The Ostromian Community I turn now to the articulation of the characteristics and distinctive properties of what I refer to as an Ostromian community, a meso-level theory focusing on the effects that particular patterns of political organization have on political contention, contingent on particular macro level conditions of the sort identified in Chapter One and discussed in more detail in Chapter Five, notably the presence of some aspects of democratic governance—though not necessarily fully consolidated democracy—and some limitations on state willingness or capacity to unilaterally target and destroy local organizations by force without violent provocation. Building on Ostrom’s account of common pool resource (CPR) institutions (1990), I define an Ostromian community to be one in which members have realized inclusive, locally embedded and institutionally-based solutions to shared social dilemmas, regardless of the nature of the problem. To an extent, the nature of the social dilemma itself is unimportant. Possibilities include not only challenges of shared resource extraction as in Ostrom’s work, but also the provision of essential services such as health care, community education, infrastructure, and locally-administered rule of law when such services are not forthcoming from higher levels of government. Indeed, given that conflict is most likely to occur and recur in the contexts of states                                                 19 An example of a relational frame employed by Tilly would be his discussion of impersonal institutions as a crucial development of democratic governance (2005). 38  of limited capacity where armed conflict is a real possibility, one would expect such self-organized communities to exist frequently in the context of areas at risk of organized violence. To presage the following discussion, Ostromian communities tend to be a) inclusive and bounded, and b) capable of taking rational decisions that are c) subsequently enforceable among both elites and membership. Equipped with such organizational resources, such communities possess significant emergent capacities to effectively manage extracommunal conflict. In most situations, non-violent contention proves to be the most efficient and effective method of pursuing shared interests. They are able to achieve such stable solutions thanks to the presence of a number of what I refer to as  “constituent elements.” In formulating these I draw upon Ostrom and other collective choice scholars (Wade 1994, Agrawal 2001) regarding the necessary conditions for communities to implement institutionally regulated shared solutions to collective action problems. I also amend their descriptions in order to accomplish two major goals. First, I re-express various concepts in more general terms, given my interest in shared solutions to significant collective action problems of any type, and not just issues of resource management. Second, given that my interests in OCs relate to the incidence and avoidance of conflict, I specify the concept of OC to include some aspect of open and inclusive identity, most likely geographically based (i.e. residency). Effectively, communities organized around some kind of identity-based exclusionary principle—whether grounded in ethnicity, religion, class, family, or some other marker—may produce solutions that engender conflict with others in the same geographical area who are not included in the resulting organization. I divide these constituent elements into three categories: foundational, governance, and relational. The foundational elements of an OC are  39  1) A clearly defined communal boundary, 2) A shared and inclusive identity, and 3) A shared internal challenge taking the form of a collective action problem.  In response to the dilemma encountered, the community has gradually organically developed, incrementally adapted, or explicitly designed and implemented a set of institutions to meet the shared challenge in an inclusive manner, using the same principles that Ostrom identifies as operating in common resource pool institutions, and operating what she describes as a “collective choice” arrangement (1990: 90-103). Such institutions feature mechanisms by which to incorporate perspectives of those affected within the community on what solutions to implement, how to implement them, and how to ensure all members share benefits and burdens equitably. I refer to these collectively as governance elements, and they include 4) Deliberating and deciding on a common policy in response to that problem by means of some mechanism of collective choice; 5) Monitoring and sanctioning defection by members; 6) Monitoring and sanctioning defection by leaders and other agents; and 7) Dispute resolution mechanisms serving to strengthen within-group cohesion.  In her account of CPRs, Ostrom identifies two further properties related to external relations and scale, respectively. These too apply to OCs more generally; I refer to them as “relational” elements. 8) First, the entities enjoy minimal recognition from external authorities (though it may be sufficient for the state to exhibit some form of benign neglect). This property, it is worth noting, places Ostrom’s own theorizing within the meso-level as I have defined it.  9) Second, in larger contexts—such as the primary case under consideration here—the entities take on a federated or otherwise nested organizational structure such that the above properties are organized at local levels and over multiple levels to maintain communal accessibility with larger scale organization.  The presence of such a socio-institutional structure has numerous implications for OCs’ interactions with extracommunal actors, providing such communities with a variety of emergent 40  capacities that are unlikely to be found in communities lacking such institutions that tend to shape their responses to extracommunal threats. I draw attention to two.  First, OCs exhibit what I describe as “social rationality,” analogous to what Fearon (1995) and Waltz (1979) would refer to as unitary rational decision-making in the international context. They take strategic decisions based upon the shared interests of the community.20 Thanks to their inclusive deliberative institutions, OCs have the potential to confront threats or challenges to the community’s security in a manner representative of an inclusive understanding of the needs of the broader community. As a result, strategies and objectives, adopted through existing decision-making institutions adapted to confront new circumstances, will tend to reflect potential costs and benefits to the community as a whole rather than some select group within it. Depending on the manner in which leaders are selected, such communities may be quite stable internally as well, with leaders incentivized to consider long-term implications of decisions taken, rather than simply opt for a more expedient route. In economic terms, the social discount rate for such communities is thus substantially lower than it would be without this institutional capacity. Leaders likewise are incentivised by the self-reinforcing nature of such institutions to act in the interests of all constituents in the community, rather than in the service of some subset of the population (Weingast 1997). Second, OCs are effective mobilizing organizations. Once decisions are taken, such communities have in place a system capable of mobilizing available human, financial, material, and informational resources. In general, they do so out of a form of contingent compliance (Levi 1989), in which all members accept the need to participate, but only so long as they remain                                                 20 Ostrom (1990) herself conceives of this more in terms of contract theory in which all parties to an agreement retain the ability to modify agreements to which they are a part. This ensures a continuing good “fit” between participants and the situations they find themselves in, even long after the initial agreement was taken. 41  confident that some coercive capacity exists—whether centrally organized or dependent on peer-based monitoring—to ensure others will comply too, and that any defectors will be suitably sanctioned. Again, such capacities, put in place to address previous dilemmas of collective action (Ostrom 1990), may be adapted to new challenges as they arise, including extra-communal threats to lives and livelihoods, in a way that minimizes free-riding and maximizes the mobilization of existing resources.21	Simply put, the presence of a locally embedded set of institutions dedicated to the collective resolution of social dilemmas generates considerable externalities that shape communities’ responses to extra-communal threats, whether from other neighbouring communities or from the state itself, in ways that affect both the form and intensity of resistance or reaction. As I describe and illustrate throughout this dissertation, over a significant range of conditions, the paths of response chosen by OCs will be ones that mitigate the risk of violence, even as they produce heightened levels of political contestation.  While the institutions are located at the meso level, they have effects on mechanisms and processes found at the micro and macro-level as well. At the micro-level, the OCs forestall individual and small groups from engaging in conflict in the service of private motives in a number of ways. First, group leaders and marginalized members alike are constrained by the presence of an enforcement capacity, ensuring that community members attempting to pursue an independent agenda do not subvert decisions taken. The pursuit of private interests tending to                                                 21 Encouragingly, this accords with Tarrow’s (1994) understanding, based in part on the previous ground-breaking work of McAdam (1982) of the need for such mobilizing organizations at the heart of any effective movement. Responses to perceived injustices do not simply emerge; some mechanism is inevitably necessary to overcome collective action problems. Inclusive OCs solve those problems in a way that tends to mitigate the possibility for violence in comparison with other, more exclusive solutions in which the costs of any conflict may be more easily externalized. 42  work against communal goals, whether in the form of personal gain, vengeance, or defection as a personal survival strategy, is rendered more costly by the provision of credible sanctions for any members that deviate from accepted strategies and associated practices. Accordingly, the inclusive nature of such communities reduces the potential for alienated subgroups or for within-group factions to emerge and pursue distinct political agendas. At the same time, institutions of the community tend to encourage feelings associated with Wood’s (2003) “pleasure in agency” among its members. Even if they remain excluded from significant aspects of the political activity at the national level, they retain significant feelings of participatory satisfaction through their involvement in robust local institutions. Thus, in line with Walter (2004), they remained less likely to endorse or seek out violent forms of political action given the possibility of effective belonging and participation within. Likewise, they remain less vulnerable to the sort of anomie and political isolation that Humphreys and Weinstein (2008) suggest provokes many individuals’ decisions to enlist in violent groups. Members of OCs have no need to seek out some form of political belonging, violent or otherwise; they already belong to and participate in a political project (whether they want to or not). At the meso level, decisions and actions taken by community representatives and agents with regard to intercommunal contention will tend to be in the service of shared communal interests, rather than those of a narrow elite or some specific dominant subgroup, whether ethnic, religious, familial, class-based, or otherwise thanks to checks on leaders’ ability to use their public offices for private ends. At the same time, the presence of significant deliberative processes and mobilization capacity encourages leaders to consider the full set of costs to the community of any particular course of action. In effect, an OC tends to reduce the scope of 43  violence to Fearon’s (1995) rationalist arguments for conflict. Put differently, the presence of an OC significantly reduces or eliminates both irrational arguments for war, and war due to selective incentives. When OCs fight, they do so because leaders and those who support them deem it to be most rational course available, the most likely way to achieve shared interests and respond to significant shared threats to lives and livelihoods. In the case of the Chapare for instance, cocalero leaders concluded that even a flawed democratic process held out greater hope of success than armed resistance. Cocalero members agreed, and all held to that line through nearly 20 years of confrontations.  Likewise, situations of “collectively inefficient” violence—that is, situations in which benefits of conflict accrue for some members of the community, but are outweighed by the net costs imposed on the community as a whole—are possible, but in general only following some prior process of institutional breakdown or more general community fragmentation (e.g. Acheson 2006; Pearlman and Cunningham 2012).22 The phenomenon of elite fomented violence described by Fearon and Laitin (2000: 864-868), such as that encountered prior to the Rwandan genocide (Prunier 1995, Uvin 1999: 259-260) or in the build-up of the Sri Lankan conflict (Kapferer 2011), are less likely to occur without an accompanying process of internal institutional paralysis, polarization, or breakdown. More general cases of conflict emerging through processes of elite fragmentation and radicalization (Fearon and Laitin 2000, Liow 2006: 10, 41) are likewise less likely to emerge. Given the strength of internal monitoring and enforcement mechanisms, cases of insurgency emerging from the margins, as was the case in Peru in the Sendero Luminoso (McCormick 1987) and the emergence of the Lords Resistance                                                 22 Though focused at a different level of analysis, see also Goldstone et al. (2010) for macro-level findings regarding the link between national polarization and resulting political instability. 44  Army out of the chaos of earlier conflict in Uganda (van Acker 2004), are similarly less likely to gain a foothold.  Finally, at the macro level, I argue that OCs fill in several important gaps in existing models with regard to three important macro-level phenomena related to conflict onset, namely (1) the asserted pacifying effects of democracy (and the mechanisms by which they are achieved), (2) the possibility of state repression producing an escalatory conflict spiral, and (3) the converse possibility of conflict emerging out of a lack of state capacity to forestall insurgency. Importantly, Ostromian communities alone cannot outweigh the effects of state-level characteristics and state-decisions. States can unilaterally decide for war; if they do so, there is nothing that OCs can do to forestall it; they can only choose how to respond.  Thus, for instance, without the possibility of national-level peaceful conflict resolution offered by democratic institutions, OCs may be much more likely to conclude that conflict is the best or only way to pursue desired ends. They can, however, fill in the participatory gap left by weak or otherwise defective participatory process. They can likewise enable significant restraint in communities’ response to low level state repression so long as they see it in their interests to do so, forestalling the emergence of the sort of escalatory spiral of repression and response seen elsewhere (Davenport 2007). When states lack the capacity to maintain the rule of law and other prerequisites of peace and security in a region, a local OC may possess both an interest in and the capacity to take up the role within its own defined boundaries, preventing the emergence of local insurgencies and other less grandiose forms of violent activity. Ultimately, on the basis of the above I predict a particular set of outcomes for Ostromian communities that have encountered a contested incompatibility with extracommunal actors, most notably—though in theory not only—the state. I argue that OCs (1) support rational group-level 45  decision-making in the pursuit of shared community interests. Given their ability to effectively mobilize resources in the service of those interests they (2) engage in high levels of conventional and nonconventional political contestation across a range of social arenas with both the state and other communities. Likewise they (3) avoid engagement in violent conflict in defending those interests provided there remain other, more utile options. Only in rare situations, those characterized by both a lack of preferred alternative strategies of resistance and a sufficiently calamitous threat, do OCs resort to violence.  Possessed of an active leadership charged with the pursuit of widely accepted communal interests and equipped with an ability to mobilize communal resources in the service of those interests, OCs are well-positioned to compete aggressively and effectively with extracommunal actors, regardless of the nature of the opponents or the field of contestation—formal or informal; political, economic, legal or social. It is for this reason I expect them to engage in contentious politics across a wide variety of fields. At the same time the OCs’ deliberative capacity, combined with the stability they provide to leaders to engage in long term planning (compared with alternative community structures), ensures that the full costs of any potential armed conflict are considered when contemplating such a step.23  The inclusive solutions to social dilemmas that characterize OCs as I employ the term may be distinguished from a variety of alternative meso level institutional structures. 24 I offer one potential typology in Chapter Five as an aid to further research. In contrast to such                                                 23 On fields of contestation, see Fligstein and McAdam (2011).  24 Ostrom, for example, characterizes the two most common alternatives as “the Leviathan,” and “privatization,” respectively, referring to the state and private enterprise as providers of desired public goods. She also allows for the possibility of weak or flawed institutional goods providers, something to consider when extending the theory to other cases, a process I begin in Chapter Five. 46  alternative situations, 25 Ostromian communities are capable of taking and implementing collective decisions, and are likewise capable of enforcing agreed-upon rights and obligations on the population, in general along the lines of contingent or quasi-voluntary compliance in situations described by Levi (1989), which Ostrom puts near the heart of the solution to the collective action problem achieved by such groups. In such contexts, elite rebellion and marginal insurgencies alike, the most common sources of non-state violence, are less likely to emerge against the expressed and enforced popular will of the larger community. Accordingly, a functional Ostromian community will tend to avoid “socially inefficient” violent conflict—i.e. conflict that runs against the aggregate interests of the community, benefiting some while producing net Pareto inefficient26 outcomes for the community as a whole—that accounts for the majority of conflicts observed globally today. 27 At the same time, they engage in enthusiastic defence and pursuit of shared interests. When such communities do engage in extra-communal conflict, whether nonviolent or violent in form, it will be for reasons and on terms that tend to enjoy the support of a stable majority of the community represented within local institutions. They will likewise make for formidable opponents given their capacity to mobilize resources and undertake (and enforce) communal decision-making.                                                 25  This in turn is based on prior findings regarding the potential for more efficient solutions that become possible in iterative versions of strategic interaction games (Axelrod and Keohane 1985), combined with the possibilities of coordination presented by institutions (Keohane 1984/2005, Weingast 1997, North 1990, Shepsle 1986). 26 Briefly, a Pareto efficient distribution of resources is one in which no person can be made better off without making someone else worse off. Conversely, a Pareto inefficient distribution is one in which some surplus value is lost due to, for instance, a leader siphoning off resources to private ends that could be put to more productive use serving public interests, or an armed group waging a war that hurts civilians significantly. 27 In this, they operate in a manner complementary with the model of in-group policing proposed by Fearon and Laitin (1996) in their account of interethnic cooperation, and indeed conform to expectations arising from Weingast’s (1997) model of the self-reinforcing characteristics of stable democracy. While the latter’s theory operates at the national level, there is no intrinsic reason analogous effects might not be observed in effectively bounded and institutionally governed subnational communities. 47  Even when a threat is perceived as significant—or indeed existential—there are often methods of resistance, such as what Chenoweth and Cunningham (2013) describe as methods of “nonviolent resistance” that, over a range of circumstances, are likely to be more efficient given the high costs and low likelihood of success associated with armed conflict. So long as such options remain viable and less costly than recourse to violent conflict, the likelihood of escalation by non-state actors remains low, as the non-violent alternatives continue to be preferred and pursued. Indeed, the above insights point to two prima facie proofs of concept. First, many of the cases that Ostrom examines in her own study of shared resource pools are remarkable in their longevity. Ostrom (1990) includes numerous examples of communities that have maintained stable institutional solutions to common resource pool dilemmas for hundreds of years, solutions that indeed predate the emergence of the modern state in which the communities in question are located. Törbel, Switzerland has practiced a form of collective resource governance since at least the 16th century. Some of the Spanish irrigation systems she examined traced their origins back to the 1400s. The Japanese villages she studied managed forests communally from 1600-1867 Though by no means conclusive, the fact that such communities are so stable, often outlasting the political regimes in which they are embedded at least suggests that such institutional arrangements are quite robust to political upheaval, an observation consistent with the idea that such communities are successful at managing extracommunal conflict. How else would Ostrom’s examples survive successive changes in state-level governing regimes? Second, many of the determinants of nonviolent resistance identified by Cunningham (2013)—in contrast with either conventional politics on the one hand, or civil war on the other—48  are neatly explained by the relative presence or absence of a particular constellation of collective action problems and solutions. Her largest conclusions are that civil war is more likely, as compared to conventional politics, when self-determination groups are larger, have kin in adjoining states, are excluded from political power, face economic discrimination, are internally fragmented, demand independence, and operate in states at lower levels of economic development. I find that nonviolent campaign is more likely, as compared to conventional politics, when groups are smaller, are less geographically concentrated, are excluded from political power, face economic discrimination, make independence demands, and operate in non-democracies. (Cunningham 2013: 291)  Though focused specifically on self-determination groups and conflicts, a number of her findings nonetheless complement my own case-specific results. Of particular interest is her finding that internally fragmented groups are more likely to engage in violent resistance. This makes sense in itself in light of Fearon and Laitin’s (2000) conclusions regarding the role of elite bidding in the emergence of violent identity-based conflict. Such bidding often hinges on the presence of latent divisions in society for such leaders to exploit. It also makes sense in the context of OCs, however: to the extent that a local institutional structure tends to limit the possibility of fragmentation among community members, one would expect less violence to ensue. Likewise, smaller more geographically concentrated groups are, other things being equal, more likely to form inclusive and bounded institutions. 2.3.1 What OCs are…and Are Not Before turning to scope conditions on the effectiveness of local institutional structures in the next section, I conclude the discussion of OCs with several caveats in an attempt to further specify what it is that I am (am and not) arguing. First, it bears underlining that there is no invariable expectation that Ostromian communities will universally be more peace-loving than other forms of community. Just as Fearon (1995) derives situations in which conflict is rational, 49  so too under certain circumstances Ostromian communities may find themselves with no better option than to engage in armed conflict. There is nothing doctrinaire about any emergent pacifism. On the contrary, when faced with a threat28 deemed too serious to ignore and impossible to respond to effectively through less costly means, a natural implication of the theory I present here is that in the absence of peaceful alternatives, such communities will engage in broadly-based and therefore relatively highly effective mobilization in support of armed conflict. 29 That war is comparatively unlikely because the high costs involved does not mean it never happens.30 Accordingly, the theory here is not a simple normative account whereby norms of internal conflict resolution construct extracommunal interactions. This is not a theory of communal pacifism, and requires no heroic commitment to peace on the part of community members or leaders in order to produce predicted effects. Nor is it the case that such communal coordination comes without cost. On the contrary, the degree of regulation within OCs, particularly those under heavy pressure such as that encountered by the cocaleros in the primary                                                 28 In theory, such groups are also capable of acting in pursuit of opportunities, rather than just responding to threats. Indeed, it was in the midst of the First World War, when British resources were stretched to their thinnest, that the Irish parliament—a body constituted by and consisting of Irish MPs elected to Westminster and thus exhibiting an interesting form of dual legitimacy—voted for war; the ensuing initial war for independence was a relatively quick and decisive victory for the Irish. In the Bolivian context, it is interesting to observe how, in the temporary era, cocaleros have aggressively pursued their interests, even against other members of the broad national governing MAS coalition headed by cocalero leader Evo Morales, as in the case of the still ongoing “TIPNIS” affair, in which cocaleros have mobilized in support of a proposed highway to be built through the territory of another indigenous group, over the latter’s vocal protests. 29 Indeed, in looking for well-known examples to illustrate such predictions, one might consider the 13 colonies to be an emergent federated OC, and the Americans ultimately won the war of independence handily. Likewise, the Boers fought the British Empire to a near standstill in two wars when the latter was at the height of its power. Such cases are intended to be illustrative—there are both similarities and differences between such historical democratic and institutional revolutionary movements and the smaller contemporary communities under consideration here—but the fact remains that in each, the rebellious side was a non-state actor possessed of institutions capable of representing and mobilizing a defined community engaged in war via representative institutions, and with the resulting support (and resources) of a significant proportion of the population.  30 Fearon, in turn, draws on the rationalist assumptions associated with neorealist and neo-institutionalist IR scholarship (notably Waltz 1979; also Gilpin 1981, Axelrod and Keohane 1985). 50  case, will often collectively agree to sacrifice significant aspects of individual liberty and autonomy of action, both politically and economically, in service of group unity and pursuit of a larger goal. Given the realities of path dependency (Thelen 1999), such choices are difficult to unmake once made by a community, and lead to a restricted space for political participation for community members even after contention subsides as communally-oriented norms and behaviours persist.  Neither is there an expectation that all in the community are equally supportive of the projects undertaken. Crucially, this is a story about institutional decision-making and enforcement. Sharp disagreements can and do occur among and between leaders and members. Rather, it is simply that such disagreements are managed within the institutional framework. 2.4 Extensions, Scope Conditions, and Caveats 2.4.1 Rational Choices, Contingent Identities While the general foundation of the research project is rationalist-institutional, nonetheless given the considerable span of time under consideration—the primary case of contention in the Chapare plays out over about twenty years—the possibility that identities, the shared understandings of the world with which they are associated and the communal preferences and interests to which they give rise, may shift. The assumption of preferences and identities as fixed that informs rational choice analysis is just that, an assumption, one that can and indeed ought to be relaxed at need, and with sufficient empirical evidence in support of the associated process of ideational change, however conceptualized.  Focusing once again on the case of an OC engaged in contention with the state, I advance two interrelated claims. First, I argue that contingent strategic decisions can contribute to the construction of social identities. Over time, these contributions themselves become increasingly 51  hardened, forming in essence part of the set of resources and constraints influencing community members’ and leaders’ subsequent decisions. Second, contrary to more common understandings regarding the role of polarization and identity hardening in contentious contexts as a force tending to increase the likelihood and severity of conflict, I suggest an alternative possibility whereby identities can be hardened in a way that entrenches non-violent responses to extracommunal threats, as communities embrace and reproduce identities consistent with some forms of contention and resistance, but not others. Paradigmatic examples include the Gandhist campaign of passive resistance in pursuit of Indian independence, as well as the later iterations of the African National Congress’ fight against Apartheid in South Africa. I seek out some more general form to account for the emergence of such explicit adherence to non-violence in the context of extreme political contention. The Theoretical Context: Ideas in Conflict Concerns of ideational causes of conflict loomed large in early sociological studies of civil violence (Gurr 1970, Merton 1949 [Humphreys and Weinstein 2008]), but were more recently shunted aside in favour of more structural variables, and finally micro-rational level incentive structures. This is puzzling, given the obvious utility of a constructivist approach to conflict studies (whether labelled as such or not). The very fundamental problem that antagonistic groups are not fully and stably defined a priori prior to the onset of civil conflict, and remain fluid over the course of violence—indeed, it is often identity transformation on the part of one or more sets of combatants that proves crucial to the emergence of lasting peace—ought to be enough to justify it. Despite this fundamental insight, while most case specific studies make note of that fact, relatively few continue to use it. At the extreme, many 52  quantitative studies of conflict treat identity as fixed and captured in some ethnolinguistic fractionalization index, or similar static measure. More recently some more clearly ideationally focused scholarship has emerged, however. Wood’s study of the war in El Salvador drew attention to “the pleasure of agency” as a cause for participation in and support of insurgency among peasants (2003). Severine Autesserre’s use of international frames to account for the continuation of violence in supposedly “post-conflict” environments (2009, 2010) also falls into this category, as does Lee-Ann Fujii’s consideration of social network effects in the perpetration of violence during the genocide in Rwanda (2009) and the performative aspects of extra-lethal violence in various contexts (2013).  Other recent studies have begun to focus on the interplay between of rational choice and ideational framing and/or effect. For instance, Kalyvas and Sambanis (2005) develop the intriguing concept of “ethnification” to describe the process by which ethnic identities became increasingly deterministic over the course of conflict in Bosnia, not only out of enmity and cycles of vengeance but as a result of the limitations facing occupiers. Lacking the ability to effectively police conquered territory and fearful of betrayal behind the front lines, armed groups would demand total surrender, and punish refusals severely; over time this served to heighten sensitivities regarding the identities along which such demarcations occurred. Likewise, Fearon and Laitin (2000) draw out a variety of of elite- and mass-led processes by which ethnic identities might be rendered more salient, in general out of pursuit of rational gain by some set of actors. Kalyvas’ (2008) study of ethnic defection makes clear how important rational calculation regarding desirable ends (notably survival) is in the adoption and “betrayal” of particular identities.  53 Ideas and OCs  My own study sits within that latter tradition. In my own context, I am interested in the extent to which, and the mechanisms through which, meso level institutional forms and the practices they encourage tend to shape actor identities and preferences over time, and vice versa. Specifically, I argue that a form of path dependency can exist, such that socialization practices put in place in pursuit of some instrumentally outcome—for instance, support for some potentially costly campaign of resistance—will shape the preferences of community members in powerful ways, such that contingent choices lead to hardened identities, often entrenched in communal institutions. Community members come to accept certain outcomes as legitimate, and even vital, while others seem increasingly beyond the pale.  Importantly, while socialization studies in the context of conflict have often focused on the production of violence, notably in the service of military recruitment and training, as well as the parallel “shadow” processes of insurgent and gang recruitment and training, this need not be so. Indeed, a key outcome of note within my own research is the extent to which peaceful resistance — particularly defined — became a touchstone within the discourse, institutions and the actions of cocaleros. Indeed, it is not too strong to say that peaceful resistance became a component of both cocaleros’ shared identity and their public image. Accordingly, any shift towards a more violent strategy of resistance would have provoked a reconsideration of both, with all the potential consequences that would have implied. 