Clientelistic Warfare

Paramilitaries and the State in Colombia (1982–2007)

by Francisco Gutierrez Sanin (Author)
©2019 Monographs XII, 480 Pages


Clientelistic Warfare analyzes the relationship between Colombian paramilitaries and the State in the period 1982–2007. Despite the attention that the paramilitaries demand, due to both the magnitude of their crimes and their specifi cities, understanding the nature of this interaction has proven to be complex. They were not a homogeneous, hierarchical force, but a protean network of highly localistic coalitions and units. Based on new and extensive empirical evidence, this book shows that even in diverse circumstances there was a set of basic mechanisms that established a link between the State and paramilitary factions, which marked the trajectory of the latter. These mechanisms, in turn, were permanently mediated by political institutions and the highly clientelistic Colombian polity. Therefore, without a close reading of the Colombian clientelistic politics and statehood, it is not possible to understand the interaction between the two entities.

Table Of Contents

  • Cover
  • Title
  • Copyright
  • About the author(s)/editor(s)
  • About the book
  • This eBook can be cited
  • Contents
  • List of Figures
  • List of Tables
  • Acknowledgments
  • Introduction
  • Chapter 1 The state-paramilitary relationship puzzle
  • Chapter 2 Democracy, clientelism, violence
  • Chapter 3 The Colombian war
  • Chapter 4 Paramilitary structures
  • Chapter 5 Trajectories and crises
  • Chapter 6 The paramilitaries, the army and the police
  • Chapter 7 Informational labyrinths
  • Chapter 8 Paramilitaries in politics: Friends, factions and murderers
  • Chapter 9 Building the state from below? The take-over of the health system
  • Chapter 10 The dynamics of coercive dispossession
  • Conclusions
  • Bibliography
  • Index
  • Series index

| vii →


Figure 1: Municipalities of the Urabá region

Figure 2: Massacres by armed group

Figure 3: Polity democracy and state terrorism in Colombia (1976–2016)

Figure 4: Homicide rates per year (1955–2011)

Figure 5: Comparative homicide rate/kidnapping rate (1970–2010)

Figure 6: Puerto Boyacá, anti-subversive capital of Colombia

Figure 7: ENP 1

Figure 8: ENP 2

Figure 9: ENP 3

Figure 10: ENP 4

Figure 11: Displacement in Colombia

Figure 12: Based on Lirong Jian, Sifeng Liu, Yi Lin, 2011

| ix →


Table 1: Some of the main Colombian guerrilla groups

Table 2: Disputes between paramilitaries

Table 3: Uribista senators involved in parapolítica1

Table 4: List of municipalities

Table 5: Examples of paramilitary pacts

Table 6: Regression Magdalena

Table 7: Regression – national sample

Table 8: Propensity scores and comparison between treatment (municipalities with paramilitary presence) and control (municipalities without paramilitary presence) before and after pairing

| xi →


I am indebted to many more people than I could possibly name here. Any glaring and unjust omission is completely unintended.

During years, the institution where I work – Instituto de Estudios Políticos y Relaciones Internacionales, IEPRI, at Colombia’s Universidad Nacional—has given me a nearly ideal environment for research. Some of my colleagues there are renowned experts in the Colombian armed conflict, guerrillas and paramilitaries; I have understood tons of issues thanks to my conversations with them. Other colleagues (especially but not only the team of Father Fernán González at CINEP, which has an illustrious research trajectory, as well as investigators at the Centro Nacional de Memoria Histórica, which has done so much to understand the victimization of civilians in the midst of our conflict) have been extremely helpful. Of course, I drew heavily on the already copious national literature on paramilitaries. I am extremely proud to have worked for already a long time with Elisabeth Wood and James Putzel from whom I have learnt so much, and who read previous drafts of the text and provided crucial feedback. Last but not least, two anonymous reviewers chosen by Peter Lang – Oxford University Press also helped me to substantially improve a previous version of the manuscript. Of course, none of these colleagues, reviewers and/or dear friends are responsible – as the formula adequately goes – for the errors, inexactitudes, or misunderstandings of the phenomenon or of the relevant literature this book may have.