2.4.2 Scope Conditions: The Contingent Promise of Ostromian Communities As a meso level phenomenon, there are limits to the effectiveness of Ostromian communities. Developments at the state level can overwhelm or otherwise over-determine results within a given country, what Straus (2012: 350) describes as “the centrality of a supreme national actor” 54  in domestic politics. Simply put, local institutional factors may be overwhelmed in the face of a state determined to, and capable of, engaging in armed violence or unrestrained repression targeting the viability of such organizations themselves. Sufficiently motivated and equipped states can bring armed violence unilaterally (Straus 2012: 347). This is less problematic than it may appear at first blush, however, as many of the cases we seek to explain are precisely those with a weak state apparatus, in which democratic institutions, to the extent they exist, are flawed in significant ways. One of the most powerful recent macro-level findings regarding conflict concerns the element of "feasibility" (Fearon and Laitin 2003, Collier et al. 2009, Hegre 2001, Hegre 2014). Simply put, civil armed conflict occurs where insurgency is most feasible because of some combination of human and natural geography, availability of financial resources, and limited state capacity. Moreover, returning once more to Straus, there is the possibility that the impact of meso-level institutions is a matter of sequencing—while a state once decided can overwhelm any local actor, certain meso level entities may be likely to forestall escalation to that point. This is, in fact, essentially one aspect of my argument, considered from a more dynamic perspective. While the state may have the capacity to mobilize and inflict violence against populations, in general it does so only in response to some perception of threat (Davenport 2007), and often as part of some pattern of dynamic escalation. The central insight of the repressive literature is that potentially vulnerable states—specifically, those lacking consolidated political institutions and/or those persisting at low levels of development—tend to engage in such violence reactively, responding to the presence of a perceived threat. While there are some objective characteristics associated with that perception of threat, much depends on the claims made by the groups themselves and, perhaps even more 55  importantly, the strategies employed in their pursuit. In the same work referenced above, for instance, Davenport (1995) suggests that the dissidents’ use of political violence, a broad range of tactics, particularly those that in some way transgress accepted cultural boundaries of contention within the polity, will be perceived as threatening by the state, provoking a more violent response as a result. There is thus the possibility that state opponents, having some knowledge of how likely a significant violent escalation by the state will be depending on types of resistance they undertake, will calibrate their strategies accordingly. OCs with leaders who understand the full costs of violent contention may thus look halt any escalatory spiral quickly, correctly foreseeing that the spiral would impose significant costs on the community without increasing the likelihood of a successful outcome. They thus will deliberately seek out collective strategies of resistance that are unlikely to provoke unwelcome violence upon their community, avoiding overly provocative actions. Potential elements of such a strategy include 1) Avoiding acts of physical political violence; 2) Working continuously to assert any actions that are taken—particularly those that skirt the boundary of what might be considered physical political violence—exist within the scope of the accepted repertoires of political contention within the state; 3) Engaging in more general campaigns to generate and maintain popular support for—or at a minimum tolerance of—goals and strategies of the opposition group, in order to buttress perceptions that the group’s practices fall within the context to support claims that the opposition falls within the boundaries of accepted or tolerated political discourse; and 4) Identifying and engaging with allies—both domestic and international—to help in advancing claims in points 2 and 3.  Structures such as Ostromian communities are thus significant in part because they constitute social actors that—barring internal fragmentation or some other sort of breakdown—are capable of a high degree of self-control and coordination within the community. Even when social grievances of whatever source provoke community members to make claims on the state or resist 56  new and unwelcome state intrusions, such a structure encourages members to do so collectively and in a manner that appears to offer the greatest chance of success at the lowest cost for the community as a whole.  Thus to summarize, while meso level entities such as OCs are bounded by the conditions of the state they find themselves in in their ability to shape outcomes—notably by macro level factors including regime type and the state’s repressive capacity—they nonetheless possess considerable agency to manipulate outcomes, particularly given the extent to which their own strategies shape states’ perception of threat, and the response likely to ensue. In Chapter Five, I return to the issue, evaluating the macro-level context in which contention occurred in the Chapare, before turning to a brief cross-country comparison of alternative cases in an effort to determine how robust such macro-level limiting conditions are, and the resulting implications for the meso-level OC theory of contention. 2.4.3 Extending and Integrating OC Theory While the arguments most often referenced in this dissertation relate to the emerging sub-field of contemporary conflict studies with its focus on insurgency and civil war, there are other relevant traditions of scholarship to which OC theory relates. The sociological study of agricultural rebellion and revolution constitutes one such tradition, albeit one that on the surface makes conflicting predictions regarding what one ought to have seen in the Chapare. For instance, while on the one hand both Paige (1975) and Wickham-Crowley (1992: 94-95) suggest that in theory areas dominated by smallholders might be comparatively less likely to give rise to rural revolutionary movements than those dominated by other forms of land occupancy, most notably share-croppers and squatters (respectively), the latter also cites and extends Scott's (1975) argument that peasants will rebel against attempts to alter their subsistence relationships. He 57  argues that threats to their ability to continue farming are most likely to meet with a violent response. Indeed, Wickham-Crowley goes so far as to claim that producers “who have been losing ground—literally—to landlords or the state, whose situations have become more precarious, or whose very existence as tillers of the soil is threatened by opponents, are our most likely candidates for radicalism” (1992: 118). Further adding to the riddle, Skocpol (1979) suggests that, among other things, peasants enjoying a high degree of independence and solidarity—which as I shall argue are two defining aspects of the Chapare cocalero community—are the most likely to be revolutionary. Thus, a theoretical puzzle. Should we expect cocaleros to be like conservative “kulaks” vilified in Marxian tradition given their status as smallholders, or revolutionaries in the face of an emergent threat to their livelihoods and their communal resources for revolution (Skocpol 1979, Wickham-Crowley 1992, Scott 1975)? This is precisely the sort of problem that process tracing, with its theory-laden within-case observational approach can shed light on. Accordingly in the concluding chapter I will return to the question of whether, and how, meso level OC theory can inform established theories of agricultural revolution. 2.5 Alternative Explanations Beyond the inevitability of explanatory factors that serve to limit the applicability of the hypothesized explanation beyond the primary case, there remains the possibility that one or more alternative explanation actually better explains the primary case itself. Accordingly, in this section I systematize the set of most plausible alternative explanations found within the relevant literature on conflict, and derive a series of testable implications for use in comparing them against my local institutional account in order to determine which provides the superior account of various crucial aspects of the narrative, consistent with the process tracing method. 58  Table 2.1 summarizes the leading alternative explanations, along with the result that I will provide for each. I group them into three general "families" of explanation, relating to fundamental necessary and sufficient conditions for insurgency to emerge and take root: motivations, recourses, and opportunity. Beyond the advantage accruing from its correspondence with several of the leading traditions of explanation currently found within the political science literature regarding conflict onset and escalation, the three-fold framework corresponds to what might be characterized as the “standard general model” of political violence first articulated by McAdam (1982) and recently restated by Davenport (2005: xiii-xvii).31 For each alternative I outline the primary mechanism(s) linking each to successful emergence of insurgency, and specify indicators for use in clarifying whether the explanation provides leverage in explaining a given conflict. In doing so, I draw in part on Humphreys and Mohamed (2005), Humphreys and Weinstein (2008), and others. At the end of Chapter Four I revisit this chart, evaluating the indicators for the presence and significance of each of the alternatives in explaining the Chapare case. There I note explicitly which alternative theories are relevant, and how (and how much, to the extent the data allows such evaluations) they bear on the outcome. Some are actually consistent with an OC account. For instance, the idea that a community would be more likely to act when confronted with an “existential threat” sits comfortably with the idea of a rational OC aware of the dangers it faces, but also of the potential costs that accompany various strategies. Likewise, both the role of macro level democratic institutions, and the potential for state-level repression are effectively                                                  31 McAdam (1982) in turn was working off a large tradition. He emphasized three general elements in the analysis of dissidents' behaviour: their level of organization, their belief in success, and the structure of political opportunities available to them to achieve their ends. Subsequent significant contributions include Tarrow's articulation of "framing" of conflict goals (1994), and Tilly's numerous reformulations of key concepts and modes of analysis (e.g. 2003). 59  Table 2.1: Alternative Explanations incorporated into my micro/meso/macro level framework. For others however, a positive result would infirm my hypothesized explanation—if for example a theory of greed-based resource Family Factor Sources Mechanism(s) Indicator(s) 1. Motivation  a) Grievance: social-identity based motivations for conflict Gurr; Stewart (2002); Fearon and Laitin (2000) No grievance means no means of motivating fighters Clear and articulated economic, cultural, and political grievances from identifiable groups in region  b) Presence of financial, material resources to support armed insurgency Collier et al. (2009), Ross (2004), Weinstein 2007 Presence of materials necessary to fund an insurgency, whether through looting or extraction via taxation Evidence of material capacity on the part of a would-be insurgent group sufficient to cover start-up costs of insurgency. May be due to presence of lootable resources, or identifiable population from which to extract resources.  c) Presence of existential threat Buzan (1991) Existential threats motivate and excuse radical actions not otherwise acceptable Subjective perception of a threat to a community’s livelihood or fundamental elements of its identity; Evidence regarding credibility of that threat  d) Availability of alternative participatory/ dispute resolution options Locke (1990 [1689), Kant (1939), McAdam (1982), Fein (1995), Hegre (2001) Grievances managed through ADR, esp. democratic processes Presence of institutions allowing for meaningful action on basic needs, pursuit of redress for grievances, etc.  2. Opportunity  a) Potential for repression by the state Davenport (2007), Tilly (1985), Gurr (1986) Threat of violence deters would-be rebels Presence of extensive security apparatus Examples of prev. failed attempts  b) Actual repression by the state Kalyvas 2005 Emergent rebels disrupted by state action Evidence of coercion undermining emergence of existing or potential insurgent groups   c) Lack of repressive threat to social groups requiring violence, due to state weakness  State sufficiently non-threatening to community that armed conflict never contemplated/ pursued Evidence of a state lacking the military and police capacity and/or political will to mobilize and repress identifiable subpopulations expressing grievances 3. Resources    a) Geographical situation favouring insurgency Fearon and Laitin (2003); Kalyvas (2006);  Physical barriers to the imposition of state control enable rebels to organize Limited state presence beyond urban centres; rural population; rugged terrain; access to borders   b) Availability of necessary ideational resources/framing Tarrow (1994) Dominant idea/ narrative to inspire disparate individuals, groups to common end Presence of a narrative linking together an identifiable group in a larger story of injustice and cooperation in resistance  c) Presence of/access to necessary expertise Tilly (2003); Gleditsch and Salehyan (2006), Lake and Rothschild (1998) Insurgency requires clear skill set, including procurement and violence specialists Presence of substantiated links, ideally with data on content of the relationships, between community members and armed groups elsewhere through which relevant expertise may be transmitted 60  mobilization, or some explanations employing the civil democratic peace theory can account for observed outcomes in the Chapare without support from the institutional account regarding the role of OCs, it could undermine the latter, suggesting that some other previously existing account within the literature is in some sense more convincing, more comprehensive, more parsimonious, or some combination of all three.  To summarize the results to come, all but two of the alternate theories fail to account for observed outcomes. In many cases, associated indicators suggest the presence of conditions that actually made armed conflict more likely, further cementing the Chapare's status as a "hard case" defying conventional explanation. The conditions requiring qualification concern macro level factors of democracy and repression. This I offer in sections of Chapters Four and Five, in part through demonstrating how the cocalero federation as an OC exhibited considerable agency over both the state’s level of repression and the effectiveness of democratic institutions. They accomplish the former by carefully controlling their own use of violence in a way that minimized state perceptions of cocaleros as threats to the state, and the latter by embarking on an ambitious and ultimately successful campaign of democratic mobilization intended to overcome a number of institutional and normative barriers to effective political participation by indigenous rural Bolivians. 2.6 Conclusion In this chapter I have presented a theory explaining how a particular kind of meso level institution that I refer to as an Ostromian community—a community with inclusive and bounded institutions to resolve collective action problems—will tend to reduce the risk of armed conflict in situations of contested incompatibility even as they work vigorously to resolve the situation in their favour. To develop the argument, I first defined the concepts of peace, contestation, and 61  conflict, before turning to the specifics of what a meso level entity entails and subsequently to the particular design characteristics associated with OCs, and how they bear upon political contention. I then examined the role of identity and in particular the effects of socialization on shared community identities, institutions and behaviours, before defining scope conditions expected to limit the effectiveness of the theory, notably regime type and regime repression. Finally I discussed a series of relevant alternative explanations found within the literature, generating testable hypotheses from each in order to facilitate comparative analysis of the those theories against the above account via the method of process tracing. In the next chapter, I turn to the specific case of the Chapare, presenting detailed evidence establishing that the community is in fact an OC, and that it possesses the characteristics I claim for OCs. Chapter Four then turns to the question of how the Chapare OC interacted with the state to produce the distinct form of contentious politics observed in the region during the period in question, including in particular the limits on violence observed by the cocaleros.  In Chapter Five, I provide a summation of the dissertation to that point before turning to questions of scope conditions for OC theory, and exploring the issue of potential generalizability. With regard to the latter, I provide a “plausibility probe” in the form of Peru—which I characterize as a “most similar” comparative case—and its experiences with Sendero Luminoso. Finally by way of conclusion I outline the primary contributions of the dissertation and outline potential next steps for additional research regarding the relationship between OCs and other meso level institutions on the one hand, and various patterns of contentious politics on the other. 62  Chapter 3: The Chapare Cocalero Federations—An Ostromian Community  3.1 Introduction Having developed a theory of OCs and how they relate to contentious politics, in this chapter I move towards an operationalization of the concept, first in general terms, and then in the specific context of the primary case of the Chapare region of Bolivia. Accordingly, I have four goals for this chapter. First, drawing of the work of Ostrom (1990) and others I provide discrete indicators for the various constituent elements of an OC, explaining how they may be identified in practice, and why they are necessary elements of the larger concept. Following this, I introduce the Chapare case, giving a history of how the region in the latter half of the 20th century. Third, building on the two elements above, I used the OC indicators to support the claim that the cocalero federations. In doing so, I draw out several hypotheses regarding factors that seek to account for the emergence of OCs—why it happened in this case, but does not elsewhere.32 In order to accomplish those goals, the chapter proceeds as follows. In the following section, I provide a brief overview of the idea of "descriptive theory" as a separate mode of explanatory research distinct from the more familiar causal approach to explanation.33 In section three, I develop a set of indicators for OC constituent elements. Then in section four I identify factors potentially contributing to the emergence of an OC. Finally, in the subsequent sections, I account for each of the OC constituent elements in the context of the Chaparean federations on the basis of empirical evidence.                                                 32 Owing to the research design and its focus on the years of contention, I was not able to undertake sufficiently precise research regarding the emergence and evolution of the federations to offer process-tracing based-explanation regarding the causal factors that account for the OC’s emergence in this case. I do, however, offer several hypotheses regarding factors that may be either solely or jointly responsible.  33 For an excellent treatment of the differences between descriptive and causal arguments, see Gerring 2012. 63    Foundational constituent elements: 1. Bounded communities; 2. Inclusive within those boundaries; and 3. Presence of a shared internal challenge taking the form of a collective action problem.  Governance constituent elements: 4. Ability to deliberate and decide on a common policy in response to that problem by means of some mechanism of collective choice; 5. Capacity to monitor and sanction defection by members; 6. Capacity to monitor and sanction defection by leaders and other agents; and 7. Dispute resolution mechanisms serving to strengthen within-group cohesion.  Relational (both intracommunal and extracommunal) constituent elements: 8. Degree of recognition and independence from the state (whether de jure or de facto); and 9. In the context of large communities a federated structure, in effect creating a network of hierarchically organized sub-communities.  3.2 Descriptive Theory As Gerring (2012) puts it, descriptive argument aims to answer “what” questions, in contrast with causal arguments that answer “why” questions. Gerring gives the example of Huntington’s (1991) theory that there have been three “waves” of democracy as an instance of descriptive theorizing; it tells us something about what exists in the world, rather than how things happen. As such, one might consider it as “ontological” theory. Descriptive theory has made something of a comeback in the last twenty years in both the comparative and international relations subfields, roughly coinciding with the rise of constructivism in the latter. The idea of identifying “constitutive norms” associated with international politics, for example, is very much in the same mode of explanation. Even so, it remains much less common than the more causally oriented mode of explanation, the linking of 64  cause and effect, that motivates much of scholarship in the discipline. Gerring (2012) and others (e.g. Collier et al. 2010: 35-36) consider this to be a mistake, arguing that description remains an important task worthy of attention in its own right in political science. In a discipline with few if any objective facts, every act of description carries with it the same theory-driven elements of selected simplification and emphasis that characterizes causal theorizing. Accordingly, it ought to receive the same kind of scrutiny and, for lack of a better word, disciplinary respect. Description often proves more “recalcitrant,” to use Gerring’s word (2012: 735), than explanation. One need only consider arguments continue over how best to define and operationalize core concepts like democracy, political culture, peace long after we have identified robust causal relationships about them to see evidence of how this can be so. In Gerring’s (2012) typology of description, the first part of the present study constitutes an exercise in typological descriptive theorizing: I posit that Ostromian communities—bounded and inclusive communities capable of implementing shared solutions to collective action problems—are one among several general patterns of meso level communal organization. As such, I am characterizing what Gerring would consider a multidimensional ontological entity, requiring various indicators for its different aspects—hence the provision of multiple properties associated with OCs, along with indicators for each, below. While development of the full set of alternative types in the typology lies beyond the scope of the present study, I do provide one potential list of alternatives in Chapter Five. 3.3 Operationalizing the OC  In this section I provide additional details about how each of the constituent elements of the OC outlined in Chapter Two function in practice, and accordingly how they may be observed and measured. Some of the indicators for the various properties of OCs are relatively straightforward, 65  while others, being more abstract, require more justification. I describe them all in turn in tables 3.1-3.4, identifying both institutional and behavioural characteristics associated with each. 3.3.1 Foundational Elements of an OC Table 3.1 identifies indicators related to the first three constituent elements. These measure necessary conditions for OCs to emerge.34 It must be clear who is in, and who is out of the community, in order to clarify who ought to benefit from and contribute to any collective enterprise, including political contention. Such is a requirement of any community.  Constituent elements 1 and 2: Bounded and inclusive communities   Institutional indicators    • Presence of a foundational constitution or covenant for the group, whether formal and written, or informal and oral, identifying physical space over which, and whether and how people may join or leave the community •    • Procedures to ensure that all members of the community are incorporated in community rights and obligations • Institutional expressions of shared identity encompassing all affected by the relevant collective action problems, who are thus part of the community   Behavioural indicators: •  •   • Existence of a widely recognized (i.e. by both members and non-members) territory over which the community exists, with shared understandings regarding the boundaries of that community •  •   • Understood practices by which members can join and leave (whether conforming with institutionally mandated practices or not) • Discourse reaffirming shared identities encompassing all within the community  Constituent element 3: Presence of a shared internal challenge taking the form of a collective action problem   Institutional indicators:    • Institutional acknowledgement of a shared dilemma requiring action.   Behavioural indicators: •  •   • Discussions of the problem in terms indicating that the participants understand the shared nature of the challenge, and the need for cooperation in order to meet it. Table 3.1: Foundational Constituent Elements of an OC                                                 34 Note that they fall short of necessary and sufficient conditions. I propose, but do not systematically test, a set of explanations for the emergence of OCs in section 3.4 below. 66   The inclusive nature of an OC, in contrast, is a more particular prerequisite condition for OCs as I employ the term. In this they differ from the CPRs described by Ostrom herself, in which inclusiveness and identity are not significant features. In an exclusionary community, or one lacking institutionalization, it is much less likely one would see “social rational” decisions; more likely the outcomes would pit some members of the group against others in the community. The key point regarding inclusivity is that these communities base membership and participation on principles such as residency or some other open category such that everyone who has a stake in the decisions taken by an OC has the opportunity to become a member, participate in deliberations, and share responsibility for their implementation. There exists some form of social contract among members. In contrast, communities that are grounded in some exclusionary identity marker, whether related to ethnicity, religion, tribe, or wealth or class, will be unlikely to produce outcomes that reflect the true diversity of perspectives among those affected. The results will be much more likely to produce division and conflict between group members and others in the larger community as a result. Indeed, one intriguing aspect of the primary case below is that the community itself became an important marker of identity, one that in many ways tended to trump pre-existing identities. As I argue in Chapter Four below, in the context of the cocalero federations, at least, that shared identity was significantly shaped and strengthened through socialization processes occurred within the intense contestation the cocaleros went through.  Admittedly, the question of “how inclusive is inclusive enough” may actually be a matter of degree. For instance, the exclusion of a small transient population of labourers from a rural decision-making institution’s representational structure may not matter much in terms of 67  outcomes, but the exclusion of a large permanent population of landless labourers almost certainly would (Wickham-Crowley 1992).35 Indeed, a question that emerges from the present research, one that cannot be adequately answered in the context of a single case, is “how” inclusive a community must be in order to exhibit the other properties associated with OCs. Finally, the presence of one or more collective action problems constitutes a definitional requirement, specifying how large an OC will tend to expand, given the presence of other necessary factors. In general, one would expect them to expand to the limit of the group for which some collective action problem exists. In the case of the cocaleros, we shall see that this implied the initial presence of numerous smaller OCs, each largely focused on meeting local challenges such as the provision of core social services like health, education, and shared infrastructure. As a new challenge emerged in the form of a threat to the coca economy on which all depended, the size of the problem made it possible, and indeed rational for the various communities to combine into a single much larger community, along the federated or nested lines that Ostrom suggested (1990: 90-99).                                                 35 In the case of the cocaleros discussed below, for instance, the federations were broadly inclusive of all families owning land in the region, but did not guarantee group membership and representation to the landless labourers that at various points were present in in the region, most notably during the coca boom, prior to the implementation of eradication as state policy. Likewise, representation does not extend to all adults in a family, but rather to one member per household. In practice, this seems to have had little effect on the overall pattern of contestation with the state—indeed, to the extent to which jornaleros [day workers] are mentioned, in the main they seem to have cooperated closely with owners in defending coca production sites and rights (e.g. García Linera 2005 et al. 399), but undoubtedly has had implications for the distributional effects in the community. With the rise of a new movement for Bolivians lacking access to land (Movimiento Sin Tierra, MST) in the last 15 years, combined with firm restrictions on the development of new catos for coca production in the region, this may undoubtedly become a more significant issue going forward. Indeed, one of the primary challenges cocaleros were grappling with during my field research was the issue of catos blancos, fields of coca production outside the quota allotted to producers in the region; as a new generation emerges without easy access to new coca catos, it seems likely the issue may become more acute. The solution currently being implemented—control social—effectively peer monitoring managed by the federations, has had some success in both identifying excess planting and reducing overall coca production levels in the region, but is undoubtedly not a perfect solution to the problem (Ledebur and Youngers 2013; multiple interviews). 68  3.3.2 Governance and Relational Elements of an OC Table 3.2 provides indicators associated with deliberative decision-making capacity. Element four concerns the presence of some deliberative capacity within the community. The mechanism has the dual effect of incorporating a broad range of concerns into the community’s goal-setting processes, and also—just as importantly—legitimating decisions once taken as representing the will of the community as a whole. This is similar to Ostrom’s own requirement that most individuals affected by the operational rules can participate in modifying the operational rules, though owing to the far greater demands placed on such communities I consider, I place emphasis not only on the ability of members to participate in decisions affecting them, but the reciprocal effect whereby such decisions, once taken, take on the legitimacy as expressions of communal will thus to be respected. Table 3.3 in turn outlines factors relevant to the enforcement of decisions once taken. These operate at two levels: ensuring mass compliance, and exercising leadership oversight. Various mechanisms provide for both. At a mass level both institutional and peer-based monitoring may take place. Such mechanisms in general will depend on some form of   Constituent element 4: Ability to deliberate and decide on a collective choice policy in response to that problem   Institutional indicators: •  •   • Evidence of an open and inclusive deliberative structure through which a binding decision can be taken and legitimated as an expression of communal will; institution may be formally or informal in nature.   Behavioural indicators: •  •   • Existence of durable shared understandings regarding how decisions are taken, what the rights and responsibilities of members are with regard to the processes of deliberation, decision, and implementation. •  •   • Presence of some central decision-making body enjoying authority over the membership; larger, more complex communities may distinguish between different elements of decision; some may separate out “executive” and “legislative” functions, for instance. Table 3.2: Deliberative Decision-making Constituent Element of an OC 69  contingent compliance, in which the large majority of members in a community tend to adhere to rules out of the expectation that others are doing so as well, with any shirkers suitably punished. Absent that expectation, far more would themselves attempt to shirk out of fear of being the “sucker” in a situation analogous to a Prisoner’s dilemma. These are effectively the same as those spelled out by Ostrom. Table 3.4 then provides two relational elements concerning how different units organize themselves within the OC, and how the OC as a whole is situated in relation to the state, the dominant external actor. Ostrom (1990) suggests that such communities will be organized in a  Constituent elements 5 and 6: Capacity to monitor and sanction defection both among members and leaders   Institutional indicators: •  •   • Presence of institutional procedures providing some form of oversight and/or accountability of group leaders with regard to the membership. •  •   • Presence of a mechanism for the application of graduated sanctioning in the event of shirking on the part of membership with regard to the shared responsibilities of group members. May take the form of peer-to-peer sanctioning (as in the case of some common pool resource institutions discussed by Ostrom 1990) or institutionally implemented penalties.   Behavioural indicators: •  •   • Regular meetings of group responsible for accountability and oversight (unless invested in an individual) •  •   • Evidence of an assessment process. •  •   • Evidence of sanctioning of leaders. •  •   • Evidence of peer-to-peer monitoring processes identifying and successfully sanctioning shirkers. •  •   • Evidence of institutional monitoring processes identifying and successfully sanctioning shirkers.  Constituent element 7: Dispute resolution mechanisms serving to maintain within-group cohesion despite disparate within-group interests   Institutional indicators: •  •   • Presence of formal or informal processes for community members to consult or employ in the resolution of problems without the involvement of external actors   Behavioural indicators: •  •   • Evidence that actors in fact use resolution mechanism rather than pursuing either state or private justice Table 3.3: Monitoring and Enforcement Constituent Elements of an OC 70   Constituent element 8: Degree of recognition by and independence from the state   Institutional indicators:    • Some degree of autonomy enjoyed by OC institution(s) with regard to state institutions. The recognition may be formal or informal (i.e. de jure or de facto); likewise it may be active or passive (i.e. marked by positive recognition from the state, or be a function more of benign neglect or reluctant tolerance).   Behavioural indicators:    • Evidence that the state either does not challenge the right or reality of such groups’ existence, or is unable to mount such a challenge successfully. •  •   • Evidence that the state recognizes OC institution(s) as an interlocutor, engaging with them in various ways such as negotiations, agreements etc.  Constituent element 9: In the context of large communities a federated structure, in effect creating a network of hierarchically organized sub-communities   Institutional indicators: •  •   • Existence of multiple community sites, with hierarchically constituted linkages, whether formal or informal, between sites, enabling the formation of joint policy across all. •  •   • Mechanisms to enforce compliance with agreed-upon policy between sites   Behavioural indicators:    • Evidence that sub-community sites do indeed work together and maintain a common front to mutually decided policy.     • Evidence that sub-community sites monitor one another, and sanction free riding and other kinds of defection from agreed upon policy. Table 3.4: Relational Constituent Elements of an OC  federated structure when encompassing multiple sites, or becoming sufficiently large so that communal mechanisms necessary for peer monitoring (and trust in the efficacy of peer monitoring) cease to obtain. In this way, reliable decision-making, implementation, and enforcement mechanisms exist at each level, while coordination across sites can be organized as necessary within the federated structure. At the same time, Ostrom suggests that some independence from and recognition by the state is necessary for her common resource pools (CPR) institutions to operate. In the context of contentious politics such requirements take on quite a different tone however, as the state has a clear interest in weakening or even eliminating oppositional organizational structures. Accordingly, the question becomes more specifically about power, and as such the behavioural 71  measures are quite important. Legislative rights to organize may not be respected by state agents in practice, and a given OC’s continued independence may be at least in part a function of its ability to successfully resist and withstand efforts to undermine it. 3.4 Explaining the Emergence of the Chapare OC  A central argument of the chapter is that the cocaleros governed themselves as a federated OC; I also provide an account of how such a community came to be, though I do not rigorously test it against other alternatives. The rigorous testing of an explanation of how and why the OC came to be, while an important task in its own right, lies beyond the scope of the present study. Importantly, in explaining the emergence of an OC in the Chapare, and by extension OCs more generally, the account here must be considered an exercise in hypothesis generation rather than testing. Each of the above factors appears to play a role in the emergence of an OC capable of decisively influencing the pattern of contentious politics subsequently observed in the Chapare, but given that I have only one “observation” of OC emergence in this study, it is not possible to effectively weigh the relative importance of each of the above factors, or assess the extent to which any one or combination constitutes either necessary or sufficient conditions. Much hangs on the question whether some or all of the above conditions are necessary for the emergence of an OC. In the case of the former, where a subset of conditions is necessary and sufficient, they may prove widespread. In contrast should it be the case that the full set of conditions—and indeed perhaps other salient factors not named here—are necessary, then it may be that relatively few communities are likely to take on OC characteristics, at least within the set of cases in which we are interested—i.e. those where conflict is possible, yet not assured. This is a question that thus cannot be answered within the context of a single case study; further research will be necessary to resolve it. 72  So, how and why did the internal institutions governing cocalero communities become OCs? I argue that they did so in response to shared challenges of collective action facing farmers following their settlement in the region. In making that case, I draw attention to these points within the following narrative:  a.  The overall lack of support from the government for fledgling communities and precariousness of life in the jungle settlements that mandated an approach of self-reliance among colonists. It is in part due to that isolation in such forbidding territory that communities turned to communal cooperation, relying on one another to achieve ends beyond those possible for settlers left to their own devices. Put differently, the isolation and terrain presented the communities with significant collective actions at the moment of founding, requiring collective response to be “hardwired” into the communities from the start.  b. The relatively recent date of that settlement of the region, with communities for the most part emerging in the last 50-60 years, opening space for the new arrivals to craft institutions in response to shared problems relatively unencumbered by choices of the past. Also relevant may be the relative flexibility to institutional change within the community. i. An alternative to consider is that the relevant point is not the recency of the OC’s emergence per se, but rather its institutional capacity to cope with new circumstances, including new threats. In the context of the present case, that implies that it is significant not just that the cocaleros constituted an OC, but that they constituted one equipped with malleable institutions that nonetheless preserved a degree of predictability for members because of their status as written constitutions generally adhered to. In theoretical terms, this might be understood in terms of institutional stickiness (Thelen 2002: 392-396; 2002: 99-101), the idea that very inflexible institutions will not respond well to changing circumstances, particularly the emergence of new threats to communal interests.  c. A degree of rough equality among members. Effectively, equality may matter for two reasons. First, there is the general point drawn from democratic theory that from a legislative perspective, unless all interests are represented in voice and vote, the community will not be able to integrate the full set of interests present in the community required for socially rational decision-making to emerge.36 Second, there is the point taken from collective action theory that, other things                                                 36 In a brief survey I was unable to find significant research regarding the internal politics of trade unions’ delegative governance structures, which the cocalero unions obviously resemble in important ways. Given the particularities of trade union delegate governance structures (explored for instance in Ford 2002), this would seem to be a worthy area of further research, both for the Bolivian and wider labour contexts. 73  being equal, broadly shared joint interests are much less likely to exist in communities exhibiting extreme disparities in wealth. Rich and poor members will have divergent interests in the community; this in turn makes collective solutions to shared challenges much less likely to emerge.   d. The presence of actors motivated by and able to respond to economic incentives rather than, for example, the subsistence-based moral economy associated with peasantry (Scott 1976). This speaks directly to the sense in which we may consider one community or another to be behaving “rationally.” A group of actors that have come together to maximize their standard of living through cooperation in pursuit of economic gains will be quick to perceive and respond to threats to that economic well being. Conversely, groups organized around and dependent on long-standing rights and responsibilities, such as those found in the traditional peasantry may act more reactively, engaging in sporadic and primarily reactive acts of violent contestation when threatened without a prior expectation regarding the likely response from the state, potentially provoking sustained violent repression as a result.  e. The availability of existing cultural-institutional forms on which to draw. i. An additional ideational component to consider is the “strength” of identity with the community as a possible explanatory factor in an OC’s relative ability to respond to a threat from the state or some other extracommunal actor. Thus, the prior salience of OC identity may prove to be a variable explaining subsequent OC effectiveness at managing conflict across cases. This too must remain a matter for further study.  f. A degree of institutionalization of decision-making structures such that offices are separate from individuals. There are several ways to support this claim theoretically. At a general level, it is reminiscent of Tilly’s (2005) characterization of modern states being defined and characterized in part by the emergence of impersonal state-social relations, ensuring the principles of rule of law—a vital prerequisite of agent oversight—hold not only over the masses, but leaders as well. This in turn ensures that a) leaders are held to account, lessening the possibility that they will act in pursuance of private ends, and b) encourages competition among actors for leadership and by extension competition of ideas for primacy, thereby making rational decision-making more likely than in circumstances in which decision-makers operate without the need to justify their ideas and decisions against other available alternatives. One might also compare it to Weber’s legal type of authority, and contrast it with traditional and charismatic alternatives (2004 [1919]).  74  3.5 The Primary Case: The Chapare The Chapare region is a tropical and subtropical area of the department of Cochabamba, lying approximately 100km to the east of the department’s eponymous capital and nearly 2 km below it. Consisting of dense rainforest, the region receives more than 3800 mm of rain annually.37 The population remains predominantly rural, with the vast majority living in the country. Accordingly, agriculture remains the most important industry, though the easternmost areas of the region, notably in the area governed by the Federation Bulo Bulo, do have some level of natural gas production as well. Table 3.5, is based on data from Salazar Ortuño et al. (2008). Federation Founded Centrals Sindicatos Families Federación Especial de Trabajadores Campesinos del Trópico de Cochabamba (FETCTC) 1975 31   299 21,726 Federación Especial de Colonizadores de Carrasco Tropical (FECCT) 1984 28 419 14,678 Federación Especial de Yungas Chapare (FEYCH) 1988 6 66 2,646 Federación Especial de Colonizadores de Chimoré (FECCH) 1964 9 58 2,522 Federación Unica de Centrales Unidas (FUCU) 1986 8 57 2,071 Federación Agraria Mamoré Bulo Bulo FAMBB) 1996 4 29 1,857 Totals  86 935 45,539 Table 3.5: List of Coca Federations of the Chapare Though known collectively as “The Chapare,” the region in question actually includes areas of four different provinces within the Department of Cochabamba, including portions of provincias Chapare, Tiraque, Carrasco, and Araní. The most important political lines of demarcation among coca producers, however, relate to the now six labour federations that govern rural producers’ internal relations, and represent them collectively to the rest of the country. The six federations are listed in Table 3.5. Of these, the first two encompass the majority of members and exercise significant influence over internal decision-making processes                                                 37 By way of comparison, at about 1100mm per year, notoriously rainy Vancouver B.C. receives less than a third of that amount. 75  in the coordinadora era—marked by the emergence of an initially ad hoc committee in the late 1980s to coordinate and direct action throughout the six federations (Salazar Ortuño et al. 2008: 127-132). The formal coordinadora was subsequently established on 24 February 1991 (Salazar Ortuño et al. 2008: 19, 175-176).  Figure 3.1: Map of the Six Federations of the Chapare. Source: UNODC. The history of the various federations’ emergence is itself interesting, offering some insight into the particularities of Chaparean and Bolivian politics. Interestingly, prior to the years of contention the members, and especially the leaders, of cocalero federations seemed more interested in fragmentation than integration. Indeed, the creation of a new federation was one way for aggrieved members to chart a new policy direction when defeat of an entrenched leadership cohort seemed unlikely. That process largely reversed itself however, with the rise of 76  a threat to collective interests in the form of the state’s pursuit of coca eradication. That development served to quickly reunite the cocaleros, producing a movement more organized and more closely coordinated than had been seen previously in the region or indeed in the whole of Bolivia. In the following sections I outline the characteristics of those federations, specifying how they constitute Ostromian communities. 3.5.1 Foundational Constituent Elements  The bounded nature of the cocalero community in the Chapare is a function of the geography of the region along with its particular settlement history, enhanced more recently by its emerging coca economy. The area was in fact settled in substantial numbers comparatively recently. The first significant highway through the area connecting the cities of Cochabamba and Santa Cruz was completed only in 1954 (Cole 1958), which opened the region to new settlement. Even after its completion, rivers continued to provide the most effective means of accessing many areas. The majority of present inhabitants arrived in the last 40 years as a result, within living memory, as part of a larger process of internal colonization of the country. Movement to the Chapare was, by turns, a function of government policy and economic pressures elsewhere in Bolivia (Spedding 2004; Salazar Ortuño et al. 2008; Painter 1994). Given the rough terrain and the relative lack of infrastructural support from the government—even in the case of government facilitated colonization—the colonists tended to settle in small, close-knit farming communities located near the major transportation routes into the region, both highway and river. The forbidding nature of the terrain ensured that the boundaries to a typical colony were self-evident upon founding. Owing to the need to provide collective assistance for one another in order to meet basic communal needs, the communities were often established with explicit communal agreements of shared responsibilities, resulting in dense networks of shared obligation 77  owing much to similar practices observed in the Bolivian highlands, the point of origin for many new arrivals (García Linera et al. 2005: 393). The communities that emerged remained impoverished compared to the rest of the country well into the 1980s, when the cocaine boom was well underway (Salazar Ortuño et al. 2008: 89-90). Whether their arrival was spontaneous or government-directed, colonists continued to eke out a subsistence existence into the 1980s, supplementing their farm goods with fishing, hunting, and gathering in the abundant rainforest surroundings (Salazar Ortuño 2008: 53-54, Spedding 2004).38 While the state undertook the project of universal education in the wake of the 1952 social revolution, literacy levels remained low (Salazar Ortuño et al. 2008: 81).39 Access to public services was likewise limited, with just four government health centres present in the region as late as 1990 (Desarrollo Democrático Participación Ciudadana 1997: 49).40 In short, new arrivals were forced to rely in the main on themselves for support.  Several elements are noteworthy in this history of settlement, serving both to heighten the puzzle of contention in the Chapare, while also pointing to an explanation for the emergence of an OC as well. As a destination for voluntary economically driven settlement, the region attracted a population initially composed of landless or otherwise resource-poor young males. Some came because of a lack of economic opportunities in their home region; others were driven by changing physical conditions such as intermittent drought in the highlands (Painter 1994) or                                                 38 The typical cocalero’s income shifted significantly in the early 1980s with the onset of the so-called “coca-boom.” The boom, a direct result of the global popularity of cocaine at the time, brought about a rise in price in the commodity sufficient to allow some cumulation of wealth among cocaleros for the first time. (Painter 1994: XX) 39 Indeed, interspersed among the calls for licit coca production were regular appeals to address shortfalls in education, health provision, and so on.  40 Indeed, even during my fieldwork in the region in 2012 and 2013, the main health centre in the regional centre of Villa Tunari had a medical centre staffed for the most part by foreign nationals, particularly from Chile or Cuba, either volunteering or funded through support from a foreign government.  78  by incentives provided by the state, whether positive in the form of inducements for settlers or negative in the closure of state industries and massive layoffs among miners and other state employees during the economic restructuring of the 1980s. RV, a long-time leader in Central Unidas federation forcefully expresses both the need that drove migration to the region, and the grievance that that want generated: Obviously in Potosi [one of the highland departments in Bolivia from which many Chaparenos migrated] there was a degree of poverty almost 92%, if I’m not mistaken. In other areas it was 70%, a high degree of poverty in different departments because there was a total absence of the state. There was nothing… When there had been nothing, the poor becoming poorer, migrating in hunger, in misery, and the rich getting richer. The money, resources were circulating in just a few hands, and the other hands were left without resources in extreme poverty. No one ever spoke of health or of education, there were…[divisions between] first and second people. Some privileged with more rights, other people in the country without rights, people in the second category.  Consequently, the region was settled by the building blocks for revolution in many other contexts: young, landless, unattached males, whether motivated by feelings of grievance against the state or driven by greed and the possibility of payment or plunder (Goldstone 2002; Urdal 2006), or indeed prompted to turn to violence as part of a survival strategy made possible in part by a sense of detachment from political processes in general, and the anomie that results from that isolation (Humphreys and Weinstein 2008).  That history distinguishes the region’s settlers from rural populations in other parts of the country as well, however. As economic migrants entering what was for them41 relatively virgin territory, they were more like frontiers-folk in many ways than typical indigenous communities                                                 41 There was a well-established population indigenous to the region prior to the period of colonization. However, the communities were small and widely dispersed, such that the settlers soon came to outnumber them by a wide margin. While some lowland-indigenous regions came under state protection from settlement, such as the Territorio Indígena y Parque Nacional Isiboro-Secure (TIPNIS) to the north of the Chapare for example, the region as a whole came to reflect the cultural and political organizational preferences of the newcomers. This at times created visible tensions between the different kinds of “originarios” (Grisaffi 2010), though the issue has achieved greater salience following the rise of the MAS to power. 79  found throughout the country. Rather than endeavouring to live with the land according to complex received traditions, as in the case of established rural Bolivian communities, or in opposition to the state and mining interests as in the unionized mining centres, the new Chaparenos were comparatively free to make their own way. RV makes the point in this way: You are asking me why did miners come here? … Most of all, they didn’t have anything. The mines were closed, and they have come here above all in search of survival, in order to survive. They have come not just to plant coca. Take me, for example. I did not come here to plant coca, but rather I came here to live. Why? Because here there had been a good life. Obviously I lacked nothing to eat. There wouldn’t be much money, but I have rice, for example—It’s a matter of seeding. I have fruits, I have hunting for fish; I have everything, though it’s a bit tight in some ways.  Beyond being economic migrants to the region, the new arrivals were necessarily self-reliant and self-employed as a result of their circumstances. A rough equality existed among members from the time of founding. Family plots remained small on average, though there were a small number of larger land grants made, particularly during the years of dictatorship. Some of these were subsequently broken up and sold to landless residents in the Chapare following a poorly documented process of land invasion. (Salazar Ortuño et al. 2008: 95-96, 113-114). Though the emergent colonies, or syndicates as they increasingly became known, still emphasized the importance of communal harmony, they and the locally grounded social contract on which they depended were built to meet the specific needs of the new arrivals. In this way, the communities differed markedly from the both the highland ayllus42 and tin mines from which other settlers came. Whereas long-established custom governed highland communities, and the miners’ unions concentrated on social and political disputes with the owners and the state, the                                                 42 Ayllus are the traditional political organizational unit of the Aymara highland and Quechua valley communities. Their organizational form is quite complex, with multiple levels of nested hierarchy. The Aymara ayllu leadership culminates in the office of the “Apu Mallku” or supreme leader, a title held by Evo Morales since 2006 (Cusicanqui 1990; Wright and Paux 2013). 80  Chapare colonies were free from constraints of history and historical struggle alike. At the same time they were able to draw on memories and organizational resources of both; as a result, the Chapare communities took on their own form, one well suited to the particular circumstances settlers encountered upon arrival but drawing on those previous organizational forms. That equality was further reinforced by the nature of coca production, which did not lend itself to the emergence of large plantations. While many cocaleros took on assistance to cultivate additional coca acreage, particularly during the peak boom years, as prices declined so did the practice (Spedding 2004). That rough equality played an important role in both the internal and external relations of the federation in later years, with coca federations enforcing limits on the size of family coca plantations as part of its agreement with the state.4344 While many upon arrival practised a form of subsistence agriculture not unlike that found throughout the highlands, the region’s producers also looked actively for cash crops to produce for sale. They were undoubtedly campesinos45 upon arrival, scratching out a living by any means necessary, combining subsistence agriculture with whatever they could hunt and forage in the surrounding jungle. EB, another leader from Centrales Unides, did not mince words in                                                 43 The disposal of land remains a complex topic in the region. On the one hand, the land market is much more open than on the Bolivian highlands where familial relations play an important role in determining ownership and usage (García Linera et al. 2005: 393). In the Chapare the land may be bought and sold on an open market. While I was in the region I saw a Chapare land sale advertisement posted outside the Chapare-bound bus stop in Cochabamba specifying the size, and the fact that it came with a cato of coca. At the same time, those transactions are regulated by the local syndicates, which claim the unambiguous right to bar the sale of property to anyone considered politically “suspect,” for lack of a better term. These principles are closely enforced in practice as well: effectively no one buys land without also buying into (or at least acquiescing to) the political project of the federation. 44 The method through which such limits are enforced, known as “control social” remains controversial both in the country and among Bolivia’s international partners, though evidence to date suggest it is proving successful in limiting excess production. See Ledebur and Youngers (2013) 45 In making the following distinction I am following the definition, common in the literature, that campesinos constituted a form of transitional mode of existence, combining elements of a subsistence existence with some presence within the agricultural market. 81  describing the hardship he experienced both prior and subsequent to his arrival, describing it as follows: The life of a miner is fucked. Well, ever since I was born, in the department of Oruro, in the province of Charapata, that is in Avaroa province. I was from a very poor family, as poor as Evo, or perhaps poorer. At least Evo played an instrument; I didn’t play anything. We had sheep and llamas, but we were poor. Well, as we were poor, I had to support myself in order to study in school. In whatever manner: selling ice cream, picking up bread, and other things I worked at after school. In ’72 we came to the Chapare with my father to colonize. We took a lot, in the brush in Lauqueña… in ’72. Those areas were deserted. There were no houses; now there’s a city. In Shinahota there were no houses, not a single house.  Imagine, as we had little money, I had to stop my studies for three years, but I had to suffer much in the Chapare. There was no food, so I was short and skinny. While I was developing my bones I didn’t consume eggs or milk, because in those times I wanted to, but there was none. We ate rice and bananas, rice, bananas, and sometimes cassava [a starchy root vegetable]. So imagine, I had to go back to thinking while studying [on that diet].   That tradition continues to the present to some extent. As Asterio Romero, a long time coca leader and Secretary General of the Department of Cochabamba [effectively the second ranking official in the department] put it in passing during an interview: In Chapare it is normal for each person to have a rifle, but not to deal with the police, or to fight, no. A 16-gauge rifle is for hunting. There are forest animals like the lowland paca [a large rodent], the armadillo, the peccary [a kind of wild hog], the tapir. People can’t go far to buy meat, so they kill those animals and eat, like so.  Even so, as they came to rely increasingly on agricultural sale for income—whether in the form of coca or the various cash crops such as pineapple, citrus fruits, passion fruit, and palm hearts that were advocated by the state—the region’s inhabitants resembled the archetypical campesino less and less, and economically motivated smallholding producers more and more. Indeed, it is notable that a number of the most significant confrontations involved government attempts to close coca markets. Such attempts provoked furious responses from the cocaleros, including the initiation of flash blockades in their defence (Salazar Ortuño et al. 2008: 557) and attacking and 82  killing state forces, an unusual act for the normally disciplined cocaleros (Salazar Ortuño et al. 2008 644-645). RM, a veteran of confrontations over coca (and at the time of our interview President of the MAS-IPSP in Federation Carrasco), described it like this: Afterwards the fight continued, in 2001 and 2002 in Cochabamba. No one could even “pijchear” [traditional method of consuming coca by chewing it] coca, as Law 1008 was applied and the “pijcheaba” was a form of drug addiction for them, whereas for us it was food and medicine that gave us more strength and more spirit to work. So, we continued pijcheando and it did us no harm, it did not drug us. But for the gringos [Americans] it was another matter, they were saying that coca was a drug. Three deaths we had in Sacaba, when they wanted to close the markets, and we have not stopped. We defended the organizations, and now we continue the coca market, but it has not been easy for all of that. Two soldiers were killed and we had three civilian deaths.  While it foreshadows discussion of conflict that is the subject of the next chapter, the passage clarifies how important a cash crop and market access had become to the cocaleros. This in turn sheds light on how cocaleros’ livelihoods had in diverged in specific ways from other indigenous actors in the country, where smallholding remained synonymous with subsistence.  Returning to the prior discussion, from the moment of arrivals the new settlers enjoyed considerable freedom to innovate around the purpose and structure of community institutions. While obviously aware of the traditions from which they arrived, nonetheless they were free to interpret, apply, and modify those understandings as circumstances required, first embracing and subsequently defending a move to a more modern, profit driven form of agriculture. Thus, several hypotheses emerge regarding the factors accounting for the emergence of an OC. It may be that such institutions require some or all of the following conditions: rough equality among members at the time of institutional founding and/or throughout the period of contention; the availability of ideational and cultural antecedents valuing organization in pursuit of collectively valued ends; and/or the presence of “windows of opportunity,” or critical junctures to design local institutions that match local requirements.  83  In any event, the relationships formed among the newly arrived inhabitants reflected the initial rough equality and the necessity of self-reliance. They reflected the challenges faced by members upon their founding. Moreover the public, impersonal and non-teleological nature of the institutions and offices made such organizations more adaptable as circumstances changed and new challenges arose. Instead, agricultural migrants to the region entered into agreements with neighbours much more reminiscent of a Lockean social contract. The goals of the communities were universally practical, even prosaic. As García Linera et al. put it, “the union in its genesis was an entity where families organized themselves to distribute lands, regulate the supply of labour, and organize the local authority to manage daily life” (2004: 390). Accordingly, in the syndicate “each member must be present to open paths, weed fast growing weeds, maintain commonly used buildings like schools and soccer fields, and so on” (2004: 389, 393). As limited as the aims were, the act of founding a new syndicate was significant for the members. As Blanes and Flores describe it, “[t]his constitutive act of the syndicate was truly fundamental; it represented the birth of a new colony.”46 There was little larger political purpose behind such organization, at least to begin with, beyond advocating communal interests and managing relations with the state in the case of the state directed settlement, or acting as a point of liaison in the context of the indirect patterns of state-social relations in the region, and the accompanying clientelist system of resource distribution that persisted into the democratic era. Indeed, that indirect system remained quite stable from the post-revolutionary period, through the military-campesino pact, and in some                                                 46 Quoted in García Linera et al. (2005: 393).  84  areas of the region persisting into the early democratic era.47 The following quotation gives some flavour to the deeply limited civil governmental presence in the region prior to devolution and the increase in local funding that attended the 1994 Popular Participation Law that, among other things, converted mayorships to elected offices and devolved some funding to local levels of governance for the first time. Prior to that change, as one expert put it,  In some cases it was just the mayor [present in a given region], in some cases it was just the sub-prefect, or the mayor, but there was no institution to support him. It was just one man. For 20 to 25 years they have managed things so. Now it has changed (Jose Luis Mamani).  Indeed, while in the wake of '52 the Bolivian state changed markedly in many ways in terms of its goals and institutions, it remained sharply limited in terms of its capabilities. Moreover, for all the progressive acts it took on issues such as universal suffrage, the principle (if not the practice) of universal education, and perhaps most importantly land reform, barriers remained between the urban mestizos who dominated government and the pinnacle labour organizations of the country on the one hand, and the indigenous rural majority of the population on the other. This was not restricted to the Chapare, either; as Faguet puts it, prior to the enactment of the Law of Popular Participation in 1994, “beyond the nine regional capitals (including La Paz) and an additional 25-30 cities, local government existed in Bolivia at best in name, as an honorary and ceremonial institution devoid of administrative capability and starved for funds. And in most of the country it did not exist at all” (2004: 6).   Accordingly, an indirect pattern of governance remained intact. Communities continued to be responsible for significant areas of public policy considered the purview of modern states.                                                 47 As I describe in this chapter and the next, while some of the cocalero federations broke relatively early with the state, others continued to maintain relatively deferential attitudes towards the state—reputedly encouraged through considerations provided to local leaders—until a new generation of leaders emerged and took control via election in the early years of contestation over coca.  85  Government restricted its activities to the imposition of rural taxes (Klein 2011, Dunkerley 1984), the provision of certain goods subsidies, market price regulation, the maintenance of public order, and the provision of basic-level education and limited healthcare.48 That pattern continued by and large during the period of dictatorship stretching from 1964-1982, as the government took as its primary goal the maintenance of regime stability. Local concerns were largely left to local populations to settle for themselves, so long as they undertook no action deemed threatening to that stability, which in practice meant avoiding anything resembling overt political opposition. Many in consequence sought to limit exposure to political actors altogether, even at the cost of access to the few state resources available. As one former cocalero leader put it, Of course, in a matter of social works for example, when a school failed, or a first-aid station, or a water system, before the political instrument emerged, people did not go to city hall [then appointed by the state] for help. Perhaps out of dignity, because everything there was conditional: “if I give you this, you gotta give me this; if I give you this, you have to vote for me.” That was the way before. It was out of dignity I think that people never asked the mayor for their due, rather they drew a line and built the school, the first aid station themselves. Until just recently my central49, Ansaldo, has had wood for its health post, but only just recently they went to construct the rooms. The community itself built it (EC).  As alluded to above, owing to the particular history of Bolivia, and by consequence the specific experiences and cultural resources that the new arrivals brought with them, the natural templates for settlers to draw upon in undertaking these new agreements were the communal highland ayllus and the miners’ unions. The most common name for the resulting form, the                                                 48 Interview participants confirmed such impressions, with several of the earliest arrivals that I spoke to describing an area almost totally devoid of government presence upon their arrival until the increasing militarization of the area during the 1970s and 80s. 49 The federated structure of the six federation includes three hierarchically nested levels of organization: the sindicatos, which are union locals normally consisting of no more than 80 families; centrals, which are groupings of about ten neighbouring sindicatos together; and federations, which vary widely in size (see table 3.5). 