The book synthesizes research results built carefully throughout the years. Carefully, but not in solitude. I had the fortune to count with a fantastic team of research assistants: Jenniffer Vargas, Margarita Marín, Christian López, Diana Machuca, Silvia Becerra, among others (I am sure that once again I have incurred in very unjust omissions; I am highlighting only some of those who worked with me directly in the analysis of the paramilitary phenomenon). Also huge thanks to Robert Blaise MacLean. He and Alexandra Gómez – as well as Peter Lang – Oxford University Press editors – played a fundamental role in helping me build the final manuscript. ← xi | xii →

Social organizations, human rights NGOs, journalists, judges, concerned people, among many others, have gone to huge lengths (not to be forgotten: often sacrificing their lives) to document and denounce the phenomenon I analyze here. Transitional justice mechanisms have also played a fundamental role in this. I am enormously indebted to magistrate Uldi Teresa Jiménez of the Sala de Justicia y Paz of Bogotá, for her emphatic and decisive will to make public what is public. Also for allowing me a direct participation in different judicial processes and in an ongoing conversation with her team about the themes I have been researching. Judge Ana Joaquina Cormanes Goenaga of the Juzgado Primero Penal Especializado de Santa Marta played as well a stellar role in terms of both access and understanding.

Heartfelt thanks to my wife and family. Without their unflagging support and love I would have probably not left my bed in all these years, let alone engaged in an undertaking of this size.

| 1 →


This book analyzes the relationship between the Colombian state and the country’s paramilitary groups, an enormous and lethal constellation of local and regional counter-insurgent units that operated between 1982 and 2007. My fundamental claim is that the Colombian paramilitaries appeared, grew and developed as part of the system of indirect rule and clientelism that has regulated the links between the central state and the country’s regions, but that at the same time deeply transformed this system, triggering large-scale interaction failures: collective action issues, principal-agent structures, and misalignments of interests. I put all of these under the general rubric of social impasses.1 A substantial portion of the book is dedicated to exhibiting the concrete mechanisms through which increasing state-paramilitary convergence and collusion did not tame or tone down, but rather boosted, these social impasses. The existing relevant literature hardly mentions, let alone explains or analyzes, those impasses in their full breadth and depth. I show throughout the book that to simultaneously understand state-paramilitary collusion-convergence and social impasses, it is indispensable to take on board state institutions, the organizational structures of the paramilitaries, and the territorial coalitions on which they were grounded. This may contribute to ongoing debates in the civil wars and state literatures.

In this introduction I describe the sources on which I based my study, offer some terminological guidelines, and describe the sequence of chapters. ← 1 | 2 →

Sources and tools

Newspapers and web portals

Since the Colombian paramilitary phenomenon was so broad and massive (see Chapters 1 to 4), it drew the attention of journalists, decisions makers and activists. Some web portals, among which Verdad Abierta (<http://www.verdadabierta.com/>) is the most relevant for the type of analysis I develop here, provide rich information about Colombian paramilitaries at the national, unit, and individual levels. The portal has a solid trajectory regarding its veracity, despite dealing with very sensitive information. There are also other high-quality journalistic sources on the web which I cite throughout the book.

The main Colombian newspapers have reported regularly about the paramilitary; some created, at one moment or another, special sections about the armed conflict. I systematically use information from the two main national dailies – El Tiempo and El Espectador – and from a host of regional periodicals. There are two important weeklies – Semana2 and Cambio3 which also constituted extremely important sources.

Justicia y Paz proceedings

I have collected and digitized thousands of hours of Justicia y Paz (herein JyP) proceedings, which amount to more than more than 20,000 text files.4 This extraordinarily important empirical corpus was produced and elicited within the context of the special/transitional justice regime devised for the paramilitary demobilization, which started in 2003 and ended between ← 2 | 3 → 2006 and 2007.5 JyP explicitly included interactions with victims as part of the process. Victims and human rights lawyers have demanded more, and not less, public knowledge about the events discussed in this book.6 In any case, I have proceeded throughout under the do-not-do-harm principles, and I do not include here any citation or point of information that could possibly harm victims.7

The basic rules of JyP were the following: the paramilitaries could be assisted by a lawyer, but instead of the typical adversarial setting of formal justice, it operated on the basis of confessions and acceptance of deeds by the accused in exchange for radically reduced sentences.8 If they failed to provide the truth, they could be expelled from JyP and lose its benefits, although it is indeed clear that there was room for adorning and massaging the facts; decisions to expel members of JyP were taken rarely, and only in face of radical transgressions.9 Cross-examination and verification of facts with the victims, however imperfect, were part of the process.

Of course, not all utterances produced in the context of JyP are true. In some cases interested parties – politicians, narcos and other paramilitaries – have put pressure on the imputados to change their declarations – even threatening them – but in general this has been a relatively marginal phenomenon, restricted to events involving political big-shots or big businesses. In other situations, JyP confessions have simply been found to be false.10 ← 3 | 4 → Also, some politicians accused by paramilitaries in JyP were found innocent due to a lack of evidence.11 In general, a lack of veracity or at least precision deficits in some JyP testimonies are easily explainable. The paramilitary could have reasons to take revenge on ex-collaborators and adversaries, out of spite or because of some other motivation. Paramilitaries also not remembered everything and had imperfect memories, though sometimes they were permitted to support their confessions with documentary evidence. Certainly, the quantity of violent events they committed against civilians is so large that it is difficult to imagine how they could report all of them with absolute precision. On the other hand, JyP is a process governed by clearly established formal rules. The paramilitary confessions produced within it are not cheap talk, and they are regularly checked and frequently challenged by the Fiscalía (<http://www.fiscalia.gov.co>) – a major state agency, which has played the role of prosecutor – other state agencies, the respective JyP magistrate, and the victims. All in all, JyP’s veracity standards are much higher than those of, for example, a standard in-depth interview, let alone commonly used civil war databases.

Some of the documentary aids used by the paramilitaries were actually transferred to JyP. The Fiscalía also tried to support its accusations with hard evidence, and to systematize the confessions of the JyP paramilitary. These databases were also valuable for my research, at least as relevant context. Of course, they are still very partial convenience samples, but they can be put to use in many ways: they contain both numbers and narratives, they show that the violent paramilitary activity was much greater than what standard ← 4 | 5 → narratives and extant databases reveal, and sometimes they can be matched with other databases to produce better estimates of frequencies of violence.

Other judicial proceedings

Besides the massive JyP documentation, I used yet another large collection of judicial proceedings (more than 1500 of them)12 from the so-called juzgados de ejecución de penas where paramilitaries were tried.13 The setting here is very different and distinctly adversarial. Some victims accused, some state agents prosecuted, and the paramilitaries and their accomplices defended.14 Once again, reporting the deeds was not cheap talk.15 The proceedings are incredibly rich in their own right, but they also came with additional information:16 reports by the security agencies of the state, paramilitary materials (including correspondence) seized by the authorities, reports by NGOs and international agencies, and so on.

Furthermore, this corpus allows me to partially compensate for some regional and temporal biases contained in JyP. The main juzgados de ejecución where the proceedings were consulted and digitized were in Bogotá and Santa Marta. The juzgados in Bogotá cover all the country, and in particular the South, where several paramilitary units roamed. To document their behavior through JyP is difficult, for many reasons (they did not join JyP in the first place, or they did but the documents are hard to get, etc.). Santa Marta provided a rich coverage of several paramilitary units, ← 5 | 6 → including the Bloque Norte, a critically important one.17 And contrary to JyP, the juzgados de ejecución also cover in detail early paramilitary experiences, including the ultra-important ACDEGAM18 case.

I used other proceedings for specific chapters. For example, the congress members accused of incurring in what in the Colombian context was called parapolítica – the partnership between a politician and the paramilitaries – inform a substantial part of Chapter 8.19

In-depth interviews and paramilitary documentation

Last but not least, over a number of years I have collected a set of in-depth interviews20 with paramilitaries, demobilized or active, and of people who lived in paramilitary fiefdoms (from policemen to survivors of paramilitary attacks). The paramilitaries also produced ideological, disciplinary and administrative documents (much, much fewer than the guerrillas, ← 6 | 7 → though), which I use here. I also had the fortune to be able to regularly consult first-class Colombian experts, in the judicial and academic worlds.

In general, even after triangulating between these sources, I avoided drawing conclusions about specific characters, and whenever I had a minimum of doubts about the role they played I removed proper names of people who did not belong to the paramilitary. I assume throughout that events identified independently by privileged observers and/or actors, who had no chance to co-ordinate their versions were more likely to have happened – a Bayesian-type plausibility building mechanism (Beach and Pedersen 2013).


Some researchers assume that databases coming from convenience samples constitute hard data. In reality, they are as ambiguous and complicated in nature as any other kind of data, perhaps more (Freedman 2005, Ball 1996). At the same time, well-curated databases can indeed be very important reference points and a fundamental tool to improve evidence triangulation and mixed-methods designs. For all this to become true, it is indispensable to know, even if only approximately, the process through which the data was generated.

I use in this book the main state database on victimization of civilians (the RUV – the Unified Register of Victims), massacres and large databases generated by the Observatorio de Restitución y Derechos de Propiedad Agraria (<http://www.observatoriodetierras.org/>), and electoral data, among other sources. As stated above, I clearly establish when a result or an argument is supported by quantitative evidence, explaining how it was built or referring the reader to the corresponding publication.