86  sindicato campesino (peasant union), neatly captures the duality of influences, though it may obscure the distinctiveness of the agreements described above. Notwithstanding that distinctiveness however, some discussion of prior organizational forms on which the campesino union locals were based is in order. A full explanation for that development would be relatively involved. In brief however, the state in Bolivia until the late 20th century maintained an indirect relationship with the majority of its citizens. Indeed, it was not until the social revolution of 1952 that the indigenous, rural majority of the population achieved full recognition as citizens; prior to that time, the relationship between the state and such groups was mediated through mechanisms such as group taxation (Dunkerley 1984, Yashar 2005, Lucero 2008). Even in the wake of universal suffrage, rural indigenous populations were unwelcome in the formal politics of the state and even in the context of the country's urban- and miner-dominated national labour movement. They were thus excluded from both the state and primary opposition organizations (Dunkerley 1984; Escóbar 2008).50 At the same time, there was a strong current of thought among indigenous and campesino Bolivians themselves resisting engagement in formal politics.51                                                  50 In particular, Bolivian rules surrounding the formation of political parties made it relatively difficult to form parties even in the era of full democracy that emerged following the transition of 1982. The country’s electoral court often rejected applications to register new parties; for this reason, the current ruling MAS-IPSP party took its name from an existing but dormant party unrelated to their project after being rebuffed in attempts to register their own party. The informal barriers included resistance on the part of many urban mestizo Bolivians to the presence, let alone participation of rural indigenous Bolivians in the country’s ruling spaces. As recently as 1926 then President Saavedra enacted a law prohibiting Indians in the country from using sidewalks or entering Plaza Murillo, site of Congress and the Presidential Palace in the capital of La Paz (Gotkowitz 2008: 62). 51 This reticence was in the former case related to the implied endorsement of a bourgeois political system such participation represented. Indigenous communities were also heavily influenced by the emergence in the late 1970s of Katarism, a movement of indigenous awakening built upon reference to images of a militant indigenous past and focused on the creation of a indigenous-based community separate from, rather than integrated with the country’s existing republic with all its Western, mestizo and creole values (Yashar 2005: 154; Patzi Paco 2007). 87  Accordingly, the natural pattern of organization for cocaleros upon arrival in the region throughout the period of colonization was to create small, largely self-governing communities. By mutual need, they encompassed all who settled in a particular area at a given time, normally no more than 100 families per settlement.52 When new colonists arrived in the community, they implemented forms of local governance reminiscent of the communities from which they arrived. While the size of plots varied, the fact that most arrivals obtained their own land upon arrival ensured that members of a given federation had a similar stake in the outcomes of the community.53 The structure of that relationship created bonds of common interest among members. More than this, the strong emphasis on the primacy of communal goals and the ongoing acceptance of this emphasis gave unions and, later, federations a great deal of leverage over their members, as campesinos’ continued residence and livelihood were contingent on their ongoing membership in good standing in the local union and commitment to whatever courses of action the group ultimately decided. In consequence, the communities that emerged resembled what García Linera et al. refer to as the sindicato campesino, or peasant union.  The “peasant union,” beyond the name, has little to do with labour unions; rather, it designates a type of traditional association unified by obligations and rights regarding family-communal land usage and local and familial political responsibilities (2004: 11, fn 1).                                                   52 García Linera et al. 2004; Spedding put the upper limit at around 80 (2004). 53 Note that during the coca boom, the region became home to a large population of labourers who did not own land of their own, known as partidarios and jornaleros (the latter denoting day labourers) (Spedding 2004). Though numbers are hard to ascertain, evidence from the period suggests that the majority left with the crash in prices in 1986. While a new problem of cato blancos—undeclared coca plots—has become more acute in the post-agreement era of peace (and by extension of cato limitation under the terms of the agreement finally negotiated between Morales and the other coca leaders and then Bolivian President Mesa), by and large non-landed cocaleros seem to have played little significant role in the region during the years of contention. 88  In keeping with the corporatist tradition in the country, each union upon founding affiliated itself with larger labour organizations, agglutinating first into larger centrals and, subsequently, federations (García Linera et al. 2004).54 I examine this process in greater detail in Chapter Four. In brief however, while the communities were locally self-organized, one of the notable legacies of Bolivia's social revolution was the adoption of a corporatist model of labour organization, such that the vast majority of the country's workers were incorporated into the umbrella labour structure of the country culminating in the Central Obrera Boliviana (Bolivian Central Union, COB). This organization extended even to country labourers, who were incorporated into a number of different larger groups, notably the Confederación Sindical de Colonizadores de Bolivia (The Union Confederation of Colonizers of Bolivia, CSCB) and the Confederación Sindical Única de Trabajadores Campesinos de Bolivia (Unified Union Confederation of Rural Workers of Bolivia, CSUTCB), who in turn sent representatives to the COB. Effectively, to quote García Linera et al. (2004: 43), “to be a citizen is to be a member of a union.”55 With the abrupt interruption of competitive democracy in 1964, that pattern of local self-reliance within union structures was only reinforced, despite state efforts to disrupt all unallied political organization in the country. With the electoral participatory channel effectively blocked, the wide variety of locally constituted institutions developed even greater importance—albeit of                                                 54 Interestingly, the process of evolution seems to have stagnated with the rise of Evo Morales and the MAS to power. Governance structures in the region have become quite stable. Indeed, outsiders and some residents (particularly non-members) spoke quite negatively about the level of control exercised by the MAS in recent years, suggesting open discourse and political debate had become less common and more risky within the federation structure following the end of contention with the state. That said, congress transcripts I obtained devote significant attention to questions of overlapping jurisdiction, communal ownership, and at certain points in time, affiliation with one central / federation or another. Each central and federation included a “Comision Tierra” which took on the task of adjudicate competing land claims, as well as lines of hierarchy running from union to central to federation. 55 Translations by the author unless otherwise noted. 89  an “apolitical” kind during the height of the military repression of oppositional political activity (Dunkerley 1984).  Among those who resisted, the COB attracted the greatest ire of the dictatorships, losing in the process a generation of leaders to exile, flight, imprisonment, disappearance, and death in a manner similar in style if not in scale to that seen in other Latin American dictatorships of the era in Chile, Argentina and Uruguay (Kohl et al. 2011, Dunkerley 1984).56 In contrast, the state's attitude towards the countryside was somewhat more nuanced. Building on an alliance established by the Movimiento Nacionalista Revolucionario (National Revolutionary Movement) that he replaced, Barrientos and his successors relied on a “military-campesino” pact to ensure stability in the country. In essence, the military provided a number of benefits to peasant leaders in exchange for support for the governing regime, particularly in its struggles against urban-based labour forces (Klein 2011, Dunkerley 1984). The dominant relationship thus was one of co-optation, leading to a relationship characterized by limited engagement verging on neglect, aside from certain general benefits such as subsidized grain gradually clawed back by the state during periods of economic stagnation, producing significant outcry and sporadic, disorganized protest in response that was quickly put down. It also provided in at least some cases more specific concessions to those holding local positions of power, though direct evidence of this is much harder to come by (Dunkerley 1984). The cost for such allegiance was high however, as the government suppressed any efforts at independent organization occurring at the margins. For instance, military leaders made significant efforts to limit or prevent campesino organization in the Chapare throughout the dictatorship period.                                                 56 The number of killed, tortured, imprisoned, and disappeared remains unknown, but estimates suggest something on the order of 40 murders, 100 disappearances, and 400 killed in total. [accessed 18 February 2015]. 90  As a consequence, producers determined to organize politically were forced to do so clandestinely, at considerable personal risk. One early leader, Agustín Arancibia, recounted his experience as follows in his diary: Colonization was conducted by the National Institute of Colonization of Bolivia (INCB). They did not allow organizations; the needs of the settlers forced us to organize clandestine unions. Knowing that I was organizing, the INCB rescinded my lot and pursued me. But my fellow colonists gave me another lot, no. 163. Not only that, I was able to organize a Central, and became Secretary General. The INCB monitored us more rigorously, and we made contacts with other departments of La Paz, Santa Cruz and Cochabamba. (198557; quoted in Salazar Ortuño et al. 2008: 23)  Other techniques of control included the deportation of families from the region, though that was more problematic for settlers who arrived with some degree of state support. The military government employed positive incentives for cooperation as well, notably in the form of favourable land grants to local leaders, who dominated many of the campesino organizations allowed to exist openly in the wake of the military government's early moves against organized labour in the country. The relationship constituted a cornerstone of the so-called military-campesino pact (Dunkerley 1984: 132-133).58 With this combination of positive and negative incentives, the military government strove to limit the independent organization of campesinos, and prevent the organization of potentially threatening alternatives. Nonetheless, the difficulties of controlling such an isolated population with which the state had relatively little direct interaction ultimately proved too much for the state to overcome, as the new colonists found                                                 57 Though the quote doesn’t make it clear, other passages in the diary clarify that he is referring to events that took place under Barrientos’ dictatorship in the late 1960s. 58 In subsequent years that relationship became increasingly strained, and in January 1974 the pact broke completely with the so-called Masacre del Vallee (Massacre in the Valley) when troops broke up a rural blockade campaign protesting the end of price controls, at the cost of between 80-200 dead. Even after the pact was effectively broken however, as I argue below, the fostered linkage between campesino leaders and the state continued to have a lasting effect on cocalero politics during the 1980s, as the federations with a history of government “tutelage” proved initially more conciliatory in their approach to state efforts to reduce coca production. 91  creative ways to coordinate their actions, as evidenced by the quotation from Arancibia above, and the organizations that subsequently emerged from such efforts.  By the 1980s, the cocaleros—for thus they were by then, having turned to coca as their most important source of income—had created a loosely affiliated network of small, localized Ostromian communities with vertical ties to different umbrella labour organizations, but with the state’s policy regarding coca production amounting to little more than rhetoric, they had not yet begun to define an active political project. The focus of those organizations grew steadily more political as state policy hardened, taking on the task of defending members’ rights to produce coca. These new objectives were implemented in addition to, rather than in place of the previous focus on internal social matters such as the provision of healthcare and education, road upkeep, and so on. Indeed, Central and Federation congresses continued to feature reports from a “social committee” discussing the state of health and education within the group’s territory alongside those devoted to politics, the “coca” issue, land and territory, and so on. As a senior coca leader, Carmelo Peñaranda, put it,  No, look: we did not invent the organizations. This comes from our ancestors, from our grandparents they come. And things were organized; we just strengthened them, a bit more incentive at the organic [i.e. membership] level; stronger. We’ve always had, from our Incan ancestry; we’ve always had organization. Don’t forget there were uprisings in La Paz by Tupak Katari. So it’s not new, there has always been organization, but back then for the organizations to speak of a political instrument was punishable, because the social organizations were condemned to serve the other parties, because the campesino was prohibited to speak of politics. For campesinos, their politics was the axe and machete, they could not speak of a structural change in the country, but eventually those social movements resisted and have gone on to take power, without there being any objection.  From that time onward, the goal and the purpose of the federations shifted, becoming inextricably bound in a national political agenda. Various respondents described this struggle in similarly compelling ways. 92  But what is the fundamental objective of the six federations? It’s to join forces, combine judgements around the executives of the federations, in order to definitively oppose not just the government, then of the right as the MNR and the ADN, because those governments were so submissive to the US government. In the Chapare, only the US government was seeking to eradicate coca crops, without offering an alternative to this eradication. (David Herrera)  Thanks to the internal organization and thanks to the union we created our political arm; from that arises the political arm. But the clear goal for us has been to seek liberation and self-determination of our peoples. (EI)  In sum, the cocaleros brought with them particular sets of values that made the emergence of collaborative community institutions seem like natural ways for the new arrivals to cope with their circumstances. The settlements that emerged in the Chapare region were well defined by their geography and relative isolation; their respective prior experiences equipped them with the ideational frame in which to construct their communities as inclusive, and indeed focused on the realization of communal goals, whether necessities of survival or political advocacy.  The diversity of settlements, and the openness of communal institutions, likewise distinguished communities of the Chapare from those found elsewhere in the country, as well as cases in other places in which armed conflict emerged. From the first, the communities included all arrivals that settled in the region. Accordingly, colonies came to be composed of members from numerous different ethnic backgrounds. Highland Aymaras settled alongside Quechuas both from Cochabamba and from other regions of the country. The latter constituted a majority, and in consequence, in the rural settlements continue to speak Quechua rather than either Aymara or Spanish. Interspersed between these two dominant groups were smaller numbers of mestizos from urban areas of the country, Guaranis from the Eastern lowlands, and a small number of the region’s original inhabitants, including the Yuracares and Trinitarios (Grisaffi 2010). Likewise, one encounters diverse religious backgrounds, including adherents to both 93  indigenous beliefs and practices and Catholicism; indeed, as is common elsewhere in the country many observe some mixture of both traditions.   The distinctions between these different identities receive very little attention either in scholarly literature or, so far as I was able to discern, in the communities themselves. On the contrary, many respondents went out of the way to stress the unity of Aymaras, Quechuas, and others in their political struggle. Comments from Julio Salazar, a senator for the MAS, are emblematic: “the coca leaf represents the cultural identity of the indigenous originario59 campesinos, who are Aymara and Quechua” (Interview, 2012). His phrasing is typical of MAS and federation representatives, strongly emphasizing the unity of the community, and bringing together multiple identity markers that previously divided (campesino yet members of a syndical federation; Aymara and Quechua together) under the figurative symbols of shared resistance, including the coca leaf. Accordingly, embedded values such as cooperation and the prioritization of communal goals were integral to settlers’ trust in reciprocity that was so vital the initial colonies. Likewise, the specific circumstances attending the formation of the Chapare colonies—notably their relatively recent formation under conditions of isolation and rough equality—is noteworthy, though given the single case design of the present study it is impossible to say whether those factors constitute causally significant, or even necessary prior conditions to the formation of an Ostromian community. The initial independence, the ongoing and increasingly institutionalized practices of interdependence, the communal principles embedded in the shared culture and values of the indigenous and mining communities from which the settlers came, all proved                                                 59 A difficult term to summarize succinctly, signifying some sense of primordial belonging. Grisaffi 2010, for instance, provides a number of examples of its use, but he too declines to define it despite its presence in the title of the article. 94  significant in the emergence of an OC in the region. Table 3.6 summarizes the implications of the observations with regard to the indicators I have identified for the three “foundational” OC constituent elements. Table 3.6: Evaluation of Foundational OC Constituent Elements  Indicators Evaluation  Institutional indicators    • Presence of a foundational constitution or covenant for the group, whether formal and written, or informal and oral, identifying physical space over which, and whether and how people may join or leave the community. The cocalero communities all featured a founding agreement among original settlers; all subsequent arrivals to the community were required to accept and adhere to the agreement and any subsequent amendments.   • Procedures to ensure that all members of the community are incorporated in community rights and obligations. Cocalero communities ensure that each household possessing land within the area of a given syndicate is both entitled and required to participate as a syndicate member, taking part in all regular meetings and decisions.   • Presence of a situation in which all members of the community would benefit from a certain action and through which costs are shared. Prior to conflict, provision of most public goods fell to the community. All benefited from the construction and upkeep of public spaces and resources, but the programs required the regular participation of all.  Behavioural indicators:  •  •  • Existence of a widely recognized (i.e. by both members and non-members) territory over which the community exists, with shared understandings regarding the boundaries of that community. Each syndicate maintained its own set of boundaries. Disputes among two or more syndicates were resolved through appeal to the relevant central(s). Disputes among centrals were further escalated. All respondents were clear regarding communities they belonged to. •  •  • Understood practices by which members can join and leave in practice. Membership in the syndicate begins and ends with acquisition of land. Syndicates reserve right to bar sale to anyone not deemed to be sufficiently supportive of community goals and practices. •  •  • Discussion of identity encompassing all eligible members of the community Emphasis on unity among all indigenous rural Bolivians in the region was constant both in institutional documents and in interviews. •  •  • Discussions of the problem in terms indicating that the participants understand the shared nature of the challenge, and the need for cooperation in order to meet it. From the moment of founding, colonies/ syndicates built shared responsibilities into the agreements governing their settlements. 95  3.5.2 Governance Constituent Elements Collective Deliberative and Decision-making Ability Unless a political line is the product of extensive discussion in our natural community organizations, it will not be valid or legitimate. (Constitution FEYCH, 2000)60  Evidence regarding the presence of this constituent element is triangulated from multiple sources. These include sets of records—ranging from agendas in some cases, to detailed summaries or even transcripts for larger meetings—for a number of political meetings for which I obtained documentation outlining the order of business for the regular (normally biennial) "congress" meetings at the central, federation, and coordinadora level at which executive officers are elected, policy positions and current leaders’ performances are reviewed, constitutions amended, intersyndical disputes resolved, and so on. I also obtained constitutions for several of the federations at various moments in time, which spell out in considerable detail the rights and responsibilities of members and meeting delegates, as well as the various senior officials.61 These written documents are reinforced and supplemented by participant accounts. In institutional terms, there are arrangements provided in each Federation’s constitution for active participation in decisions taken. All members have a right to be elected as local, central, and federation level officials, and as delegates to the regular meetings of each group. Delegates have explicit rights to voice and vote in meetings, and take part in the commissions that craft reports and motions for action on various topics. Likewise, component organizations are able to move motions at meetings of umbrella organizations. Significant evidence exists                                                 60 “Report of the 9th Ordinary Congress of the Federation of the Traditional Zones of the Yungas Chapare.” 2000. Cochabamba: Chapter 1, article 4, p. 35.  61 I have copies of three copies of federation constitutions, one centrale, and a statement of the charter for the Movimiento al Socialismo–Instrumento Político por la Soberanía de los Pueblos: Federación Especial de Zonas Tradicionales Yungas Chapare 2004; Federación Agropecueria Mamore 1999; Federación Tropical Cochambamba 2003). 96  regarding the scope of the deliberative process undergirding decision-making, and includes both substantive and legitimizing symbolic elements. In the course of my archival research I reviewed more than a dozen congressional reports—essentially digests of varying detail and quality summarizing the proceedings of the congresses of various Federations and Centrals—they are typical of such meetings, as all exhibit similarities in terms of rights and responsibilities exercised by meeting participants, executives, and so on.62 For instance, the constitution included in the 2000 FETCTC Congress report identifies the following rights and responsibility for all members (2000: 76-77): a) To accept the principles of the program of FETCTC; b) To have membership in one of the organizations [i.e. syndicates] either as member of the base or as an executive member; c) To implement and enforce this statute; d) To pay fixed monthly union fees and to know the state of accounts; e) To integrate within self-defence committees as needed or permanently depending on the case; f) Having obtained [sic] union dues, to be entitled to attendance in the union permanently, including all unionized members and their families should they, as consequence of their activities, be persecuted, imprisoned, confined, exiled or killed; g) Within the FETCTC all members enjoy the broadest freedom of expression, in accordance with the country’s constitution; [and] h) All leaders must become legitimate defenders of community interests. The offenders of this article can be subject to criminal proceedings in accordance with regulations.  The reports for all such conferences follow a similar formula, not unlike those seen in the larger national-level labour organizations such as the CSUTCB and the COB and indeed delegate-based organizational structures found elsewhere. They include initial speeches by dignitaries, attendance taking, reports from office-holders, and the election of a new executive. The reports also provide varying levels of detail regarding the formation and presentation of reports and                                                 62 A methodological note: these documents were relatively difficult to obtain. There is no central repository for them; they are only available through contacting the organizational secretaries charged with assembling them at the time. As such, the collection of several together constitutes a significant new research source. I had initially hoped to assemble something like a full set, but this proved impossible.  97  recommendations by standing committees covering topics ranging from organization finances and due collection, to meeting conduct, to the analysis of different dimensions of the contemporary political situation, with particular attention paid to both the coca situation in particular and land issues in general. Various committees also manage internal disputes in form of petitions from member organizations (such as the adjudication of land disputes and resolving lines of affiliation between syndicate, central, and federation), and review of leader behaviour. Those reports constitute much of the substantive content of the meetings, and are subsequently approved by the congress plenary. Thus, symbolically, each member of the six federations plays a role in the decisions being taken relevant to them, at whatever level, via the election of delegates to the regular meetings of centrals, federations, and the coordinadora. Substantively, those who wish to become politically involved can (and indeed must) do so by seeking positions within the federation. All the “bases”, the constituent union locals, send representatives—indeed, attendance is mandatory, with absences punishable by fines—with rights of voice and vote. The representatives in turn take part in the various commissions. Similar, if more modest, processes take place at more frequent intervals at the Central and union local level. Local unions can make representations to Centrals, and both can make representations to the Federation, the ultimate authority for the community. Indeed, as Spedding (2004: 299) describes, the federations take on functions of government at the community level.63                                                 63 This authority is broadly constituted, encompassing the totality of responsibilities the federations have arrogated to themselves in the absence of the state. 98  When the situation calls for it, meetings can be increased in frequency and intensity—particularly at the syndicate level. One former leader of the Centrales Unides federation describes it as follows: The meetings were happening every week to defend ourselves against [the government’s eradication program] Plan Dignidad.64 We had to go meetings each week to guide our colleagues. There were also meetings to organize the compañeras, the women, because when we the men arrived at the front to confront [security forces] they put a bullet in us directly.65 (EB)  With meetings of the federation congresses occurring only biennially, in practice, day-to-day management of federation business, including strategic planning takes place within the executive, with specific responsibilities assigned to each office. The list of elected offices is extensive at the level of Federation, ranging from the leading executive and general secretaries, to secretaries of housing, education, coca, civil defence, finance, the political instrument, even natural disasters, sports, and so on. For instance, the constitution outlined in the FETCTC Congress from 2000 provides for the election of 19 secretaries and three other members-at-large (2000: 67). The same document makes that arrangement official: “the executive secretary of the FETCTC is the highest authority of the Chapare peasant, and their decisions can only be reviewed by the assemblies and regional conferences of the FETCTC” (Article 14, p. 68).  The duties and powers of the executive secretary are also extensive, encompassing responsibilities for representing the federation, communicating on its behalf, enforcing and implementing decisions taken, and—in cases of emergency—taking any decision she or he (in practice it has always been a he) deems necessary on his or her authority, to be reviewed by next federation congress. At the same time, the eligibility criteria necessary to compete for any senior                                                 64 An aggressive and controversial plan of eradication introduced by President Banzer in 1997. 65 Accordingly, a common tactic the cocaleros employed was to ensure that the first row of marchers, blockaders, and protesters were women, as security forces were less willing to open fire on female demonstrators. 99  office ensures that only members with long histories of participation will ascend. This principle is borne out in the life histories of many respondents who spoke of assuming gradually more senior positions within the federation structure. Among the eligibility criteria for the FETCTC executive secretary, for instance, are the following:  To be a leader of the Federation (executive and general secretaries) requires: a. Be Bolivian by birth and be in full exercise of citizenship. b. Be a campesino, identifying with the sacred duties [and] interests of the campesino class, and have 10 years experience in tropical agricultural work and production of coca leaf, and not have been subjected to total reduction. c. Have an honourable history and have no outstanding charges or findings of guilt  d. Be nominated by the campesino bases and unions. e. To be executive secretary, one must have served as secretary general for two consecutive periods and three discontinuous. f. To be general secretary one must to have served as secretary general at a central for at least two consecutive periods and three discontinuous (Article 12, pp. 67-68).  In practice, this grants tremendous power to the executive leadership in implementation of goals and the day-to-day management of the cocalero community, and most particularly to the executive committee of the coordinadora established to coordinate cocalero federation actions.66 Through the use of mechanisms ranging from the regular collection of syndical dues including funds specifically for political organization, to a federation-controlled local radio station over which information could be disseminated and calls to action could be made—not to mention the hierarchical command structure through which orders could be placed from coordinadora through federation, central, and finally syndical executives—within the context of the general                                                 66 Indeed, evidence regarding Evo Morales’ time in office suggests that one of his chief preoccupations was ensuring that he coordinated decisions with the coordinadora. As one respondent put it, “From what I’ve seen in my meetings, I think they [the six federations, i.e. the coordinadora] outweighs the mayor and the unions. Here it’s the federations—there are six federations. Whatever the six federations decide, that goes for even my President. They are strong, the federations. They’ve studied people, and they take note of everything” (Anon 2). 100  mandate and goals established by the cocaleros federations at congress, the executive was able to act quickly and decisively as circumstances dictated. With respect to the binding nature, both the institutional rules and practical examples make clear that decisions taken at the meetings of the federations—and a clear hierarchy of authority exists within the organizations, with biennial federation congresses serving as the maximal authority within each federation, at which time executive elections occur and significant changes in federation policy can be made—take precedence over those reached at lower levels, ensuring a powerful level of cohesion in decision-making and implementation, with union locals bound to uphold operational decisions.  The same principle applies to changes in the organizational structure as well. For instance, article 56 of the FETCTC constitution (2000): specifies explicitly that “the proceedings of the [biennial] conference are verified through majority support of the majority of union locals; the decisions taken shall be mandatory for unions, and those that do not abide will be sanctioned (2000: 76). Other passages of the proceedings (e.g. resolutions passed and rejected on pp. 81-82) make clear that no central or syndicate can unilaterally change borders or affiliation without the approval of the appropriate higher authority. Though the delegative structure ensures robust deliberation accompanies any major decision—and particularly cases in which significant changes in policy or leadership are contemplated—it is worth noting that the quality of deliberation and democracy within the federation is limited in two respects: one ideational, and one instrumental/agential. The ideational limitation is perhaps not unlike that encountered elsewhere, though the mechanisms are more explicit and the effects far more obvious. Specifically, participation in the union at any level involves heightened exposure to the process of socialization discussed elsewhere in this 101  dissertation, referred to by participants as training, strengthening or re-enforcement, or education. Though no one within the movement described the process in such terms, two long-time non-member residents of the region with considerable exposure to federation politics did. Union representatives receive greater exposure than typical cocaleros through participation in more regular meetings, particularly given that most meetings include some form of education, testimony, or other message with the function of reinforcing the shared commitment to struggle. Common themes include descriptions of past suffering, the political history of the cocalero movement, accounts of crimes perpetrated against the cocalero community by then-present governments, celebration of the actions taken to resist state action by cocaleros, most notably including the great marches on the capital of La Paz, and so on.  