Due to their massive character, there is no way of meticulously reading the entirety of my sources, let alone analyzing them. In order to solve this ‘big data syndrome’ and be able to systematically use my sources, I developed a textual data analysis tool programmed in Mathematica® – a programming language in its own right and for me a nearly ideal data-managing tool, as it is a universe within which there is no Chinese wall ← 7 | 8 → between quantitative and qualitative data –21 and in many senses tailored to work with textual corpora such as those described above. The workings of the tool allowed me to combine attentive analysis of chosen documents and first-hand knowledge of the cases I describe – which cannot be replaced – with carefully built automatic queries throughout each of these corpora.

As Guenter Lewy – speaking in another context about the use of massive textual sources on perpetrators – has said so well, ‘none of these sources, taken alone, is infallible, but taken together they provide a reliable picture’ (Lewy 2017: viii).


I try to avoid unclear or gratuitously exotic terminology throughout. It is inevitable, though, that some idiosyncratic expressions will be used, so they deserve to be clarified from the outset.

a. Paramilitaries instead of paramilitary. There were many kinds of paramilitaries in Colombia. When speaking about them generically, I will use always the plural. When speaking about all of them as a set, I will call them the ‘force’ (sometimes, simply to not repeat, the network or, for the period 1997–2002, the federation). In effect, despite the fact that they frequently shot each other, they considered themselves part of a single, counter-insurgent undertaking (see Chapter 3). Below the ‘force’, I will refer to the ‘unit’ or the ‘set-up’, that is, the basic components of the very unwieldy paramilitary network, like Blocs and Fronts.

b. From time to time I will call the paramilitaries ‘men in arms’. The presence of females as combatants in the paramilitary was negligible. In ← 8 | 9 → their demobilization it was around 7 percent,22 but only because they included not only fighters but also informants and social activists.

c. Intra-systemic actors. It would be tempting to speak about ‘illegal’ and ‘legal’ actors in the Colombian war. However, in Colombia the dividing line is not so clear, as legal actors could be massively involved in illegal activities, and the paramilitaries enjoyed a semi-legal status for long periods (Chapter 2). Thus, I speak about intra-systemic actors, that is, actors that could act as if they were totally legal. The paramilitaries were never accepted as fully intra-systemic, and an important part of their war was always to establish links with those who did enjoy that status.

d. Provision of security. Social actors, not only elites, sometimes demand security, and sometimes violence and coercion against adversaries. States and armed groups provide them. Henceforth, I will generally call the provision of security, violence and coercion simply ‘provision of security’.

e. Administrative territorial units (departments, regions and municipalities). Before the 1991 Constitution, Colombia had basically two subnational entities: the departmental level (which included also the figure of Intendencia and Comisaría for less developed territorial units) and the municipal level. The 1991 Constitution granted more importance to the municipal level, eliminated the figures of Intendencia and Comisaría – granting all of them departmental status – and created other territorial units. Presently, Colombia has 32 departments. Whenever I speak about regions I do not necessarily refer to the latter, but to collections of municipalities and/or departments which enjoyed a common identity (Bechara 2015, García 1998).

I almost always use the past tense, not because I believe that the Colombian ordeal with the private provision of violence and security has come to an end, but because my period of choice is 1982–2007. ← 9 | 10 →

The chapters

The book unfolds in the following fashion. The first chapter presents the analytical stakes and methodological challenges involved in discussing the relationship between the Colombian state and the paramilitaries. The second puts forward the basic historical and institutional context for the rest of the book, and shows how Colombia has witnessed a long-term coexistence between some key democratic institutions and highly privatized political violence. It also discusses the precedents of the paramilitary phenomenon. Chapter 3 provides the basic definitions and discusses the institutions that established a firm and stable link between the state and the paramilitaries. Chapter 4 focuses on basic paramilitary organizational patterns, identifying both commonalities and variation within the paramilitary cloud. That is, the paramilitaries were protean, but shared some common traits. Both variation and commonalities were institutionally fixed, and are indispensable to understanding the paramilitary war and the social dynamics it triggered. The fifth chapter discusses paramilitary trajectories and boom and bust cycles. Taken together, Chapters 3 to 5 show how the paramilitaries by institutional design were embedded in a system of exchange between regional demands for private security provision and central state agency sponsorship and support mediated by indirect rule and clientelism, which in turn weighed heavily on the manner in which they established their organizational structures. All this gave rise to highly violent and destabilizing trajectories.

Chapter 6 is centered on the relationships between the main armed bureaucracies of the state (army and police)23 and the paramilitaries. These relationships are at the core of the success and expansion of Colombian paramilitarism. At the same time, even they were marked by brutal social impasses. Chapter 7 considers such impasses in the light of ‘informational games’ and conflicts.