To cite a typical example, the XVII Biennial Congress of the FETCTC in 2000 (2000: 91-94), at the height of the most intense period of confrontation, concluded a testimony from a compañero, Carmelo Rojas, listed under the heading “Related History of the Federation of the Tropics of Cochabamba.” In it, he recounts his family’s arrival in the region, the difficulty in founding the first Federation in the region (which he identifies as taking place in 1956), the wresting of land from larger landholders in the 1950s as part of the relatively chaotic land reform (ushered in by the MNR in the year following the social revolution). He goes on to describe the onset of state-perpetrated violence in the region following the appearance of cocaine. He cites some specific examples, including a cocalero doused with gasoline and burned alive. He recalls being persecuted, detained, transported to the capital for questioning, and various details of specific confrontations between cocaleros and the military forces. He concludes with a 102  dedication to those who fought previously, those who “died in defence of the coca,” and with an exhortation to continue the struggle, organize better, and remain unified in order to win. Similarly, the 2004 meeting of the Federación de Zonas Tradicionales Yungas Chapare (FZTYC) concludes with a review of the movement’s history. Such passages are common to meetings. More frequent, smaller union seminarios (seminars) at the syndicate level often focused on the goals and strategies of the cocalero movement and obstacles facing it. Some of the meetings were and are relatively technical, covering aspects of various laws proposed, and how they affect cocalero interests for good or ill. Even more significantly, from the standpoint of organizational continuity, is the mandated ability and duty of current office holders to vet future candidates for office on the basis of what for lack of a better term might be described as ideological purity. To cite an example, the constitution of the Federation Sindical Agropecuaria Mamore (Mamore Agricultural Union Federation, or FAMA) specifies that a member must meet three conditions before standing for office (1999: 8): 1) be an active member, with demonstrated experience, 2) be at least 17 years old (orphans excepted), and 3) appear before any secretary of the union with transparency, and without a history contrary to revolutionary syndicalism.  In short, candidates were pre-approved by previously elected officials explicitly charged with ensuring their compatibility with the previously agreed mission of the federations. This helps to account for the long slow progression so common in the different life histories I encountered. With similar practices in place throughout the organization, the result was a high level of continuity within the organization the most intense contention, from 1988-2004.67                                                 67 Each constitution also spells out the final objectives of the federation, which in theory could be changed by vote at a biennial congress, but in practice remained stable throughout the course of conflict with the state. As with the cocalero movement as a whole, each federation’s motto is “¡Kausachun coca! ¡Wañuchun yanquis! (A Quechuan phrase translating roughly as “Long live coca! Death to the Americans!”). The goals for each include pursuit of 103  Indicators Evaluation   Institutional indicators: •  •   • Existence of a mechanism by which the opinions regarding different courses of action may be aggregated from all members, and by which different alternatives may be weighed before a binding decision is taken (e.g. through some sort of representational institution); institution may be formally or informal in nature. Delegative system provides for representation of groups at successive levels of discussion; same mechanism tends to legitimize decisions taken.    Behavioural indicators: •  •   • Existence of durable shared understandings regarding how decisions are taken, what the rights and responsibilities of members are with regard to the processes of deliberation, decision, and implementation All leaders I spoke with were clear regarding how the group reached decisions, and how they would be implemented. Many repeatedly stressed the importance of unified support behind decisions, once taken. Many emphasized the importance of participation by “the base”, i.e. rank and file members. •  •   • Presence of some central decision-making body or bodies enjoying authority over the membership; larger, more complex communities may distinguish between different elements of decision; some may separate out “executive” and “legislative” functions, for instance. Clear lines of authority running from the biannual congresses and more frequent meetings at lower levels to the executives at each level charged with enacting policy within the community. Table 3.7: Evaluation of Governance OC Constituent Elements (Deliberative Decisions) The result of this is a fascinating organizational structure, one that simultaneously allows for airing of diverse opinions among members and their representatives at various levels of organization, yet where the overarching goals and strategic direction of the organization has                                                                                                                                                        sustainable development, unity among the Centrales and their member unions and communities, secure guarantees regarding land and territory, providing of the legal rights of each subgroup without risking the common good, promotion of trained cadres of leaders, and the mobilization of all members in response to external influences (Estatutos de la FAMA (Mamore) Chapter II, Article 5, p. 5). 104  remained remarkably stable through more than thirty years of dramatic political contention. That process of policymaking within each Federation has this dual quality of inclusive deliberation and subsequently binding decision-making. Table 3.7 summarizes these results. The following section outlines how the community ensured that those decisions, once taken, were enforced. Enforcement and Oversight The fundamental principle of militant MAS - IPSP is that no member may pursue personal interests in politics, in trade unions or community. Our militant [movement] must serve the political, union and community. (Constitution of the FEYCH).68  The enforcement element actually operated at two levels (mass and elite), and via multiple mechanisms for each. I explore each in turn. Regarding elites, the federation maintains a set of controls through their standing economics committees, supplemented variously by Comisiones de Vigilancia (vigilance committees) and Comisiones de Evaluación (evaluation committees) (e.g. FETCTC 2000), which specifically scrutinize the expenses and behaviour of officials elected to public office, effectively functioning as an internal audit committee. Each federation, and all the centrals I was able to review constitutions for, included articles specifying the actions for which leaders (and members) might be censured, and mechanisms by which such punishments might be effected. Infractions range from poor language in meetings, to unjustified absences, to behaviour reflecting poorly on the federation, to betrayal of its goals. Typically, punishments ranged from fines, to reprimands, to suspension, to expulsion. Congress reports from the FETCTC, Mamore, and Chimore all include variations of this. The level of scrutiny exercised moreover is quite fine-grained at times. The FETCTC Evaluation committee for instance noted the maltreatment of the municipal car. A women’s                                                 68 “Report of the 9th Ordinary Congress of the Federation of the Traditional Zones of the Yungas Chapare.” 2000. Cochabamba: Chapter 1, article 4, p. 36. 105  federation report cautioned against drunkenness. Other declarations were more symbolic, as with the denunciation of a highland Aymara leader, Felipe Quispe, who the cocaleros believed at the time had betrayed the larger campesino movement in the country. It is perhaps no surprise that a pervasive fear for leaders such as Morales was of being seen as a sell-out (Sivak 2010). The social value placed on solidarity in this regard, was extremely strong, a factor discussed further in Chapter Four’s treatment of socialization within the cocalero community. At the mass level, methods of policy enforcement range from conflict mediation performed by elected officials within the union local to more formal procedures of sanctioning and censure. Evo Morales likes to tell how he received complaints from lawyers, who claimed he was robbing them of business by saving couples from divorce (Sivak 2010). A number of the federation officials I spoke with provided accounts similar to that of Evo's, reporting that a central role of the "secretario de conflicto" or union corregidor (roughly a magistrate) or some similar official would be to mediate conflicts within the syndicate and adjudicate disputes and accusations of wrongdoing within the community (Asterio Romero interview; also Spedding 2004).  One interview participant described a method of punishment reminiscent of medieval rough music as a “message” sent to a non-member, intimidating a former hotelier in the region.  I run a high risk, I mean, I've had problems to the limit. I had problems because they wanted to fine me, I had problems because I do not want to join [the union] and have an obligation. This hotel has had problems in the past because the previous owner sued Evo Morales and won—for the 30 day strike she sued for damages because they hurt [business], and she won. The union was then very angry at her. Wanted her to go, bothered her. They threw dynamite down here, coca leaf waste appeared at night everywhere; where cars were parked they threw rocks and broke the windshield, I mean, it was very difficult for her. Eventually she sold it. (Anon 3, an urban resident of the region)  106  Cocaleros directed one of the most public of such displays at then President Hugo Banzer. In December 2000, the president arrived in the region in the company of the U.S. Ambassador to declare victory in his eradication strategy, and to ceremonially destroy the "last leaf" of coca. 15,000 campesinos led by Evo Morales protested nearby and covered an 800m stretch of highway in coca leaves in a clear repudiation of the President's message. (The episode caught international attention, with reports appearing in The Economist, for instance, some weeks later). Some sanctions took the form of some type of corporal punishment, often public, and extended even to the regulation of domestic life. EP, a former leader of the Women’s Centrales Unides Federation, recalled  There are roles for compañeros, yes. Today, we devalue it a little, but before a married woman or a married man, if they made an error, geez ... I remember how I heard and saw a man in my Central, they covered him with water and sand and washed him. The women did the same to her. That’s how I recall it. All women would sanction the woman, and all men would sanction the man; it was done equally in the syndicate. (EP)    In one particularly notorious form of communal sanction, suspected transgressors—often accused of robbery—would be tied to a “palo santo” tree, inhabited by poisonous ants.69 At the extreme, as in other parts of the country and indeed the continent, suspected criminals have been reported killed under the rubric of “community justice.” The illegality of such incidents, combined with the relative remoteness of the Chapare, makes it very difficult to ascertain just how widespread the practice was in the period under consideration, though increased availability in the years since 2005 suggest it occurs in the Chapare with some regularity.70,71                                                 69 One such incident was picked up by the Associated Press, and thus widely reported in 2014. The families of the two involved denounced the attack as constituting torture. ( 70One reported media survey identified 180 such deaths reported across Bolivia during the period 2005-2013, including 14 in Cochabamba in 2013 alone. Firm numbers remain elusive however; the same news article reports that the government ombuds office (Defensor del Pueblo) received reports of only 53 cases over the same period. 107  In the case of the more formal sanctioning, punishments for members were similar to those listed above for officials, including expulsion from the Federation. Some constitutions include specific procedures for appeal of decisions as well. While in practice there were relatively few instances of sanctions applied, they did happen. For instance, the following account comes from the land and territory committee proceeding the 20th congress of the FETCTC, in 2009: “For failing to act in accordance and for threatening with a firearm six compañeros on watch in the community the Cocas, the compañero Gregorio Ortega and his people remain expelled from the Tropic of Cochabamba, and we recommend the new executives enforce this” (2009: 38). There were and are other, more mundane ways in which the federations can ensure compliance. The most common are fines for small transgressions, including in particular absences from meetings. In keeping with the labour antecedents of the movement, reinforced by the realities of the circumstance, solidarity constituted the basis of their power as political organizations. Accordingly, regular and reliable attendance by members is a crucial prerequisite, with a combination of fines and social pressure encouraging any producers who might be inclined to free ride not to do so. It is essentially the model of contingent compliance applied to the problem of mass political mobilization, one that proved highly effective at resolving that particular collective action problem when combined with the significant process of socialization and enforcement of communal justice. Written records, testimony, and personal observation alike                                                                                                                                               ; accessed 2015-02-20. See also Goldstein (2004). 71 The incidents for which we do have some record from the period tend to involve outsiders, as in the case of David Andrade López, a police officer killed with his wife and four other members of various government agencies in October 2000. All five bodies recovered bore significant marks of trauma. In discussing the incident, the then Defensor del Pueblo referred to widespread rumours of brutality on the part of Andrade prior to his death (Interview Godofredo Reineke, former Defensor del Pueblo; Ledebur  2002: 14; Salazar Ortuño et al. 2008: 526). 108    Indicators Evaluation   Institutional indicators: •  •   • Presence of institutional procedures providing some form of oversight and/or accountability of group leaders with regard to the membership. Each level of organization and leadership provides for the review of leadership performance in the form of reports to congresses, as well as specific oversight committees to review economic performance following the election of MAS officials to local and national office, as described above. Specific duties are spelled out in a number of constitutional documents.  •  •   • Presence of a mechanism for the application of graduated sanctioning in the event of shirking on the part of membership with regard to the shared responsibilities of group members. May take the form of peer-to-peer sanctioning (as in the case of some common pool resource institutions discussed by Ostrom 1990) or institutionally implemented penalties. Constitutions of federations and centrals provide for graduated sanctions to both leaders and members who violate the parameters of their membership or office, ranging from small fines to expulsion from office, and/or the community.  More informally, varieties of rough music are used to ensure compliance; these can be extended to individuals outside the community. •  •   • Presence of formal or informal processes for community members to consult or employ in the resolution of problems without the involvement of external actors. Various processes for the resolution of intercommunal conflicts, via the federation structure. These ranged from formal group appeals heard and adjudicated at congresses, to informal interventions in communal life by local union officials, notably the secretario de conflict. Table 3.8: Evaluation of Governance OC Constituent Elements (Enforcement and Oversight) 109    Behavioural indicators: •  •   • Regular meetings of group responsible for accountability and oversight (unless invested in an individual).  In each congress document reviewed, the meeting and pronouncements of some audit committee or committees (names, precise focuses vary) recorded. Most extensive example stems from the records of the largest federation, the FETCTC, which scrutinized not only its own leaders, but all officials elected by the MAS as well. •  •   • Evidence of an assessment process. The minutes of various meetings identify monitoring processes for leaders and members; likewise testimony from respondents discuss accountability to “the base” on the part of leaders, as well as informal enforcement of rules on members. •  •   • Evidence of sanctioning of leaders. Various levels, including “vigilance” and “evaluation” committees, as well as through reports to the full congress from key leaders, and evaluation of specific actions at relevant committees; several instances of sanctioning observed in the record, including in one case for “nepotism” (XVII Congreso del Tropico de Cochabamaba 2002). •  •   • Evidence of peer-to-peer monitoring processes identifying and successfully sanctioning shirkers. Difficult to obtain directly. The presence of rough music undertaken without reference to formal processes indicates there was such monitoring and enforcement undertaken. First-hand travelling experience suggested an extremely high level of communal vigilance in the region, with members of the community quick to act in coordinated fashion whenever they observe something as a threat. See for example Prest (2014). •  •   • Evidence of institutional monitoring processes identifying and successfully sanctioning shirkers. Reports of expulsion; many references to levees occurring at various levels. •  •   • Evidence that actors in fact use the mechanisms to resolve those disputes, either willingly or at the behest of other community members. Anecdotal evidence from several respondents regarding the mediating role played by local union officials, including domestic disputes. Table 3.8: Continued 110  reinforce how seriously federation representatives took—and continue to take—the issue of attendance. Each event includes an explicit roll call, with the number of participants from each sub-organization recorded; absences are noted, and fines applied.72 The position of undisputed authority within their particular geographical ambits provide the federations and member organizations further mechanisms through which to maintain compliance among members. For instance, the federations themselves exercise authority over the registration of farming properties and the resolution of boundary disputes. As stated above, federation officials regulate land sales as well. Likewise, the relative monopoly over political activity in the region ensures that any hoping to become politically engaged will do so through the auspices of the federation, and in a manner therefore consistent with its broader mission.  Table 3.8 summarizes these findings, highlighting the presence of oversight and sanctioning mechanisms observed for both leaders and members, along with the available resources for conflict resolution. 3.5.3 Relational Constituent Elements While the cocaleros maintained a federal structure, the lines of hierarchy shifted over time. In the eras prior to contestation the most important lines of affiliation ran from local unions and centrals to departmental and national-level umbrella groups such as the COB and later the CSUTCB. As tensions heightened during the 1980s with the implementation of increasingly stringent coca control strategies, the importance of vertical lines lessened in favour of regional affiliations. The results can be seen over the course of the 1980s, as cocaleros looked less and                                                 72 Even today, with the movement stronger than ever before, indeed, in firm possession of the levers of government, discipline remains sharply enforced, something I observed firsthand during fieldwork. Attendance is carefully taken at each official event, absences are noted when a given Central or union local fails to send the requisite number of representatives. At other occasions, members receive chits following attendance at a seminario (training seminar) or following a demonstration. 111  less to the COB and other national organizations to speak on their behalf in negotiations with government and representations to the public. The miner-backed COB’s significant decline following the closure of the country’s mines only hastened the process.  That is how I came to know the mine. In my time the miners were fine, they had a goal, had good schools for their children, they enjoyed writing materials, notebooks, books. After García Meza, in ’82, they planned a reorganization because “the mine is bad, because the mine is failing, because it’s bankrupt” — a mountain of pretexts. With that pretext they tried to relocate the miners, and yes the miners lowered their pants. They were offered $5000 or $2000 per year worked. If you have worked for 10 years, and it’s $2,000 per year, how much do you have? $20,0000. And so the miners said, “All set! I’ll go to Cochabamba or La Paz and buy a house.” (EB)   It was a perception shared far beyond the cocaleros themselves. A mine union leader, Simón Ramírez, remembered it in these terms: It was like a boat going down, it was everybody fighting to survive. They wanted to destroy us and they have achieved it. What they have not been able to do in many years with weapons in their hands, they managed to do simply with a decree, decree 21060. (Interview quoted in Jeppesen 2014)   In consequence, cocaleros increasingly relied on their own officials to press the case to government. At the same time the cocaleros became increasingly self-reliant in their representations to other communities. They began reaching out to other actors as equals, and subsequently as leaders among opposition groups in the broader indigenous and rural campesino movements. As I discuss in Chapter Four, they played a central organizing role in the establishment of the MAS political movement/party.  The process of evolution in cross-community linkages was ongoing. The efforts recorded in Arancibia’s diary for instance, referenced above, make clear that linkages outside the region were a priority even at a time when political opposition was thoroughly suppressed and atomized, with leaders killed, exiled, or imprisoned. As early as 1971, leaders in the Carrasco and Chimoré had created the Confederación Sindical de Colonizadores de Bolivia (Syndical 112  Federation of Bolivian Colonists, CSCB), devoted to the representation of colonist-specific interests in the face of limited state support.  In subsequent years, particularly in the latter part of the 1980s, the cocaleros proved adept at placing their own issues at the forefront of national labour organizations’ agendas, most notably coca (Healy 1991: 93; Salazar Ortuño et al. 2008: 132). By 1987, COB had adopted coca legalization as a core concern for both its economic and political committees. This process of connecting is one that a number of interviewees mentioned. For instance,  Once there were six federations, compañero Evo made contact with the manufacturing sectors, mining sectors, and the highland cooperatives. We undertook exchanges: some came to the tropics of Cochabamba, while Evo and others went to the Altiplano, [and the] mining and manufacturing sectors (EB).  Sivak (2010) emphasizes the importance of that networking process for Evo Morales as well, both domestically but internationally.  At the same time, some experienced organizers outside the Chapare saw the considerable potential that the cocalero movement offered to the broader Bolivian opposition movement, and sought out ways to encourage greater activism and engagement on the part of what Stefanoni and do Alto characterize as a weak syndical structure in the region prior to that time (2006: 40-41). For instance, Filemón Escóbar, a long time miner and union leader who played a crucial role in the development of cocalero politics in the late 1980s through the 1990s, both as an advisor to Evo Morales and as an active educator of cocalero federation membership regarding the possibility of an electoral option. I discuss both aspects of his involvement below.  A nationally prominent labour organizer within the COB since the 1950s, Escóbar’s own involvement in Bolivian politics predates even the settlement of the Chapare, let alone its emergence first as a locus of conflict and subsequently as the heart of the country’s ruling MAS 113  party.73 In the mid 1980s Escóbar started to work with campesino unions through his work as a cultural advisor for the COB. Increasingly, as other indigenous political initiatives in the country declined, he focused his energy increasingly on the Chapare. By 1992, cocaleros were the foremost proponents of the idea of formal participation in government via a political party organized by and for indigenous Bolivians. Accordingly, they spoke in defence of the “thesis of the political instrument” advanced at the first and only Assembly of Indigenous Peoples organized by the CSUTCB, generally regarded as the foundational movement of what would eventually become the MAS-IPSP. While it would be impossible without exhaustive interviews of participants—and perhaps not even then—to specify exactly the extent of his influence, it is clear that Escóbar played a crucial role in the cocaleros’ adoption of a democratic political course. Accordingly, it is a topic I return to in the following chapter. Independence from the State One lingering question that recurs throughout the thesis is, why did the state not simply pass legislation declaring the federations illegal and/or employ military measures to eliminate the organization? As the above implies, the cocaleros were able to maintain considerable independence throughout this period. This continued “survivability”, what might be described as the grudging acceptance of the cocalero movement’s existence and importance as interlocutor, was the result a confluence of institutional/constitutional, political, cultural, and power-based factors. Constitutionally, the revolution of 1952 ensured labour federations a legal place within Bolivian political life; in the years that followed, virtually all sectors of the formal economy, as well as the countryside, were organized into a diverse array of union structures. During the                                                 73 He was, for instance, judged sufficiently notorious as a COB leader and political activist to warrant internal exile during the country’s military dictatorships (Escóbar 2008).  114  dictatorship era, the presence of such structures proved useful to the state in maintaining a degree of engagement with and—more importantly—control over different segments of the population by coopting or containing that union structure. With the return to democratic governance in 1982 and the respect for constitutional law that it entailed, the state had a legal obligation to respect the existence of such federations, or risk losing its own fragile legitimacy.  Bolivian political norms further buttressed the federations’ independence; corporatist unions became part of the accustomed patterns of Bolivian politics as in other countries in Latin America. They thus acquired the legitimacy that comes with habit widespread and over time. To attack the existence of the federations would have been to undermine a distinct element of Bolivian political cultural found throughout the country. At the same time, one must consider the limitations of the state’s capacity to undertake such delegitimization. In effect, the state did make efforts to disrupt cocalero unions during the 1990s and early 2000s, but proved unsuccessful in their attempts. Various former cocalero leaders I spoke with testified to the vulnerability they felt—several discussed fearing for their own lives, as well as their families’—but also to the measures they were able to take to preserve their safety, including flight to the jungle. RM tells such a story, recounting the measures to maintain organization and avoid reprisal in areas under military occupation. The forced reduction was tremendous; almost the entire territory of Ivirgarzama [a town in Carrasco province] was militarized. The dead we buried in groups, but it was forbidden to have meetings and gatherings. It was prohibited because there was no freedom; we had to meet in the hills, next to Senda 6 [a small road outside the town]. Other central also did so because during the day the people did not go out. By night only we organized ourselves using passwords to know where we were going to meet nocturnally. It was not easy to defend coca in order to legalize it little by little. The fight cost enough. That was the story.   115  Along those same lines a former security agent I interviewed spoke of the fear that soldiers had during postings in the region as well. While able to travel with some security during the day, at night, soldiers on patrol and thus outside of barracks were vulnerable to attack when they wandered too near the edge of the military encampment, exposing themselves to bombardment from near cocaleros in the trees or brush, and—cocalero protests to the contrary—on a limited number of occasions to rifle fire as well (Ledebur 2002: 14).74  In effect, Bolivian security forces buttressed by US funding and personnel were capable of occupying the urban areas of the region and significantly restricting the mobility of cocaleros by day, but they had little success in undermining the basis of cocalero organization in the region even at the height of repressive efforts.75 In sum, thanks to a combination of de jure, de facto, and normative ideational factors, the cocalero federations were able to persist and indeed grow in power over time. They did so through internal mobilization and organization, and through processes of alliance-building, both in the broader Bolivian opposition movement and even abroad. By the mid 1990s, the cocalero federations assumed and subsequently solidified a position of leadership. I discuss this process in further detail in the following chapter. Table 3.9 summarizes the findings of the preceding section with regard to relevant indicators.                                                   74 A quotation from Lazar (2006: 195) highlights the sensitivity Bolivian opposition groups exhibited to charges that they engaged in sniper attacks on state security forces. An opposition organizer speaking of the period prior to the resignation and flight of President Sanchez de Lozada in 2003 said, “well, the days went by and the government went on promoting itself more (se va potenciando mas), with haughtiness, arrogance – he called us snipers (francotiradores), ‘snipers linked to drug traders’ (‘franconarcovinculos’), sniper-I-don’t-know-what (‘franco no se  que’), all that.’” As I discuss in Chapter Four cocaleros worked hard to rebut any suggestion that they engaged in the types of violence that would justify military repression in response, and one of the most common charges was of sniper attack. 75 It is, indeed, somewhat reminiscent of Billy Joel’s anthem Goodnight Saigon (confirmed, I hasten to add, by more scholarly treatments such as McFate and Jackson 2006).  116  Constituent element 8: Degree of recognition by and independence from the state Indicators Evaluation   Institutional indicators:    • Presence of a formal relationship between the OC institution(s) and state level institutions.); it may be active or passive (i.e. marked by positive recognition from the state, or rather a function more of benign neglect or reluctant tolerance). Cocaleros enjoyed legal protections under the country’s 1952 constitution, which afforded all Bolivians the statutory right to organize in labour groups.   Behavioural indicators:    • Evidence that the state either does not challenge the right or reality of such groups’ existence, or is unable to mount such a challenge successfully. Cocaleros were able to work around state efforts to repress their organizational structures through a variety of strategies. •  •   • Evidence that the state recognizes OC institution(s) as an interlocutor, engaging with them in various ways such as negotiations, agreements etc. Negotiations between the state and cocaleros were constant throughout the period of contestation. (Salazar Ortuño et al. 2008)  Constituent element 9: In the context of large communities a federated structure, in effect creating a network of hierarchically organized sub-communities.    Institutional indicators: •  •   • Existence of multiple community sites, with hierarchically constituted linkages between them that enable shared decision-making With the emergence of the coordinadora in the late 1980s, formalized in 1991, the federations achieved a region-wide integrated organizational structure. •  •   • Mechanisms to enforce compliance with agreed-upon policy between sites. Each organization possessed committees and executive officials tasked with ensuring various sites met obligations among constituent members.    Behavioural indicators:    • Evidence that sub-community sites do work together and maintain a common commitment to mutually decided policy.  Evidence of regular meetings at each hierarchical level to decide upon, explain, and enforce policy.    • Evidence that sub-community sites monitor one another, and sanction free riding and other kinds of defection from agreed upon policy. Widespread evidence of monitoring in congress records; some examples of sanctioning. Table 3.9: Evaluation of Relational OC Constituent Elements 117  3.5.4 Conclusion In this chapter I have presented evidence regarding the historical emergence and present institutional structure of the Six Federations of the Tropics of Cochabamba. In doing so, I have established that the federations collectively developed the properties of an Ostromian community; as such, they were and are able to articulate and execute a collective strategy of resistance when confronted with a determined government campaign of coca eradication. Specifically, I have argued that the federations constituted locally embedded institutional solutions to a set of collective action problems that existed in the region long before the onset of contention with the state over the production of coca.  The specific elements include an communities that were well bounded in which membership was open to all who settled in the region, with membership in the new settlements quickly becoming important components of inhabitants identity that provided a solid basis for cooperation with others in the community regardless of their background. The communities also possess institutionalized mechanisms of deliberation and decision-making that encouraged end indeed enforced participation both in the decision-making process itself, and in the subsequent implementation of decisions taken; these include monitoring and enforcement mechanisms acting for both leaders and the mass of cocalero membership. In the following chapter, I turn to the question of how this structure influenced the arc of contention observed in the region throughout the era of attempted coca reduction and eradication.   118  Chapter 4: Defending the Chapare—OC in Contention  4.1 Introduction 15 November 2001 is remembered as the Massacre of Senda VI in the Chapare. On that day, three campesinos were shot and killed in confrontations with state forces, as the cocalero federations continued a campaign of “lighting blockades” of the region’s highway to protest both the ongoing government campaign of forced coca eradication and the heavy military presence in the region in support of it. Another five were wounded by gunfire, and twenty or so more assaulted by other means at eight different sites during a crackdown on a region-wide cocalero protest. Strange to say, in some ways the name “massacre” is more grandiose than the events warrant, as the day in many ways resembled many others observed in the region throughout the period of contestation, differing more in intensity and breadth than in form. The deaths came on day fifteen of a campaign of flash blockades maintained by local committees associated with different syndicates. It was not the first time deaths occurred during a blockade in the region, and would not be the last. Again and again, cocaleros would dart out of the surrounding jungle and clutter the roads with logs, rocks, and anything else close to hand, rendering a stretch of highway impassable before disappearing back into the brush. It was a systematic effort, involving thousands of producers across the region coordinated through the nested federation structure. Local leaders organized committees to maintain blockades in their region, with the larger campaign coordinated at the level of central, federation, and ultimately the coca federations’ coordinadora. 119  Such casualty figures would not look out of place had they come from any low-intensity armed conflict around the world, and yet they did not emerge from one, or trigger an escalation. On the contrary, violence remained relatively one-sided even in the wake of the Senda VI deaths. By December, the blockades, combined with rising pressure of public opinion that emerged in opposition to the state’s heavy-handed approach, led to a “coca summit” in which coca and government representatives met and exchanged viewpoints. Though the blockades and ensuing summit ultimately did not resolve the issue, it did produce a temporary halt in eradications and a reduction in state-perpetrated violence. It also demonstrated once again the legitimacy of cocalero leaders as interlocutors capable and deserving of face-to-face negotiations with the government, and illustrated the resolve of cocaleros and the viability of their strategies of resistance. Cocaleros accordingly remained committed to their aggressive yet limited action at sites of eradication, large acts of protest and blockade both in the region and in the cities of Cochabamba and La Paz, and through formal electoral competition. In the pages that follow I provide empirical evidence, primarily in the form of interviews with participants in the conflict, clarifying how the form and extent of political organization in the region decisively shaped the emergence and form of that resistance. The presence of an inclusive and region-wide set of institutions through which the cocaleros represented themselves both enabled and encouraged cocaleros. It likewise both enabled and carefully delimited the deployment of more violent forms of resistance specifically in the defence of coca plantations against state incursions to carry out eradication.  Whereas Chapter Three described how the cocalero federations constituted Ostromian communities, the present chapter explores how that organizing structure shaped the course of contention in the region as the state took a harder line against coca production in response to 120  increasing U.S. pressure during the 1980s and particularly the 1990s. Accordingly, in this chapter I first briefly outline the course of the conflict before delving into the ways in which cocaleros’ organization shaped specific processes of contention. I do so in four sections. Following this introduction, I provide a brief history of Bolivian politics, with particular attention paid to coca, its economics and role in Bolivian (and more specifically Chaparean) society, along with an overview of the contention itself, noting the overarching evolution of state policy throughout, and how cocaleros responded. The heart of the chapter then consists of a series of sections in which I identify and analyze selected processes illustrating how the cocaleros’ organizational structure shaped the form and content of cocalero resistance to increasing state attempts to reduce coca production.  • First I examine the gradual splintering of cocalero federations, followed by their relatively sudden and effective reunification when confronted with the state’s growing opposition to coca production, and the perceived threat to both their livelihoods and way of life that they represented. • Second, I examine the cocalero response to state attempts to introduce alternative development schemes as a way to induce voluntary eradication on the part of cocaleros, and how the OC structure accommodated the resulting divergence in producer interests—a process depending greatly on the inclusive conception of community that informed the cocalero federations. • Third, I examine the process of consensus building that occurred in the region around particular strategies, focusing in particular on a process of socialization that included both agential and institutional dimensions. • Fourth, I consider how the cocaleros mobilized resources for contention in practice, and in particular how the presence of the OC structure tended to at once distribute the burden widely, and ensure broad compliance with the shared responsibilities. • Fifth, I consider the organization and deployment of the self-defence committees (SDCs) as a specific example of both the capacity for “social rationality” exhibited by the cocaleros, and the ability of the community to rigorously enforce communal decisions once taken. • Finally, I consider how the cocaleros’ effectively contested the state’s own presentation of the terms of conflict, as each attempted to frame the nature and scope of the context, successfully resisting attempts to be labelled as threats to the state, reducing in the process the likelihood of violent state repression in response.  121  Finally, a concluding section summarizes the implications of the evidence presented for my primary hypothesis and the alternatives. To review briefly, the primary hypothesis developed here is that the OC structure present in the Chapare region prior to the onset of contention generated distinctive patterns of resistance, including high levels of coordinated contentious political activity, but stopping short of armed resistance. Such an outcome occurs, I argue, as a function of the cocaleros’ OC and its distinctive properties, both constituent elements I identified in Chapter Three, as well as particular emergent properties such as social rationality in decision-making, as well as the substantial resource mobilization capacity that OCs exhibit as a result. Here in Chapter Four, I argue that OC leadership was able to engage in rational decision-making in response a threat to a core cultural symbol and economic resource that was grounded in a reflection of a broad understanding of communal interests. Thanks to the deliberative delegative structure of the cocalero community, strategic decisions were informed by considerations of the net costs of various alternatives, including violence, to the community as a whole; once taken, decisions were well supported by the legitimacy that came from that decision-making process and the efficient communal mechanisms of mobilization made possible by the cocalero federation infrastructure. Indeed the presence of that institutional framework widened the cocaleros’ “repertoire of resistance” considerably, enabling strategies that dependent on sustained community-wide mobilization difficult to generate under other governance structures. In that context, the cocaleros opted to pursue change in the government’s coca policy through a three-pronged strategy of formal political organization, large-scale unconventional yet non-violent political action, and the carefully delineated deployment of non-lethal violence in defence of local territory. Those strategies, I 122  argue, reflected a clear understanding of the possibilities for, and limits on, resistance in the context of an imperfectly democratic state. Beyond the immediate instrumental rationality of the decisions at the moment they were taken, cocaleros at both senior leadership and grassroots levels consistently took actions that ratified and reinforced those strategies; indeed, so much so that support for the “political instrument” became a significant constitutive element of the cocalero identity, one reinforced through both cocalero discourse and organizational practices. The cocalero identity that emerged during and to some extent as a result of the decades of contention differed in important ways from previous efforts to politically mobilize indigenous and rural Bolivians. I argue that those novel shared conceptions of cocalero identity, entrenched as they were within institutional rules and norms by both discourse and practice, tended to prescribe some forms of political activity, including participation in formal political competition, nonconventional resistance and local self-defence committees, even as it proscribed others, most notably violent resistance and revolution. Following the presentation and analysis of the six illustrative processes listed above, the chapter concludes with a summation of results.  4.2 Historical Background: Coca and Cocaleros in the Chapare In this section I situate the contention briefly in the context of both Bolivia’s larger political and economic history, as well as the region’s increasingly coca-centred economy. Bolivia’s political history is a tumultuous one. It has the “dubious distinction” of undergoing more than 150 coups d’état since 1825 (Kohl and Farthing 2006: 37). Some changes have been particularly dramatic. For my purposes, Bolivian history may be usefully subdivided into four distinct periods during the last 60 years which for the sake of convenience I'll refer to collectively as the "modern" era: the national revolution of 1952 led by the Movimiento 123  Nacionalista Revolucionario (MNR), and its aftermath (1952-1964); military dominance under a series of dictatorships (1964-1982), most notably those of Generals René Barrientos (1966-1969) and Hugo Banzer (1971-1978); democratization and liberalization under a series of “pacted” governments based upon agreements among parties to share power in ways generally aligned with middle- and upper class urban mestizo interests, a period marked by tremendous economic upheaval and renewed political activism within the country (1982-2005); and the current “post-liberal”76 period under President Evo Morales, the country's first leader from Bolivia's indigenous majority. Throughout the modern era, Bolivia has experienced a stunning variety of political contention, varying in form, objective, and intensity. Protesters were often peaceful, engaging in a “standard” repertoire of activities associated with popular protest, including work and hunger strikes, protests, marches, and road blockades. The state's responses varied considerably and were at times brutal, particularly during the dictatorial era. That said there were a handful of attempted insurgencies in the country in the modern (i.e. post-revolutionary) era, all of which foundered. Notable examples include Che Guevara's ill-fated 1967 expedition, known as the National Liberation Army of Bolivia (Ejection de Liberation Nacional de Bolivia, ELN), the even less successful Teoponte Guerrilla of 1970, and the Tupac Katari Guerrilla Army (Ejército Guerrilla Tupac Katari, EGTK) (Klein 2011: 242-243).  At the same time, the state has undergone a series of significant economic transformations. Following a shift to import substitution and industrialization in 1956, the                                                 76 Post-liberal is a term used by Yashar (2005) to denote the period of indigenous-led governance in Bolivia, one that highlights breaks with the country’s previous consensus concerning the implementation of Washington Consensus principles such as privatization, government downsizing, and opening the country to international trade and investment flows. 124  Bolivian economy began to stagnate through the 1960s and 1970s. It became increasingly dependent on aid and high tin export prices, subsequently supplemented by natural gas exports and what proved to be unwise “petrodollar”-enabled loans. By the mid 1970s Bolivia faced a growing set of economic problems, which only worsened during the years following the country's transition to democracy. To deal with the crisis, which among other things included a plummeting currency and hyperinflation, in 1985 the newly elected Victor Paz Estenssoro brought in Supreme Decree 21060. This law instituted a broad series of austerity measures as part of an International Monetary Fund (IMF)-coordinated structural adjustment program (SAP), referred to in Bolivia as the New Economic Policy (NEP) (Kohl and Farthing 2006: 62-70). While the plan did restore macroeconomic stability, the transition threw tens of thousands of Bolivian state-funded employees out of work as the government shed jobs and privatized industries. Of particular note, the restructuring included the shuttering of a significant portion of the country's tin mines, a move that among other things effectively broke the opposition that organized labour had posed to structural economic reform. Given that corporatist labour organizations spearheaded by the militant miners’ unions had played a significant role in Bolivian politics since the start of the modern era and had successfully stymied previous attempts, the importance of this development ought not be understated. In the short term, it allowed the government to pursue a significant agenda of economic and political transformation with limited opposition. In later years however, it actually served to generate new forms and sites of political protest, as relocated miners assumed leadership roles in opposition movements across the country, notably in the indigenous city of El Alto near La Paz, and to a lesser extent in the Chapare as well as discussed below. 125  Additional efforts to privatize the economy continued into the early 21st century in the natural resource sectors. The final, most contentious, and ultimately unsuccessful acts included an attempt to privatize the water utility in Cochabamba, and a series of struggles over control and revenues associated with the natural gas industry (Olivera and Lewis 2004, Kohl and Farthing 2006: 162-167). These two episodes served to provoke massive protests. These in turn produced the resignation of two presidents in the space of a few years, with a third—the very same Hugo Banzer, reinvented as a right-wing democrat—resigning in the same period due to ill health. It was during this final era of government turbulence that the MAS (Movement Toward Socialism, or Movimiento al Socialismo) developed from a fringe political movement with representation concentrated in coca growing regions, to a national party with support from allied social groups across the country.77 4.2.1 Economics of Coca in the Chapare Coca was not always a primary crop in the Chapare. Indeed, until relatively recently there were no crops and indeed few inhabitants; even in the first decades of significant permanent colonization in the 1950s and 60s, coca was grown largely to meet local needs of farmers, constituting just one of several crops—including rice, yuca (a type of potato), palm heart (palmito), papaya and other tropical fruits—that together with resources gathered, fished, and hunted from the surrounding jungle constituted the combined produce of a typical farmer-settler in the region. Taken together, those resources often provided little more than a subsistence level                                                 77 It remains a point of debate the extent to which the party can be distinguished from the broader social movements that produced it and continue to provide the backbone of its support, most notably rural campesino and indigenous urban organizations such as the CSUTCB, and the national women’s labour group Bartolina Sisa. 126  of income in a region marked by widespread poverty well into the 1970s (Salazar Ortuño et al. 2008: 89-90).78 While coca production in Bolivia dates back to the pre-colonial era—indeed, records of its use apparently date back to 3000 BC—its mercurial rise to prominence as the country's most valuable and controversial cash crop constitutes a very recent phenomenon. Production in the country increased in the 1970s, driven by both the rising international demand and Bolivia’s own economic disruptions. During the cocaine boom of the 1970s and 1980s the narcotic became the drug of choice in both US and Europe. Within Bolivia, a massive drought in the country's highlands in 1983-4 increased the flow of dislocated migrant labour looking for alternative livelihoods (Painter 1994: 6). The former led to a rapid increase in price, while the latter ensured a supply of labour to either work on established farms or clear out new areas for production.  Though not the only region producing coca in the country, by the mid 1960s it had become the most important. The Chapare region proved uniquely suitable to the coca plant given its high temperatures and significant rainfall, as well as a particular compatibility of soil and landscape. In 1967 that it passed the Yungas region of La Paz—the other significant coca producing area of the country—as the region with the greatest land area under coca cultivation; by 1970 the Chapare accounted for more than 4/5ths of the country's total output (Painter 1994: 3). Its isolated, relatively unpopulated jungle areas lent it additional advantages during the height of the coca boom, as planes smuggling the leaf, or in some cases partially processed coca paste, landed on and took off from small hidden runways scattered throughout the region. In the                                                 78 In previous eras the primary export varied further, as in the rubber boom of the late 19th and early 20th centuries. Prior to that, when groups indigenous to the region (and not just the country—a fact that itself occasionally attains political salience as some question whether indigenous groups that settle in new parts of the country remain  "originarios," or instead ought to be considered themselves as colonists) constituted the bulk of the population, the economy was largely subsistence based, with little in the way of significant exports. 127  ensuing two decades, the total area under cultivation in the Chapare grew from 2,650 ha in 1970, to a high point of 51,798 ha in 1987.79 The price of coca grew rapidly throughout the period as well, finally peaking and then crashing in 1986. A 100 lb. bag of picked and dried leaves reportedly sold for $300 at one point in 1986; in the five following years prices were on average closer to $60/bag. Even after that precipitous decline, coca remained more lucrative than other comparable crops however; accordingly, most of the region’s producers preferred it to alternatives, even when the latter were heavily subsidized. Combined with its other advantages, including multiple harvests (about four per year in the Chapare), its suitability to the climate, its portability, and the easy access to purchasers, coca became the primary crop of the region. Farmers were able to take their harvest to market by mule, bicycle, or in more recent eras by motorcycle, or trufi, a form of shared taxi-bus. Conversely, heavy fruit such as pineapple had to be hauled at the farmers' expense to Cochabamba or even La Paz for sale, absorbing a significant portion of the total proceeds in the process. Indeed, lack of access to markets remained a central weakness dogging alternative development projects throughout the era, a point flagged by many interview participants. In sum, from the mid 1970s to the present, coca remained the most profitable crop in the region by far. Even with the various alternative development schemes offered and implemented by the state, for most of the region’s inhabitants, coca remained the primary source of income, and the best hope for a sustainable livelihood. Indeed it seems plausible, if beyond the scope of the present dissertation to prove, that organization of cocalero communities forestalled more significant direct involvement by cocaine traffickers, or other armed group seeking to capture                                                 79 Painter 1994: 15, based upon numbers from SUBSEDAL, superficie y producción de Coca en el Chapare y Yungas, Mimeo (La Paz, Bolivia: SUBDESAL, 1991) 128  additional value from the coca and cocaine trade, in the region. With cocaleros able to supply the leaf so efficiently, there was no reason to engage in vertical integration of the industry along lines seen in other coca producing countries. With cocaleros watching one another for adherence to shared principles of growing, rapacious competition among producers failed to materialize. Isolated disputes, when they occurred, were dealt with within the federation’s own internal conflict resolution processes as described in Chapter Three and illustrated below. 4.2.2 Symbolism of Coca Beyond the economic aspect, coca has long played a role in indigenous Bolivian life, not just in the Chapare but also throughout the highlands. It serves as medicinal herb, mild stimulant, herbal tea, social lubricant, treatment for altitude sickness and indigestion, source of strength for labourers working on empty stomachs, alongside numerous symbolic functions. Writing in 1981, as pressure was first rising to eradicate the plant, Catherine Allen described it in unequivocal terms: "coca is an integral part of Quechua life in Peru and Bolivia and is a powerful symbol of cultural identity" (1981: 157). The names and specific content of traditions vary by region and purpose. Allen, for instance, writes of "hallpay," a coca sharing ceremony she observed in Southern Peru. In the Chapare region and surrounding areas, inhabitants use the term "pijcheo" to refer to coca mastication in general. The total use of the plant remains widespread, with one 2013 report finding more than 3 million Bolivians—nearly 1/3 of the country—reported using the leaf, with more than a million identifying the practice of pijcheo specifically.80 Grisaffi (2010) puts forward the argument that the coca leaf actually assumed even greater importance during the course of the struggle, as it came to do significant symbolic work                                                 80 . Los Tiempos Accessed 6 October 2015. 129  for cocaleros and even other indigenous actors in the region not directly involved in the production of the hoja sagrada (sacred leaf). It provided a unifying symbol for key concepts such as sovereignty, nation, and indigeneity, a function particularly important among discursive communities that lack a shared conception of those concepts of their own. “It is without doubt the perfect empty signifier” (2010: 427).81  Senator Julio Salazar used quite similar terms in an interview when describing coca leaf’s symbolic and cultural role:  The coca leaf represents our cultural identity as indigenous campesino peoples, who are Aymara and Quechua. The coca leaf has been used for centuries [unintelligible] before colonization. We are speaking more than 500 years ago; at the time of the colony. Now in this juncture our coca leaf, as a representative of our cultural identity is used in traditional uses, cultural rituals, social—when there are events. So that’s part of our survival and moreover the coca leaf in its natural state does not hurt men; it is not poison, but rather food and medicine.  Among other things, it provided a convenient mechanism through which to answer what might otherwise have been awkward questions regarding the extent to which recent arrived colonists may claim to be originarios [first peoples], entitled to culturally embedded rights as a result, and how their claims compared with the small minority actually descended from the region’s first                                                 81 Grisaffi credits Laclau (2005) with the concept of an “empty signifier”, a cultural artifact onto which discursive communities ascribe meaning. I consider Grisaffi’s argument correct, to a point. The coca leaf undoubtedly did vital symbolic work in unifying the inhabitants of the region against unwelcome state interventions, conveniently standing in for a cause to be championed in the context of a host of issues ranging from land control to economic development, to cultural and political autonomy. The leaf was commonly invoked in broader calls for indigenous control of tierra y territorio, or land and territory. Indeed, it is noteworthy that tierra y territorio became the founding principle of the instrumento político that proved fundamental both to the two-pronged strategy advocated by Escóbar 2008 in the context of the Chapare (electoral politics plus popular protest), and the creation of the country-wide coalition of indigenous political movements that pushed the MAS to victory (FSUTCB: 1995). Having said that, the idea of coca as an “empty signifier” may overstate the case, since the leaf in fact exhibits concrete physical and economic characteristics that both generate its utility within indigenous culture, and shape the pattern of contention over its production in the region. Rallying around coca is not simply a discursive act; on the contrary, it is explicitly linked to the continuing livelihood of the Chapare and its inhabitants, representing as it does a very real escape from subsistence existence that remains the fate of most other small agricultural producers in the country. 130  inhabitants. 82  Put more positively, the coca leaf became a symbol of resistance, providing a common source of identity that allowed the region’s inhabitants—who in reality came from remarkably distinct backgrounds—to rally together against unwelcome state intrusions in the region. Original inhabitant or settler; Quechua, Aymara, or Guarani; ex-miner, rancher, or former teacher—the coca leaf carried not only economic but also cultural significance for all of them, eventually becoming the symbol of indigenous peasants’ existence in the Chapare. As such, any attack on it would be treated with the utmost gravity, threatening identity and livelihoods alike. 4.2.3 Arc of State Involvement in Coca The Bolivian state’s approach to coca changed and changed again over the course of the modern era. The general arc was from a situation of effective neglect, to symbolic opposition through the 1960s and early 1970s, to increasing—albeit illegal—involvement through the latter years of dictatorship, to an increasingly, if reluctant, embrace of serious efforts at reduction and ultimately eradication over the course of the 1980s. The intensity of that effort continued to ebb and flow through the course of the 1990s, with a balance of forces combining to limit the state’s uncontrolled embrace of violence, but peaked under President Banzer (1997-2001) and his immediate successors. Dictatorships 1964-1981: From Benign Neglect to Avarice While the government had been a signatory to the 1961 Vienna Convention on controlled substances under which the trafficking of cocaine, and indeed even the harvesting and sale of coca is considered a crime, the government of Bolivia passed no domestic legislation rendering coca production illegal until 1988. It was not until the beginning of the 1980s that the country                                                 82 The relatively small populations who actually are original inhabitants of the region identify strongly with the jungle as their ancestral home. In contrast, Quechua and Aymara culture identifies more strongly with an Incan past, and more commonly regard the highlands as their region of origin. 131  began pursuing coca control of any kind in earnest. Essentially, prior to the 1970s, there was no reason to antagonize rural Bolivians for whom the coca leaf held such value, particularly given how important their support was for successive regimes (Klein 2011). Even after trafficking of cocaine and its precursors accelerated during the growing boom of the late 1970s and early 1980s, there were considerable personal incentives for state actors to go slow on attempts at coca reduction and eradication. Healy, for instance, points out that notwithstanding the profits for cocaleros themselves, the elite groups that benefited most directly from the cocaine trade—rather than the more general production and sale of coca—existed for the most part outside the region. Healy (1986: 105-106) identified three distinct groups, while acknowledging that significant overlap existed between them. These included • Landed elites of Santa Cruz, whose families had avoided the land reform redistribution of the 1950s, and who had developed experience in international trade in subsequent decades through their agribusiness activities, notably coca and subsequently soybean;   • Large scale cattle ranchers in the Beni department, who enjoyed dual advantages of relative isolation, aiding the illicit landing and take-off of small aircraft, and significant support from the Bolivian state in the form of modernization processes (this latter advantage, Healy asserts, stemmed from the tendency for senior military officers to take on ranches in the region, ensuring close links with the central political force in the country); and  • From the late 1970s on, the military itself played a significant role in drug trafficking, most notoriously during the Presidency of General Luis García Meza (1980-1981) under the tutelage of his Minster of the Interior, Colonel Luis Arce Gómez.   Given that such a prominent segment of the country's elite found the industry profitable, it is no wonder that the government proved reluctant to take action to reduce coca production or even decisive steps against the more obviously illegal activities associated with cocaine processing and trafficking.  There were some moves to institute what would become the three main pillars of Bolivian antidrug policy in the ensuing decades: interdiction, eradication, and alternative 132  development. In 1973 for instance, the Banzer government created a national directorate for the control of dangerous substances (Painter 1994: 78). In 1974, the US agreed to fund Proyecto de Desarollo Chapare-Yungas (PRODES), an alternative development project. Even so, the industry grew, and the Bolivian government seemed to be in no hurry to take decisive action.  It was under García Meza in 1980 that the country reached a nadir, becoming known internationally as a narco-state, with the president and his supporters openly funded by and supportive of drug trafficking cartels in the country (Hargreaves 1992). The excesses of the era, both in the use of repressive violence and open criminality, played a prominent role in convincing Bolivians of the unacceptable dangers that came with military dictatorship. Perhaps just as importantly, it forced the United States, first under President Jimmy Carter and continuing under President Ronald Reagan, to keep its distance—a marked change from the previous period, when successive dictatorships had enjoyed significant US support as allies in the Cold War.  The notoriety of the García Meza dictatorship and its aftermath proved a turning point in a number of ways. First, it sounded the death knell for overt military involvement in Bolivian politics. Within a year of his resignation, years of unstable military rule ended as even diehards in the military high command concluded that a retreat to the barracks was the only acceptable way forward. García Meza's immediate successor was himself replaced by another general charged with overseeing fresh elections. Second, it heralded a return to democracy and may have even provided the fledgling new democracy greater legitimacy and popular support. Bolivians I spoke with still recall the excesses of the late dictatorial period. As I discuss in Chapter Five this actually contributed to the subsequent stability of Bolivian democracy, convincing the population that democratic institutions constituted the only viable option for the country. Third, it proved a catalyst for even greater international—particularly US—pressure for Bolivia to take drug 133  enforcement policy seriously. Less than three months after García Meza's resignation, his successor Celso Torrelio enacted "Legislative Decree no.18714 Regarding Control and Combat Against Dangerous Substances, a move towards greater control of coca and cocaine. The decree included several elements that would become hallmarks of antidrug efforts throughout the era: proclamations of grand plans either with the support of coca unions, or followed by such agreements regarding specific elements of implementation, marked by accusations of breech on both sides soon after, and a gradual drift setting in until the arrival of a new government, an increase in US pressure, or some combination of the two prompted a renewal of the process. The 1980s: Democratization and a Reluctant Hardening The year 1982 saw both the long-awaited return to democracy, and the country's first comprehensive plan for coca reduction, known as the Plan Quinquenal (five year plan). In what would prove an important, though limited source of stability through the early years of confrontation between cocaleros and the state, this initial plan and its successors emerged following negotiations between the government and cocaleros. While committing the government to coca reduction targets for the first time, it also linked those reductions specifically to the implementation of alternative development projects. Though not a formal part of the plan, the approach also reinforced a norm of consultation in the articulation of state coca strategies—an implicit acknowledgement of the cocalero power to organize and resist unwelcome and unpopular government actions, even in the pre eradication era, and a template for conflict resolution even at the height of violence and resistance under President Banzer’s Plan Dignidad.  Despite the presence of such commitments, successive governments in the 1980s remained somewhat ambivalent in terms of concrete actions regarding coca production and trafficking, undertaking sufficient action to antagonize cocaleros and trouble minor drug 134  traffickers, but not enough to significantly affect actual levels of production and export. Painter (1994: 82) for instance observed that the DEA and UMOPAR (The Unidad Móvil Policial para Áreas Rurales, Mobile Police Unit for Rural Areas)83 increasingly believed that harassing small producers would be one aspect of a successful policy designed to reduce both the supply and price of coca paste. Such strategies at the time created concerns among observers that the result might be to “Colombianize” the politics of the coca growing regions, or even render those areas susceptible to the kind of recruitment undertaken by Sendero Luminoso in Peru’s coca growing region of upper Huallaga Valley (Painter 1994: 95). In one of the country’s chief ironies, that reticence proved important during the economic crisis and subsequent restructuring of the mid 1980s and beyond. Coca, as a significant source of employment and international hard currency for the country, acted as a crucial shock absorber easing the pain inflicted first through the chaos of economic crisis and hyperinflation, and second as a result of the structural adjustment used to stabilize the economy (Painter 1994: 54). In addition to these trends and transformations at the national and international levels, respectively, there were additional changes in the Chapare itself that further facilitated the expansion of coca as the staple crop for the region, as transport systems opened up the region to greater settlement, including the network of river ports (Spedding 2004: 93) and subsequently a highway connecting Cochabamba with Santa Cruz in the east of the country.84                                                  83 A branch of the Bolivian national police within the larger FELCN.  84 This highway, even before its completion, played a central role in shaping the history of the region. Prior to its construction, Painter finds reports that stretches were used for take-off and landing by small aircraft. After its completion, the road, while at times quite slow, became the primary route linking the east and west of the country and by extension the northernmost highway traversing the continent south of the Amazon basin. As such, it quickly took on great importance in the country’s economic activity, sending a significant flow of traffic through the heart of the Chapare as a result. This provided cocaleros with a significant strategic option that they employed to great effect at various times: the blockade. With the road winding about 100 km through the region, the cocaleros able to institute a system of blockades and checkpoints stretched over more territory than the military could effectively clear 135 1990s and the Turn to Militarization Under increasing pressure from the US and its hardline approach to drugs, the Bolivian approach finally hardened in the 1990s. The US adopted a “distant interdiction” approach under the “Bennett Plan” developed by Bill Bennett, George H.W. Bush’s “Drug Czar.” The approach focused on interrupting supply before it reached the US border, leading to more direct involvement in counter-narcotics training and operations in drug producing and trafficking states throughout the region, including Bolivia. The Cartagena Accord in 1990 proved a notable landmark. The agreement’s secret “Annex III” called for the direct involvement of the Bolivian armed forces in drug interdiction operations in exchange for US military aid. Once it became public, the agreement became quite controversial and was widely criticized across the political spectrum and even by members of the governing MIR-UDP coalition, exposing as it did two extremely sensitive political topics in addition to coca itself: the sovereignty of the country and ability to pursue policy independent of the United States, and the role of the armed forces in a country then less than a decade removed from military dictatorship (Painter 1994). Accordingly, the move towards an increased military presence in the region proceeded slowly and uncertainly in the wake of the agreement. Nonetheless, by April 1991 there were US trainers at work in the country (Painter 1994; Salarzar Ortuño et al 2008: 178), and shortly thereafter reports emerged of forced eradications taking place in the Chapare. Cocalero response was swift. Leaders denounced the new level of militarization in the region and the accompanying reports of abuse, including physical attacks, theft and burglary, and extortion.85                                                                                                                                                         at any one moment. As the following discussions makes clear, a number of the moments of high tension of the era involved blockades of the road. 85 Presencia 14 April 1991, reported in Salazar Ortuño et al. 2008: 179. 