Then I turn my attention to the political system, discussing the means by which the paramilitaries were involved in different forms of political ← 10 | 11 → intermediation and clientelistic competition (Chapter 8). The mechanisms unearthed by Chapter 8 appear in pristine fashion in Chapter 9, which is dedicated to the description of the process of paramilitary take-over of the health-system in the North of the country in the later years of their bloody saga. Both chapters also show that reforms within the state changed the paramilitaries, but that these in turn changed the state. The last chapter puts all these dimensions (institutions, organizations) and types of links (via security agencies or via politicians or both) together, focusing on the phenomenon of coercive land dispossession by the paramilitaries. These three chapters analyze the interaction of the paramilitaries with regional coalitions and the political system from different angles. Such interaction was rich and variegated. The fact that the paramilitaries were based on coalitions which included regional elites and local notables provided them with a significant area of intersection with clientelistic politicians. At the same time, the ferocious competitiveness and ruthless factionalism of already established clientelistic networks was juxtaposed with the paramilitaries’ own factionalism (which was overtly homicidal), thus creating an explosive and unstable mix.

Since my core proposition establishes that the paramilitaries appeared in a concrete institutional build-up (indirect rule and clientelistic exchange), and that this made them the carriers of far-ranging social impasses, all chapters save the first one and the conclusions discuss institutions, coalitions and organizations as key explanatory factors, and social impasses as key outcomes. In all chapters I describe in detail how localistic the paramilitaries were, and how territorially grounded their relationship with the state was. In the Conclusions, I wrap up, and put forward some analytical implications.

I find no better way to close this introduction than to return again to Lewy’s words (2017): this is a sad and horrible story. May it contribute to its victims, and to truth, reparation and non-repetition.

1 Because the term “social dilemmas” has been habitually applied only to collective action issues.

2 The main media sponsor of Verdad Abierta.

3 First Cambio 16. It disappeared in 2010 after being undermined by the government.

4 Tribunal Superior de Bogotá, Sala de Justicia y Paz, Magistrada Uldi Teresa Jiménez López.

5 See Verdad Abierta (2008). ‘La desmovilización: el proceso de paz (2003–2006)’, Verdad Abierta, 11 February <https://verdadabierta.com/periodo4/>. Accessed 1 May 2018.

6 Verdad Abierta (2015). ‘Las deudas de Justicia y Paz con las víctimas’, Verdad Abierta, 9 April <https://verdadabierta.com/las-deudas-de-justicia-y-paz-con-las-victimas/>. Accessed 1 May 2018.

7 A good example of a potentially harmful citation would be self-justificatory paramilitary rhetoric, which re-victimizes the people they attacked.


XII, 480
ISBN (Hardcover)
Publication date
2019 (February)
Paramilitaries and the State in Colombia (1982–2007) Clientelistic Warfare State Clientelism Paramilitaries
Oxford, Bern, Berlin, Bruxelles, New York, Wien, 2019. XII, 480 pp.,12 fig. b/w, 8 tables

Biographical notes

Francisco Gutierrez Sanin (Author)

Francisco Gutiérrez-Sanín is a researcher at the Instituto de Estudios Políticos y Relaciones Internacionales at the Universidad Nacional de Colombia. He has a PhD in Political Science from Warsaw University. Throughout his career he has worked and published on civil war, patterns of violence against civilians, statehood and state fragility, political indicators, political parties, and clientelism. He participated in the Comisión Histórica del Conflicto y sus Víctimas summoned by the Colombian government and FARC peace table.


Title: Clientelistic Warfare
book preview page numper 1
book preview page numper 2
book preview page numper 3
book preview page numper 4
book preview page numper 5
book preview page numper 6
book preview page numper 7
book preview page numper 8
book preview page numper 9
book preview page numper 10
book preview page numper 11
book preview page numper 12
book preview page numper 13
book preview page numper 14
book preview page numper 15
book preview page numper 16
book preview page numper 17
book preview page numper 18
book preview page numper 19
book preview page numper 20
book preview page numper 21
book preview page numper 22
book preview page numper 23
book preview page numper 24
book preview page numper 25
book preview page numper 26
book preview page numper 27
book preview page numper 28
book preview page numper 29
book preview page numper 30
book preview page numper 31
book preview page numper 32
book preview page numper 33
book preview page numper 34
book preview page numper 35
book preview page numper 36
book preview page numper 37
book preview page numper 38
book preview page numper 39
book preview page numper 40
494 pages