136  In the following months, numerous patterns emerged that would recur with varying levels of frequency and intensity throughout the following twelve years of contention, building upon existing repertoires developed during the previous decade of more limited confrontations. Reports of confrontations between eradication parties and local comités de autodefensa soon emerged.86 On 18 June 1991, the cocaleros under the newly formalized coordinadora began a blockade of the Cochabamba-Santa Cruz highway. On the first day of blockade, 91 cocaleros were detained and subsequently released following negotiations. In response, other cocaleros staged a “pijcheo-vigilia”—essentially a mass “chew-in” of coca leaves—in protest. The blockade ended the following day, after an agreement between the maximal national labour representatives of the COB and the government (Salazar Ortuño et al. 2008: 180-181).  Nonetheless, eradications continued, as did cocalero resistance. The first march from the Chapare began just five days later, on 24 June, for “Sovereignty and National Dignity”. The march was in support of a series of demands focusing primarily on the involvement of Bolivian armed forces and the interventions by US agents in the region (Salazar Ortuño et al. 2008). The march, involving between 400-700 campesinos, each representing and supported financially by one syndicate, lasted eight days before being broken up by army and police forces, resulting in the death of one cocalero. Security forced detained Morales and 20 other leaders (Pinto Ocampo 2004: 10). The campaign of eradication continued until November 1991, with numerous clashes as massed groups of cocaleros confronted armed eradication teams. One confrontation prompted by an attempt by DIRECO (Dirección de Reducción y Control de la Coca, the state agency for coca                                                 86 El Mundo, Santa Cruz 17 May 1991. Reported in Salazar Ortuño et al. 2008: 180.) 137  control) and UMOPAR agents to undertake forced eradication of a coca crop in Sindicato Santa Fe saw four soldiers wounded by machete on the government side, while one cocalero was killed and about 10 others wounded (Salazar Ortuño et al 2008: 188). In both May and November, another standard element of confrontation occurred in the form of negotiated agreements between cocalero representatives and the state, in which cocaleros reaffirmed their opposition to narcotrafficking along with their own right to produce coca, and committed to end more controversial strategies of resistance in exchange for commitments to honour human rights, avoid military targeting of cocaleros, and pursue more effective strategies of alternative development to support campesinos who abandoned coca. The agreement lacked firm responsibilities or timetables, but did reinforce the legitimacy of cocaleros and their claims as interlocutors of the state (CEDIB 1992; Salazar Ortuño et al. 2008: 179-184.) While 1992 was a year of relative calm, thanks in large part to Bolivia’s achieving its coca reduction target, by 1993 confrontations began increasing once again. The country missed its targets in both 1993 and 1994, as it seemed voluntary eradication had largely run its course. Some producers who had participated in voluntary reduction began to openly replant while claiming that promised strategies of alternative development had failed (Marcy 2010: 9987, Salazar Ortuño et al. 2008: 191-192). By July 1993, the coordinadora announced a re-initiation of the SDCs.88  By the end of 1993, newly elected President Gonzalo “Goni” Sanchez de Lozada announced a new plan of coca reduction known as “Opción Cero” (Option Zero). The renewed emphasis on forced eradication led to new rounds of protests. By 1994 events had escalated once                                                 87 Quoted in Kleiman and Hawdon 2011. 88 Opinión 21 July 1993, quoted in Salazar Ortuño et al. 2008: 199-200. 138  again. In response to US pressure following the missed reduction targets, the government launched a new joint operation, Nuevo Amanecer (New Dawn), involving some 800 soldiers from the Fuerza Especial de Lucha Contra el Narcotráfico (Special Force for the Fight against Drug Trafficking, FELCN), UMOPAR, and police. (Contreras, 1994:10; Salazar Ortuño et al. 2008: 219). During this period, Morales mused publicly about the possibility of civil war resembling that seen in Colombia. The Federation in its rejection of Opción Cero began to use the language of “ethnocide and genocide” in reference to fears that the government would go so far as to engage in depopulation of the region to meet coca reduction targets. 89 Having five deaths at this point is very serious and we should reflect deeply, if we want to maintain the precious social peace in the country… this problem [indiscriminate violence and deaths] will lead to an armed uprising on the part of my fellow growers, who are actually exceeding control and leadership…. At this rate, the compañeros of the Chapare will go beyond us [the leaders] because in many meetings they have assured me that there are thousands of “Evos Morales” [and] with Evo or not, will continue the fight in defense of coca plantations…. This time I felt in the flesh and I've heard many times that farmers are willing to resist because the government has decided to eradicate them… Now if they come, we are ready, whether we lose or we win. (La Razon 3 Sept 1995)   The cocaleros again responded on numerous fronts, most notably organizing the “March for Coca, Dignity, and National Sovereignty” from the Chapare to La Paz. The march produced significant new confrontations. State forces detained Morales and fellow cocalero leader David Herrara in an unsuccessful attempt to prevent the march. During the first day, 200 cocaleros were reportedly detained, and the group faced tear gas fired from helicopters about 10 km from the starting point of Villa Tunari. A column of some 1500 cocaleros continued on to Cochabamba                                                 89 For example, this passage appeared in 1995: “producers of coca in the Cochabambino tropic and the Yungas of La Paz finalized the XII National Meeting, ratifying their rejection and total repudiation of the Option Zero as ‘irrational and inhumane, because it is not a proposal against narcotrafficking, but rather the syndicates; a policy of ethnocide, genocide, and economicide,’ said Evo Morales Ayma, president of the Coordinating Committee of the Five Federations of the Tropics” (El Mundo 20 January 1995).  139  and eventually to La Paz, attracting national attention in doing so. The marches also had a catalyzing effect in the region. AC, an urban resident of Villa Tunari and executive in one of the “OTBs”90 in the region described it as follows:  In June of ’88, there was a massacre here in Villa Tunari, when they fired bullets at the unarmed campesinos. Several died, several drowned… There was resentment, but regardless they started that forced reduction of coca. The campesinos above all organized themselves with the current president at the head. They began to defend their areas and through the blockade, and the marches above they—marched, and they socialized…. They made the first marches for land and territory and dignity and the homeland. That was the way they organized all of the tropics and began a fierce defence of the coca fields.  While the march resulted in a new agreement between cocaleros and the state, the confrontations continued. By early 1995, the state was again under US financial pressure and announced the resumption of forced eradication of any new coca plantations. In response, cocalero leaders introduced new, more escalatory language of resistance, and continued to organize both high profile protests and on-the-ground resistance in the form of self-defence committees. In the same year, the cocaleros planned and executed a women’s march on the capital La Paz, in order to speak “woman to woman” with the wives of the president and vice president. Though the government employed a variety of tactics to forestall the marchers—including everything from new offers of negotiations (albeit with conditions), to an offer of rides—but the women reached the city and the meeting happened in due course in early 1996. Though those talks initially ended without new agreement on eradication policy, the cocaleros were able to maintain publicity and                                                 90 OTBs, formally Organización Territorial de Base (Grassroots Territorial Organization) are local representative organizations created and granted access to state financial resources with the enactment of 1994’s Popular Participation Law (Altman and Lalander 2003). They were intended to provide an alternative form of local representation where more traditional forms—ayllus, unions, and so on—did not already exist. See Altman and Lalander (2003), and also Lucero (2008: 134-135). Accordingly they existed as one of the few organizational structures in the region that the federations could not coopt and dominate directly; nonetheless, by the time of my visit they were able to exercise significant influence even in the absence of direct control over the urban organizational structures.  140  pressure as a group of 50 cocaleras began a hunger strike. Meanwhile cocaleros and their allies in the CSUTCB initiated blockades in support. Eventually, with the intervention of the COB talks resumed in February 1996 between cocalero federations, the cocalero women, the COB, and the state. That meeting resulted in an agreement, though without a lasting resolution of the central issue of coca production, the subsequent calm was again short-lived (Pinto Ocampo 2004: 12-14). Maximal Repression under Banzer and Successors, 1997-2003 The most intense period of confrontation occurred following the 1997 election of former dictator turned democrat, General Hugo Banzer, and continued under his immediate successors. Deaths became much more common from 1998 onwards, as the state actively pursued coca eradication, and the cocaleros persistently organized in opposition to newly constituted and reinforced eradication expeditionary forces that proved more willing to engage in violence than previous state agents in the region. While Banzer’s Plan Dignidad retained financial support for farmers who voluntarily eradicated coca plantations, it also for the first time introduced a system of declining compensation in exchange for eradication in an effort to encourage early participation: by the beginning of 1999 no individual compensation would be provided, and by 2002 no communal compensation would be provided either; typically, previous plans had provided for both types. While designed to shift cocalero incentives, the plans failed to do so meaningfully. With the general failure of alternative development schemes at that point to produce sustainable and widespread results, coca, even if illicitly produced and with crops regularly destroyed via eradication, remained the financially superior option to other available alternatives. The strategies, with the significant additional force involved, pushed cocaleros closer to their reserve point regarding their own use of violence. As figure 4.1 shows, from 1998 onwards 141  a small number of casualties appeared with regularity among state forces as well, since cocaleros began to respond more violently in a limited number of cases. As with previous rounds, the cocalero response to the newly announced plans was swift. Cocaleros declared a state of emergency and launched a series of significant marches, protests, and a blockade within first four months of 1998. The latter resulted in multiple deaths, dozens of injuries, and hundreds detained in mass arrests. Figure 4.1, produced using data from Salazar Ortuño et al. (2008: 805) makes clear just how much violence escalated in those final years. The hashed line indicates deaths among the cocalero federations as a result of confrontations, while the solid line denotes deaths among state agents.   Figure 4.1: Regional Deaths by Presidential Period Source: Salazar Ortuño et al. 2008 Similar patterns exist regarding the total recorded detained and abused during the latter years. Salazar Ortuño et al. (2008) record a total of 77 incidents of physical abuse from 1993-1997, 112 during Banzer’s administration from 1997-2001, 154 during the one year of Jorge Quiroga’s presidency, and another 80 from Sánchez Lozada’s return to office in 2001 until his resignation in 2002. Those numbers reflected the steady drumbeat of incidents across the region. 0	5	10	15	20	25	1980-82	 1982-85	 1985-89	 1989-93	 1993-97	 1997-2001	2001-2002	2002-2003	2003-2004	civilian	deaths	 state	deaths	142  While the biggest flashpoints were attempts by the state at forced eradication and cocalero attempts to stage blockades, marches and other forms of demonstration, there were more quotidian instances of violence as well.  As the conflict wore on, state strategies varied. They became quite invasive under Banzer, including the occupation of small towns and expulsion of their populations, the closure of bus stations to prevent cocaleros from travelling to Cochabamba for protests (Salazar Ortuño et al. 2008: 565, 645), and the occupation and destruction of equipment in the cocaleros’ radio station in an attempt to hamper coordination efforts.91 The primary variance was one of intensity of the violence rather than the type, with two major exceptions. The first was the creation of a Joint Task Force of military and police officials, known by its acronym FTC (Fuerza de Tarea Conjunta), a force responsible for a significant increase in both the success of forced eradication efforts, and the social cost that resulted (Ledebur 2002). The second was the deployment of the FTE (Fuerza de Tarea Expedicionaria), contract soldiers that the Defensor del Pueblo described as mercenaries operating outside the law (Ledebur 2002). Both of these new formations appeared to resort to the use of violence more often than regular military forces or police deployed in the area. Their actions produced the heaviest casualties of the conflict. Additionally, during this period coca leaders worried increasingly about the threat of assassination. Violence directed at them was not new in the history of the conflict. Morales for instance endured several violent encounters during his time working for the federations,                                                 91 In one typical example, on 6 June 2001, a military force imposed martial law on the syndicates of Puerto Zudáñez and Santa Isabel, and expelled the population. Representatives of the Defensor del Pueblo of the Chapare verified the events. A Joint Task Force set up an encampment near the local soccer field, took over the school, and the road to a number of the settlers. The events were similar to another more violent incident in 1998 in which several houses were burned out, which itself was in retaliation for the detonation of cazabobo that injured three soldiers. (Salazar Ortuño et al. 2008: 464, 465, 561).  143  including one incident in 1989 in which he was beaten and left apparently for dead (Sivak 2010: 43; Salazar Ortuño et al. 2008: 157).92 The number and intensity of incidents created additional fears among leaders, however. Feliciano Mamani, then a member of the federation coordinadora, was shot in the leg during a blockade in 2000 by a police officer. Cocaleros alleged that there were also a team of snipers among the military forces deployed to oppose the blockade, with specific orders to target leaders (Salazar Ortuño et al. 2008: 512). A notorious event occurred in 2001 in the shooting death—allegedly by FTE forces—of Casimiro Huanca, then the Secretario ejecutivo (the senior-most official) of the Chimore Federation, during a peaceful protest (Ledebur 2002: 12).  Beyond such objective reports, it is clear that coca leaders themselves considered the danger to be extreme. Asterio Romero, a senior leader of the cocaleros, described sleeping in the open with marchers, as in part a function of cost effectiveness, but also a matter of security: …no special treatment, no executive who would take accommodation, to the hotel, none of that. I for example was an executive, a follower of the President; I slept outside with the bases [grassroots], so I have rarely fallen prey. If people thought I was in a hotel, they would have brought the police and everything, but I was not there; I was with the bases. I put on a ch'ulo [traditional woollen cap] and jacket.  The testimony from RM in the previous chapter regarding the need to meet in secret in the jungle using passwords to confirm identities gives a similar impression. Likewise, DM, an ex-leader of Federation Mamore-Bulo Bulo, described both the fear that came with being in a leadership position, and the precautions it necessitated: I with my colleagues in that time of struggle against forced eradication by the Joint Task Force, we suffered many consequences: the abuse, the mistreatment, psychological, even children. Women and men, in that time were perhaps very afraid to be leader; to be leader                                                 92 He reported experiencing several assassination attempts, and took elaborate security precautions as a result, moving constantly and rarely sleeping in the same place for an extended length of time (Sivak 2010: 98). 144  in that season was life and death, the leader in that time was going to live or die. It is for that that the University Casimiro Huanca of Chimore was built in his name, because he died from a bullet while blocking the road with products. Protesting the fact that alternative development had not progressed, he got a bullet from the army. And here in the same federation, personally when I left my house I wondered whether I’d arrive or not. I left my family because it was very risky and dangerous. It was not a very secure life.  And again, describing his work as a leader:  … afterwards things calmed down. They left, and we did too. The confrontation became calm. Under my management we lost three lives. That would be what I remember. There is the testimony of the doctors as well.  Our Headquarters was Central “22nd of May,” and there at any moment the bullet might arrive, targeting leaders. At times we slept in the bush, among the mosquitos. There was no choice of leaving because we were being chased, searched, and I did not sleep in my house. At times, I took my wife, my children out, because I could not sleep in my house. We were all pursued. (DM)   Such testimony makes clear the level of threat perceived by cocaleros; they believed that it was not just their livelihoods but their lives at stake. That perception of threat, I argue, is important to hold in mind when considering the decisions subsequently taken. In advocating both conventional and nonconventional political participation while rejecting anything resembling armed resistance, cocaleros were not only aware of the risks they ran if the government came to view the movement as an intolerable threat to stability but also—and this is the crucial point—able to manage the totality of the community’s response in ways that significantly limited opportunities for anyone, including marginal actors not deeply embedded in the cocalero federation structure, to engage in violence likely to provoke a heightened state response. Dénouement: Four Presidents and a Peace Agreement   The pattern of eradication, protest, negotiation and temporary calm repeated itself through the final six years of contention. By 2002 however, it had become clear that the policy was failing, not only in terms of the toll it took on inhabitants in the community but in its ongoing ability to 145  control coca production. While the eradications under Plan Dignidad reduced coca production in the country in 2000 to its lowest observed levels since the 1980s boom, by 2001 the area under production began to increase once again (UNODC 2000, 2002). At the same time, 2000 marked a turning point in the wider Bolivian political context as the first of a series of crises occurred in the form of national protests over a range of issues. With the resignation of President Banzer in 2000 due to ill health, his successor Jorge Quiroga lost in the following election in 2002, and Sanchez de Lozada returned once more to the presidency. The state by then was under sustained pressure on economic, political and social fronts throughout the country. The war on coca had lost much of whatever popular national support it had enjoyed. Indeed, in the 2002 election that returned Sanchez de Lozada to power, Congress chose him after he captured just 22.5% of the popular vote. Morales competed for the first time and finished a close second, with 20.9% of votes cast nationally. The result provided the clearest signal yet that the cocaleros and their cause enjoyed widespread sympathy in the country. The MAS party likewise fielded a slate of congressional candidates and finished with the second largest caucus in both the Chamber of Deputies and Senate. The party’s support, while concentrated in Cochabamba, extended across the country by then, with MAS representatives elected in six of the country’s nine Departments, including strong representation in La Paz, Oruro, and Potosi—the highland region, in which Aymaras constituted a majority.  Accordingly, following his election, Sanchez de Lozada undertook direct negotiations with Morales in the search for an acceptable solution to the coca issue. The process outlived his time in office, as he resigned in the face of a new wave of protests, this time centred in El Alto, the Aymara city beside the capital La Paz, an episode subsequently known as the “Gas War.” It 146  fell to his replacement, Carlos Mesa, to conclude the process, even as violence continued fitfully in the Chapare.  Finally, however, the government agreed to the central cocalero demand, the right to grow coca legally throughout the Chapare, and not just in the “traditional” zones identified in Law 1008. On 3 October 2004, the state and cocaleros signed something not unlike a peace treaty, committing the cocaleros to support continuing eradication efforts required of the state in treaties with the US in exchange for the right for each family in the federations to produce one “cato” of coca, roughly 1,600 m2. The state also committed to an array of new protections for human rights, compensation for the families of those killed, and care for those injured. With that, the conflict effectively ended. 4.3 Subcases: Key Processes Defining Cocalero Response In the following section, I present and analyze a number of processes and episodes that function effectively as within-case episodes or subcases. Each is exemplary of the effects that the existing OC institutional structure had on the course of contention. I examine how cocaleros, which had begun to fragment into a number of different and imperfectly aligned federations during the 1970s and early 1980s, quickly acted to resume close coordination of activities as soon as a genuine threat to their shared interests appeared. Effectively, I argue that the presence of a nested OC structure, instituted at the level of community but linked through the nested federation structure to higher levels of organization (central and federation) allowed cocaleros to swiftly overcome potential collective action and coordination problems that would have faced other communities facing a similar challenge without the benefit of an established institutional framework. I then consider the how the federations responded to a threat of fragmentation within the cocalero movement at the height the confrontation in the form of associations of non-coca 147  producers nominally allied with state interests in the region. Here I argue that the OC institutions provided a framework to manage internal conflicts among producers in the region within the federation structure, thereby forestalling a more extensive and permanent fragmentation of the region’s producers on the one hand, and the potential for intra communal violence that would have resulted from the conflict between defectors and cocalero loyalists. I subsequently examine the process of consensus building around the idea of electoral struggle supported by nonconventional forms of protest, with the significant elements of socialization that it entailed. In this subcase I examine how coca leaders, particularly Filemón Escóbar and the emerging leaders of the 1980s including Evo Morales and David Herrara undertook a campaign of seminars in the region to convince cocaleros in the region how they might best resist the state. In doing so however, the seminars had the effect constructing important elements of cocalero identity, which over time became entrenched through discourse as well as through institutional structures and practice. I then turn to questions of resource mobilization for the repeated campaigns of nonconventional political action, notably including marches to the national capital of La Paz, as well as blockades in the Chapare region. After this, I consider the emergence and activities of the comités de autodefensa, or self-defence committees, formed in the region to respond to the state’s increasing emphasis on forced rather than voluntary eradication. Here, I argue that the OC structure provided cocalero leaders and members alike with the means to engage in fine control of the level of violence employed in their resistance. In this section, I also consider the ability of the federations to influence even non-members’ efforts to engage in violent insurgency against the state (or indeed, anyone else). In doing so, the OCs provided leaders with the power to engage in limited acts of low-level violence that would not provoke unbearably violent responses 148  from the state, while simultaneously discouraging other actors from engaging in armed resistance without their consent (or more likely, had such an outcome come to pass, their active involvement). As such, the OC structure enabled the federations both to eliminate two common sources of armed violence conflict: 1) the acts of rogue agents within organizations with poor systems of command and control, and 2) the emergence of violent groups on the margins of society, with violence propagated by actors with little to lose and/or few links to the community. Simply put, the presence of the OC did not prevent the emergence of armed conflict outright, but ensured it would be the result of a deliberate decision by cocalero leaders, either at the behest of, or with the support of their base. Finally, I examine how cocaleros consciously engaged in attempts to frame how the confrontations with the state ought to be framed. Here I argue that the cocaleros had a strategic interest in appearing simultaneously vigorous, even implacable in opposition to state policies of eradication, while at the same time NOT posing any real threat to the state. This accounts not only for the carefully delineated limits on the use of violence, but also what amounted to a delicate public relations strategy targeting both domestic and international audiences in which cocalero representatives consistently resisted the idea that they themselves practiced or indeed condoned violence—even as they warned that violence could result from state actions. The chapter concludes with a section summarizing how other leading theories of conflict perform in light of the history described above and subcases delineated below. 4.3.1 The Unification of the Cocaleros Though the larger story of cocalero organization, as I argue, is one of bounded and inclusive community-based resistance, it is not a case of linear agglutination over time. On the contrary, at their founding the communities were, as I argue in the previous chapter, quite isolated not only 149  from the state, but from one another as well. It is precisely for this reason that new arrivals had to turn to one another for support. While the local syndicates joined together over time to form larger centrals and eventually federations, it was not a linear process; on the contrary, at key points in time cocaleros actually fragmented politically, potentially impairing or even undermining their capacity for collective action. How that process unfolded, and how the moments of fragmentation gave way to renewed coordination, constitute key moments against which to test my primary explanation against competing alternatives.  In general terms, two distinct periods of fragmentation occurred among cocaleros within the region. The first concerned the splintering of the federations during the years of dictatorship and the first decade following the return to democracy, while the second—discussed below—related to the emergence of “associations” dedicated to the production of non-coca crops, typically with assistance from the state and international agencies provided in exchange for voluntary eradication. I discuss each in turn, considering both rational-institutional and identity based aspects of the process of unification.  Analysis varies considerably regarding the cause of the formation of successive independent federations, with some accounts placing greater emphasis on local grievances among elites, and others emphasizing differences among cocaleros as to how to orient themselves relative to the state. The reality appears to be a combination of the two. Garcia Linera et al. (2005) attribute the break ups to "internal divisions," expressed most notably with regard to the divided federations' affiliations with different umbrella labour groups at the national level, most notably the Confederación Sindical de Colonizadores de Bolivia (Syndical Confederation of Colonizers of Bolivia, CSCB) and the Confederación Sindical Única de Trabajadores Campesinos de Bolivia (United Syndical Confederation of Peasant Workers, CSUTCB), an 150  organization established during the 1970s specifically to represent the interests of rural workers often excluded in the urban and miner-dominated national labour movement. Others such as Healy (1991: 89) note that the two earliest federations took different approaches in relation to the state, with the Federación Especial Agraria del Chapare Tropical continuing under the "tutelage" of successive Bolivian governments. Conversely, the Federación Especial de Colonizadores de Chimore existed as a more autonomous campesino-founded and led organization. Accordingly, it was the latter that first broke publicly from the Pacto Militar-Campesino in the 1970s.93 Other divisions carried political significance as well. The Federación Especial de Yungas Chapare (FEYCH) emerged in 1988 as a direct result of Ley 1008 enacted that same year. The law, which spelled out the country's policy towards coca, cocaine, and other controlled substances, made explicit the state’s commitment eliminate coca production in the Chapare region. Among many other provisions, the law divided the country into three zones with respect to coca production: a “traditional” zone in which coca could continue to be produced legally, a “transitional” zone in which production would be phased out, with the remainder declared as an “illegal” zone in which coca production became immediately illegal. The largest traditional zone emerged in another part of the country entirely, in the Yungas region of the department of La Paz. Most of the Chapare fell outside the traditional zone, the area that was identified as such broke away from the larger Federation of the Tropics to form the FEYCH, allowing the cocaleros                                                 93 The Military Campesino Pact was an alliance through which peasant leaders agreed to support the military governments of René Barrientos and his successors against opponents, largely located in the urban and mining centres. In exchange, they received certain considerations, including subsidies on food and fuel. The pact effectively ended in 1974 when security forces attacked and killed a number of campesinos that had staged a blockade to protest cuts to those subsidies. 151  there to chart a path independently of the threatened (and therefore more confrontational) majority.  More generally, as Stefanoni and do Alto note, the campesino labour movement as a whole in the Chapare was comparatively weak well into the 1980s, of little consequence even to leadership within the country’s union movement, let alone the governance of the country (2006: 40-41). That all changed during the latter part of the 1980s however. Among existing historical accounts of the region's politics, two factors dominate; to their account I add a third, below. The first was the arrival of former miners to the region. As part of the great economic restructuring of the mid 1980s—most notably in the form of Decreto Supremo 21060 (Supreme Decree 21060, the law instituting the IMF mandated Structural Adjustment Program, designed in large part by then Harvard economist Jeffrey Sachs—the state closed most of the country’s state-run mines. Taken together, the changes put 23,000 of the country's 30,000 miners out of work (Kohl and Farthing 2006: 71, citing Crabtree et al. 1987), along with another 25,000 rural teachers and 10,000 other state employees.  Some miners returned to their homes; others settled in the major cities, most notably the Aymara city of El Alto. A smaller number sought out the small alluvial gold mining deposits in the La Paz, Pando, and Beni departments (Painter 1994: 149 fn 31), citing testimony in Crabtree et al 1987). The Chapare absorbed a small portion of that outflow as well.94  Throughout the late dictatorial and early democratic era, coca emerged as one of the country’s few growing sources                                                 94 While specific data are not available regarding inflow and outflow of populations throughout the regions broken down by occupation, there is some indirect evidence suggesting the overall flow of miners to the region was limited. Painter (1994: 16), on the basis of DIRECO’s own database (reported more fully in Painter and Bedoya Garland (1991)) reports that only 7% of the farmers in the region arrived during the 1985-1989 period, when ex-miners were migrating. Conversely, 42.5% arrived during 1980-1984, coinciding with the period of extensive drought in the highlands and the precipitous rise in coca prices. 152  of employment, and a potential route out of subsistence existence in which so many Bolivians remained mired. As such, it (however inadvertently) provided a significant and rare source of stability for the country in the face of a growing economic crisis during the first three years of democracy from 1982-1985, under the ill-fated presidency of Hernán Siles Zuazo of the Popular and Democratic Unity Party (Unidad Democrática y Popular). As economic opportunities declined elsewhere, new producers flooded the region, resulting in the coca boom. While limited in absolute numbers, the miners who arrived brought with them a particular experience of labour organization and resistance to state policy, and tended to quickly become involved in political resistance; this pattern was observed in other locations where miners settled such as El Alto as well.95  Interviews within the region confirm that such was more or less the case. Several of the leaders I spoke with were either former miners themselves, or the children of ex-miners. Likewise, in substantive terms, most interview participants that I spoke with about the subject described how miners supported and improved upon the structures and patterns of resistance that were already present in the region, rather than developing new ones out of whole cloth upon arrival (e.g. Escóbar 2012). According to Julio Salazar, a MAS senator at the time of our interview in 2012, The miners were in search of sources of work. They especially came to the Chapare, and thanks to those miners we reinforced the syndical organization. I say that it’s because of the neoliberal parties the miners had to relocate. Took away their jobs and so, looking to survive, they came to the Chapare. For me imperialism is to blame for this migration.  Q: Was there a union structure before relocation?  JS: There was, there was. In other words I would say the Chapare is a centre for other cultures. There are the Aymara, the Quechua, Guarani, Trinidadians, shamans, Yuracarés                                                 95 Arbona (2008) for instance details the role that former miners played in organizing protests in El Alto, adapting as they did so to their mining union structure to the regionally dispersed neighbourhood associations through which the city was organized.  153  … many people have joined us in the Chapare. Why? Because we come from the east, west, from all departments. We were obligated to out of hunger and misery… here we are owners of these lands as Aymaras and Quechuas.  The second factor driving cocaleros’ political emergence was the policy and actions of the government itself in its increasing opposition to the cultivation of coca. From 1980 to 1988, the state's position on the subject steadily hardened, as the above account suggests. It shifted from the utter dependency of the short-lived but notorious regime of Gen. García Meza that finally—and in hindsight decisively—discredited the idea of military in government in Bolivia, to a focus on coca reduction in the early years of the decade. The shift led finally to 1988’s Law 1008 making eradication of “surplus production” of coca the stated goal of Bolivian antidrug policy.  Key points of transition include the initiation of forced eradication in 1981, while still under military dictatorship (at US behest); the planned use of chemical defoliants; the promulgation of a Plan Quinquenal (Five Year Plan) for the “reduction of the cultivation of coca and the repression of narcotrafficking” in 1982, to the initial militarization of the region in 1984, to the introduction of a second plan, to the Plan Trienal (Three Year Plan) in 1986 calling for the eradication of all “surplus production” in Bolivia including the majority of production in the Chapare. Throughout those evolving policies, a recurring yet steadily escalating pattern of contention, negotiation, and temporary resolution between state representatives and cocaleros marked the situation within the Chapare.  Indeed, for a time, it seemed the government might successfully enact a strategy of divide and conquer against the politically separated cocalero groups. The internal divisions that emerged during the 1970s and early 1980s came to capture significant differences of opinion regarding the state policy of coca reduction in the region. Following the promulgation of the Plan Trienal in 1985, for instance, the leadership of the Federation of the Tropics (the largest of 154  the federations at the time), signed an agreement to allow a sharp reduction in coca production in the region, establishing in writing principles such as a combination of cash for producers and development funds for communities in exchange for eradication. To that end, the agreement also included a price of US$350 in exchange for one hectare of voluntarily eradicated coca. One communiqué from ANAPCOCA, a group representing coca producers nationally, characterized the agreement as a “ridiculous stupidity,” insisting on an indemnification of $3,000 per ha (From Los Tiempos, 14/11/1985; quoted in Salazar Ortuño et al. 2008: 96). The same organization declared the leaders who signed that agreement, most notably Eudoro Barrientos, Executive Secretary of the Federation of the Tropics, as “traitors and enemies of the country's coca producers” (Salazar Ortuño et al. 2008: 90). Regardless, signing such agreements did not save cocaleros from instances of violence, with the state's counter-narcotics and counterinsurgency police force UMOPAR (referred to colloquially as “leopardos”) engaged in acts of violence as eradication continued. They targeted not just the obstructionist Carrasco Federation, but cocaleros more generally. The effect was galvanizing. Meetings ensued among cocaleros, both among themselves via national organs such as the Comisión Nacional de la Coca, as well as under the auspices of the national labour groups COB and CSUTCB. In 1988, at the first national meeting of coca producers, representatives called for unity in resistance to the government's now open pursuit of eradication. By 1989, the then five federations signed agreements of mutual defence, and created for the first time coordinated “comités de autodefensa” or self-defence committees through which to coordinate actions of resistance (Salazar Ortuño et al. 2008: 133, 154). By 1992, the then five federations had established a formal coordinating body, the Coordinadora, charged with achieving unity of deliberative action on the part of cocaleros.  Others also argue that the state’s heavy-handed 155  approach served to motivate and unite cocaleros. As one anonymous respondent, a long-time resident of the region—but not a member of the federations—put it, What I wanted to say was that the armed forces and the police in their interventions, with their presence, have at the same time strengthened the cocalero organizations. The greater the repression, the bigger the organization, just like anywhere. (Anon 1)  At this point, it is worth taking a step back. The narrative of progressive unification and resistance in the context of increasing state repression fits with some previous research findings regarding responses to repression. That literature is on the whole indeterminate however, as we continue to lack a theory linking repression to resistance. Sometimes the former causes the latter, and sometimes it either prevents or summarily crushes it. When resistance does emerge, sometimes it is violent, and sometimes it is not. Research remains unclear as to when we ought to expect one outcome or another. To be sure, we have some theories linking resistance with repression; for instance, the “inverted U” hypothesis suggests that unconsolidated regimes, whether characterized by some hybrid or anocratic regime type, or by an absolute lack of resources that prevents the consolidation of any type of government, tend to be associated with greater risk. Here again, however, there is no deterministic link between regime type and repression, or more importantly between repression and either onset or avoidance, a point that, as mentioned in previous chapters, Davenport (2007) makes explicitly in an expansive review of existing literature. It is only in recent years that the systematic exploration of what distinguishes violent from nonviolent resistance has begun in earnest, primarily in the form of work by researchers such as Chenoweth and Stephan (2011) and Cunningham (2013). Cunningham (2013) for instance usefully identifies a number of the most common correlates of civil resistance, while Chenoweth and Stephan (2011) provide a detailed results-based account of how it has produced successful political 156  change in a variety of cases. We continue to lack case-based research regarding the specific mechanisms by which actors opt for nonviolent resistance while eschewing armed rebellion and, just as importantly, the conditions under which we expect to see such outcomes. This then is the third—and novel—explanatory element I introduce in the above account—the role played by the presence of a specific kind of prior institutional presence. It is the presence of an OC, I argue, that proved instrumental in constituting the form and extent of the resistance that emerged. Without it, neither the new wave of arrivals with experience in collective action nor the emergence of a new and shared threat would have been sufficient to assure a unified and non-violent response. One could as easily imagine a scenario in which the new miners, flush with funds from the state buyout, would have generated considerable acrimony with their sudden arrival in the region, sowing divisions between established producers and the newcomers. Likewise, the presence of a state-based threat to the cocaleros could easily have fragmented the cocalero movements, with a large portion defecting to the government side; and/or small subset of cocaleros, or other external actors, pushing to acquire control of a significant share of the coca production in an effort to capture a significant share of a lucrative trade using violence if necessary. None of these counterfactual outcomes came to pass however; on the contrary, just as the anonymous interview respondent quoted above described, as the challenge and threat heightened so too did the cocaleros’ commitment to unified action in resistance. Rather, there were a series of meetings and a rapid changeover of leaders in the late 1980s and early 1990s following which cocaleros put in motion a sustained campaign of coordinated resistance. On this point, 1988 proved a decisive year on both the government and cocalero side. It was the year in which the state passed a law definitively calling for eradication of coca in the 157  region. On the cocalero side, Evo Morales was elected leader of the largest coca federation (Tropico de Cochabamba) in 1988. Cocaleros at the time saw the election as representing the emergence of a new generation of leaders, one that quickly replaced an established old guard (Stefanoni and do Alto 2006). The change in leadership heralded an equally significant change in policy, and indeed appeared to be such at the time as well. As David Herrara, another senior coca leader, describes it, In ’86 here in La Costanera [a stadium in Cochabamba used for large meetings], the front which we offered was called FAMA, the Frente Amplio de Masas Anti-Imperialistas [Broad Front of the Anti-Imperialist Masses]. It came from the thought that the imperialists were invading us politically, economically, socially. So, the only way to address that was to raise awareness, and to make known the reasons for the defence of Bolivians against the traitorous government of the time: the MNR, and afterwards the MIR and ADN.  Whereas the older generation tended to seek agreement rather than confrontation with the state, Morales, Herrara, and others like them took the opposite approach, employing inflammatory language, seeking to heighten confrontations, and refusing to compromise on crucial points, most notably the right to grow coca. In addition to the change in leadership and the passage of the new law, the year also saw new heights in the level of protest and violence in the regime. In the weeks prior to the passage of Ley 1008, cocaleros blockaded the region's major highway for five days, temporarily occupied offices of the Secretary for Alternative Development (complete with several hostages). The most significant events took place on 27 June 1988, roughly three weeks before the law's eventual passage on 18 July. On that day, an estimated 5000 cocaleros arrived at offices of DIRECO in Villa Tunari to protest the alleged use of chemical agents in the defoliation of coca crops. After unsuccessful attempts to force a meeting over the subject, cocaleros initially occupied the offices, but were subsequently confronted by military forces of UMOPAR (allegedly under the 158  command of US Drug Enforcement Agency agents) who dispersed the assembled producers using a combination of tear gas and live rounds. The episode resulted in between five and fifteen campesino deaths, and more than a dozen injuries and arrests, and an unspecified number of disappearances (Lee 1988: 95, Salazar Ortuño et al. 2008: 142-143). Further incidents occurred on the same day in another area of the Chapare. In Ivirgarzama, UMOPAR agents broke up a public campesino meeting, killing one and wounding four in the process. In Chimoré, UMOPAR agents blocked passage of a group of cocaleros intending to travel to Cochabamba to protest the law, compelling them to return home and allegedly robbing them in the process.  Evo Morales and other cocalero leaders who rose to the fore in the early years of the conflict were what Tilly (2003) would refer to as political entrepreneurs. There is every possibility that someone as driven as Morales would have become politically active, and effectively so, under a range of circumstances; however, the presence of the coca federations decisively shaped the form and direction that that participation took in the case of the Chapare. I argue that the effect is regular, and predictable, providing an important point of reference in attempting to predict the possible outcomes of intercommunal and intrastate conflicts—it is precisely because the cocalero communities existed as parallel OCs that they were able to merge so quickly and relatively painlessly into a single coordinated political actor, with each site embedded in a single federated (what Ostrom describes as “nested”) OC governance structure.  Indeed, the degree to which conformity within the union structure was enforced was significant, even by Bolivian standards, such that the cocalero movement is described as monolithic by outsiders even today, more than a decade after assuming power in the country as a 159  whole.96 Nonetheless, that movement would never have occurred without that pre-existing institutional structure. Events would escalate and deescalate in successive waves over the subsequent 15 years, but the cocalero community remained deeply integrated over the course of that contention, bound together not only by common interests and institutions, but a shared culture and emergent identity as well. It is to these ideational considerations that I now turn. 4.3.2 The Associations: A Challenge to Cocalero Unity A new challenge to cocalero unity came at the height of the conflict with the state, in the form of a parallel set of producer organizations. Known as “associations”, these groups brought together producers in the region who had opted to grow substitute products advocated by the government and the international actors providing support, notably the US via USAID. Members of the associations were viewed with considerable suspicions by many of the cocaleros I spoke with.  They were part of the government. They supported the government, they were like intelligence for us: they knew where we were meeting, and with those associations they hit us. We have been charged because of them as well, they have testified against us. After, a compañero told me, “I did not know it was you, I’m sorry.” But the indictment continued. They were like enemies. (ER)  [The associations] were an attempt to disorganize, an idea to separate the cocaleros. That was the mission of alternative development. Alternative development never occurred. From the beginning it was a way to disorganize the organizations. (EB)  First, they generated internal conflicts. Associations created internal fights, created conflict. They even bought the conscience of leaders, gave them free land, free fertilizer. (RV).  The members of the associations themselves described this as a misunderstanding.  Before, of course, in the 90’s, there were misunderstandings. They misinterpreted us, did not look on us well because as associations evidently we needed much help in order to                                                 96 Critics at times speak of a “dictadura sindical” (syndical dictatorship) existing in the cocalero federations, an allegation cocaleros themselves are at pains to deny. See for example: “"Dictadura Sindical", el video que causó la detención del periodista Richard Romero Cossío – YouTube” Accessed [October 9 2015] 160  produce. We needed improved roads; no one here could do this, so we had to ask the government, that’s what USAID financed in the Chapare. Now yes, thanks to USAID we improved first, our road, and in terms of construction a new bridge. As we worked, we needed to take charge, and if we required a bridge and the national government could not finance it easily, through this USAID program we succeed in financing it. Accordingly, we were grateful to the USAID funding, as we would any other who appreciated the help. That seemed to bother many, who did not understand well, as some rejected the help from USAID. We in contrast received it with open arms. A help is a help; if someone brings us assistance, we’ll take it in that form. That generated some jealousy and confusion then, but in my community we did not have a problem (SM, President of an Association at the time of the interview).  Part of the animosity was due to frustration with the economics of the situation. The associations became the organizations through which funds entered the region for alternative development, and the coca federations were thus excluded from the process:  …they did receive more aid money sent from other countries via NGOs. Because of this, we received nothing. They got money, for example, for palm hearts; nobody administered it, no one audited those moneys that they [the associations] handled, it stayed that way. (GM)  Association members likewise underscored the potential for conflict and indeed, they and cocaleros reported similar instances of friction between the cocaleros and asociónistas, as the two groups developed distinct economic interests that pushed in different directions with regard to political organization and resistance. The problem most of all was a misunderstanding whenever the federation decreed roadblocks. Obviously for us as producers roadblocks did not bring us any profit; rather, they harmed us. In the case of fruit for example it has to be harvested during the week, we cannot let it keep for another day, because it will spoil. In contrast, coca harvested today can keep for a week; they had that option. (MA, banana producer)  In those years, when there were problems with the federations and all that, [my] section was very prejudiced with regard to the blockades, because almost 90% of production spoiled. Sometimes we tried to do something, but producers said at the time: “You do not support coca,” “you do not support the marches, the blockades,” but nonetheless like any member of the organizations, we are together with them, but the people understood it in another way, that we as producers were linked more to other organizations such as 161  USAID, that’s how they looked at us, and on that side where the problem was. That’s why they looked poorly on us, but we above all were working, and we excelled. (SM)  As a result, the asociónistas tried to restrict their participation in the SDCs and blockades committees, driving a further wedge between the federations and associations.   Q: And what relation did you have with the self-defence committees?  A: No, we practically marginalized ourselves on that side of things. We did not involve ourselves in that issue because we were preoccupied in selling our fruit, you see, preoccupying ourselves in other social functions. (SM)   Accordingly, those tensions broke through into intimidation and violence:   There was an epoch in which there were confrontations between them (federations and associations). The syndicates and centrals went and burned the nurseries. There were very serious problems. But then there was a change of government [to Banzer], and from there proceeded a strengthening of the syndicates, especially when Evo Morales was first elected as [Congressional] deputy. Once again, they [some association members] began to leave the associations and reinforce the unions. (Anon 3)  Yes, there were [fights] as well. Not in our community, but in others by order of the federation I believe they went and macheted crops, burned them. If some technician came to give assistance, they burned their motorcycles. (MA, association member).   SM reports experiencing pressure as well.  Q: And the pressure from the federations how was it? Social pressure? Have they burnt or destroyed produce? What form did the pressure take?  A: There had been that type of pressure, problems, persecutions of the “heads” of the producer organizations. But another thing is the lack of support, we say, in the subject of orientation, the subject of organization, a failure to organize, because we speak of a productive organization that had to be very different than a syndicate.  Q: Sure, it’s like a business association, nothing more?  A: Yes, it’s a football association: Under no pressure, it has to be voluntary. It comes out of your pocket if there’s anything you have to do, so it was with the association. One has to pay to for what one wants it to be [unintelligible]. Above all economic support, technical assistance was lacking and for that reason [many associations] have disappeared.  Q: Have there been fines from the syndicates as well? 162   A: There have been those; we continue today to pay fines. We pay, I pay when I have to miss a meeting. And if I have to pay a fine… we pay. Right now we are sanctioned.   Q: How much, more or less, do you pay?  A: Whatever they set. At times in Cochabamaba they tell you 100 Bs (about US$14). You have to pay. They are syndical decisions and they continue even now. They’ve always been there.  Q: It doesn’t matter that you’re members of the association?  A: No matter who you are, if we’re affiliated with the syndicate—I am in the syndicate, on the margin of it in the association—but all the same we must comply with the requirements of the syndicate. If I am not present actively, then a monetary fine comes and I must comply. (SM)  What is remarkable however is that while such suspicions were rife on both sides, and while it is clear that the asociónistas experienced intimidation and abuse, they continued to be active members of the community, and were expected to contribute to the activities of syndicate and federation just like any other member. Many of the most serious points of contention, notably the failure to participate in federation activities, particularly those that cut directly against asociónista economic interests like blockades, were dealt with via established syndicate and federation procedures such as fines and other sanctions. There was no great schism, and producers on both sides continued to respect the idea that one’s identity as a member of the Chapare federations—and the obligations that implied—remained valid even for those who sided with the state on such an important issue.  Eventually, tensions declined. In part this was undoubtedly due to the significant failure of the alternative development project. Though I am unable to locate hard numbers, it is clear that some producers returned to coca, while others abandoned the region entirely. Others however, such as the two asociónistsas I spoke with, persevered. The existence of the inclusive 163  institutional structure of syndicate and federation, undergirded by a durable understanding of membership in the federations being synonymous with membership in the community, a function land ownership in the region and nothing more, proves decisive in explaining how inhabitants of the region avoided greater violence or some more decisive split between cocaleros and association members.  The sequencing of events in the emergence of the cocalero federations as an OC matters in accounting for this outcome. The institutional structure of the OCs predated the emergence of coca-centric production; until the mid 1970s, it was common for all producers to grow a variety of plants. The defining principle of belonging was one’s presence in the community, not a particular form of economic or political activity. Accordingly, that inclusivity continued even after coca became the dominant crop; federations and syndicates remained open to—and indeed, insisted upon—membership and participation from all who lived in the region and benefited from shared activities. As a result, disputes among producers remained just that. Existing mechanisms and procedures could be used as necessary to manage the internal conflict that resulted from state economic intervention in the region. In a sense, the institutions and informal bonds of community linking producers in the region together proved robust enough to manage a significant break in the shared interests that had previously bound the federations together. The link between community members strained, but never broke.  4.3.3 Cocalero Culture: Winning the Fight for the Political Instrument Thus far discussion has focused on an agent-based account of the emergence of cocalero resistance, one emphasizing the role of discrete choice in the context of existing institutional and material constraints. This is, however, just one part of a larger story, one complemented and enhanced by considerations of the role of identity and culture in shaping cocalero strategies. 164  While the decision to bring the federations under a central coordinating body made good sense on rational grounds—it presented the best chance of achieving the difficult shared objective of re-legalizing the production of coca—that contingently taken decision nonetheless made possible the emergence of a subsequent, and profound transformation of identity among the region’s producers. While to some extent that transformation was “organic,” occurring as a result of the process of confrontation with the state itself, it was also in part the result of a deliberately undertaken process of socialization. To some extent, the repertoires employed by cocaleros—most notably, the nonconventional yet nonviolent methods of protest including marches, demonstrations, blockades, and hunger strikes—were similar to those seen throughout the country (Eckstein 2001). Indeed, Nash (2001) makes clear just how important ritual and cultural practice was in shaping protest in Bolivia, providing a frame of reference. Though interpreted and repurposed in different locales to suit different ends—from the highland producers resisting unwanted state incursions, to the class-conscious mining protests contesting the power of the country’s elite, to the cocaleros seeking the right to produce coca (Laserna and Villarroel 2008)—cultural practices shaped the form, intent, and even the scheduling of protest (Nash 2001). In other ways however, the cocaleros were remarkably innovative in their approach, and themselves shaped the cultural practices that they inherited. The changes in protest forms were themselves evolutionary over previous practice, but the results were revolutionary. Various examples exist. They were not the first rural indigenous group to attempt to forge a formal political presence, but they proved far more successful than others in generating widespread support for their effort (Stefanoni and Do Alto 2006, Van Cott 2003). While blockades were a common tactic in the country given the rough terrain and limited road network connecting the 165  major cities (Eckstein 2001: 11), the cocaleros were through a combination of geography and organization able to use blockades to better effect than other previous attempts. Geographically, the region featured a single highway connecting two major centres, Cochabamba and Santa Cruz, and by extension constituting one part of a cross-continental road network. With more than 100 km of highway stretching through federation-controlled territory, during the moments of confrontation it proved impossible for the state to clear away blockades faster than cocaleros could put them in place. One blockade in 2000 lasted for more than a month (Farthing and Ledebur 2004). As discussed above, the cocaleros’ organization was a vital prerequisite of that level of broad and sustained mobilization; however, a question remains how such evolution in Bolivian political practice was possible, and indeed came about in the Chapare. Why did the cocaleros as a community decide to undertake such significant and sustained levels of political mobilization, including the breaks with past practice of compromise that it implied?  One part of the answer, I argue, is the practice of regular meetings so deeply embedded in the cocalero unions. One of the remarkable elements of syndical life for cocaleros in the Chapare was the importance of attendance at these gatherings. While some were related to the everyday business of living communally, a substantial portion consisted of “seminars”: educational presentations covering a variety of topics, often focused in on the history of politics and political struggle in Bolivia, and in particular the challenges and opportunities facing them in the Chapare. In this, the syndicates emulated well-established union practices in Bolivia and beyond; indeed the importance of education regarding the “truth” of history and political struggle has been a prominent element of leftist thought dating back to at least Gramsci and his emphasis on the need to counter ideological hegemony enjoyed by the bourgeoisie in mature capitalist economies, and 166  beyond that to the early Marxian concept of false consciousness of the proletariat (Eyerman 1981). As former miner “JM”— by then relocated to El Alto — quoted in Arbona (2008: 34) put it, we got our best education in the plenary meetings and congresses – they were the university that trained us. To listen to the speeches by [Federico and Filemón] Escóbar, [Simón] Reyes, and comrade [Juan] Lechín was to get an education there and then – they were our university professors. (October 2005)   That process of education was further heightened in the tropics, where mandatory attendance and participation were both strongly enforced following the increased emphasis on solidarity that came with the increase in tensions. As Healy (1991: 89) described back in 1991, during the early years of contention, “various factors — increasing income, harsh treatment, constant threats from Bolivian law-enforcement agencies, and state efforts to reduce their coca-leaf fields — have combined to convert them into some of the most conscientious, dues-paying members of Bolivia's rural sindicatos.” Chapter Three described in some detail institutional and behavioural dimensions of that enforcement. Accordingly, the program of seminars had a commensurately greater effect, given the more faithful attendance.  Leaders looking to advance particular lines of resistance deliberately and—it would seem—successfully used the frequent meetings to bring cocaleros to their own viewpoint. As one long-time (non-cocalero) resident described it, the seminars were continuous, and played a significant role in what she described as the “catechism” of the cocaleros (Anon 2): A2: From what I have observed in my meetings, I believe that the federations have more power than the mayor and the local unions. Whatever the six federations decide, even my President does. The federations are strong. And the people have learned, they understand absolutely everything. It is not like before when they went out to march, and they did not know why they were marching; now they know why they march, they have been catechized, they have been taught.  Q: Is this through the seminars in the tropics? 167   A2: They have seminars, meetings, once a month. There are advisors there, always advisors.   Q: How long has this process of teaching gone on?  A2: Certainly, they were taught by a group, led by Evo Morales, who went everywhere to train others as syndicalistas.   David Herrera remembers the project along similar lines in this larger excerpt from a conversation quoted above:  Q: When, more or less, did these seminars begin?  A: In reality, we started the seminars around 1989 or 90.   Q: Very early.  A: Yes.  Q: Where did this idea come from? Was there a connection with the mining syndicates? or, was it…  A: Not so much. It was more an idea and thought of the leaders. We did not forget. We, for example—Evo, me, don Néstor [Bravo]—the first time we offered ourselves as candidates of the Federation of the Tropics…  Q: In ’86.  A: In ’86 here in La Costanera [a stadium in Cochabamba used for meetings], the front which we offered was called FAMA, the Frente Amplio de Masas Anti-Imperialistas [Broad Front of the Anti-Imperialist Masses]. Then came the thought that the imperialists were invading us politically, economically, socially. So, the only way to address that was to raise awareness, and to make known the reasons for the defence of Bolivians against the traitorous government of the time: the MNR, and afterwards the MIR and ADN. That was the purpose of the seminars that we undertook.  Q: Was it a useful tool?  A: It was a very useful tool, a way to sensitize people.  As for content, Herrara remembered it in these terms. 168  The seminars in the Chapare, although they existed at the level of the union, the central, and the federation, first always took place at the level of the unions because out of the bases came the new leaders, and for that we give training courses and leadership training. We explain why, for instance, we defend the coca leaf, why we defend it. Sure, now I’m speaking to you summarily, but there we explained the gifts of the coca leaf, the alternatives of the coca leaf, for how many years the coca leaf existed, why the coca leaf serves us, why—we also explain in the seminars—we defend land and territory [tierra y territorio], what use the land provides to the campesino and the to the state, what things we have to defend. One of the fundamental themes was sovereignty, that of sovereignty, because we have to depend on another state that is much more a debtor than our own, the United States.   Because they want us to submit, that is the reason why they bring us crumbs, because they bring via the subject of UN conventions cars and agreements: where they give $20 or $30 million but you have to buy an American Ford with it. Or you have to contract 10 American professionals. Those are the themes that were discussed, and always the people were sensitized in the tropics of Cochabamba because everyone ought to know why they were marching, why they were in the blockade, and why they had left work and gone out to defend. Those are the themes of our seminars.  Figures like Morales and Román Loayza97 constantly addressed member cocaleros. VG, a long-time coca leader who was still active in the executive of a central in Federation Bulo Bulo when we spoke, remembered Loayza’s words nearly 20 years later.  I’ll always remember, that it was with compañero Román Loayza, in those time he was a good leader too, one has to acknowledge, though later his mindset slipped to the other side, but he has been a good leader. He told us, “compañeros, we can no longer have them whip us because we cannot bear it. We must think on our own politics.” And we did. It was our own politics, and our own thinking. But we must always value our old leaders. It was they who thought more.  Other accounts focus more particularly on the strategic dimensions of the seminars. Filemon Escóbar, for instance, in his written account points to a more strategic focus. Escóbar, a lifelong union organizer and self-described Trotskyite, first came to national prominence as a mining leader in the immediate post revolutionary era. From the late 1970s onwards he increasingly                                                 97 A federation leader, he was among the first four deputies elected for the Izquierda Unida, a precursor to the MAS (Movimiento a Socialismo) that eventually took power with Morales as president. 169  focused on rural indigenous political organizations. In his memoir, he connects his varied efforts to bring about change electorally both to an earlier school of thought among Bolivian labour activists and to the political mobilization of the cocaleros (2008).98 By his own account, Escóbar alone undertook more than 650 seminars across the tropics, and eventually across the country, preaching a message of revolutionary change through elections. In his description of those seminars, he recalls emphasizing the history and contemporary reality of populist resistance in the country. In that, he drew out the importance of Aymaras and Quechuas not only in “direct action” (i.e. acts of public protest) but also in supporting guerrilla resistance movements in the country. He invoked the example of Guevara to illustrate how resistance—whether peaceful or violent in its approach—was destined to fail in the country without support of rural indigenous Bolivians.  The central thesis of the seminars, according to Escóbar, was the need to “vote for ourselves.” After illustrating how “neoliberal” parties—the MNR, MIR, and ADN—all prospered through cocaleros’ votes implementing changes running against the interests of cocaleros and other indigenous Bolivians with their support, he made the case