Making Police Reform Matter in Latin America 9781685853013

Citation preview

MAKING

POLICE REFORM MATTER IN LATIN AMERICA

MAKING

POLICE REFORM MATTER IN LATIN AMERICA Mary Fran T. Malone, Lucía Dammert, and Orlando J. Pérez

Published in the United States of America in 2023 by Lynne Rienner Publishers, Inc. 1800 30th Street, Suite 314, Boulder, Colorado 80301 www.rienner.com and in the United Kingdom by Lynne Rienner Publishers, Inc. Gray’s Inn House, 127 Clerkenwell Road, London EC1 5DB www.eurospanbookstore.com/rienner © 2023 by Lynne Rienner Publishers, Inc. All rights reserved Library of Congress Cataloging-in-Publication Data Names: Malone, Mary Fran T., author. | Dammert, Lucía, author. | Pérez, Orlando J., author. Title: Making police reform matter in Latin America / by Mary Fran T. Malone, Ph.D., University of New Hampshire, Lucía Dammert, Ph.D., University of Santiago de Chile, Orlando J. Pérez, Ph.D., University of North Texas at Dallas. Description: Boulder, Colorado : Lynne Rienner Publishers, Inc., 2023. | Includes bibliographical references and index. | Summary: “Explores the scope of police reforms enacted in Latin America over the last three decades, their impact on police-society relations, and how sustainable they are proving to be in the current climate of democratic decline”— Provided by publisher. Identifiers: LCCN 2023005751 (print) | LCCN 2023005752 (ebook) | ISBN 9781685853532 (hardcover ; alk. paper) | ISBN 9781685853013 (ebook) Subjects: LCSH: Police—Latin America. | Police administration—Latin America. | Police misconduct—Latin America. Classification: LCC HV8160.5.A2 M35 2023 (print) | LCC HV8160.5.A2 (ebook) | DDC 363.2098—dc23/eng/20230522 LC record available at https://lccn.loc.gov/2023005751 LC ebook record available at https://lccn.loc.gov/2023005752 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. Printed and bound in the United States of America The paper used in this publication meets the requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48-1992. 5 4 3 2 1

Mary Malone dedicates this book to her children, Sonia “Bailey” and Eva Barth-Malone. With your kindness, generosity, and humor, you make the world a better place. I can’t wait to see where the next adventure takes us! Lucía Dammert dedicates this book to her daughter, Camila, and son, Matias. You both inspire me every day with your curiosity, your kindness, and your unwavering spirit. I am so proud of you both and grateful for the joy and love you bring to my life. Orlando J. Pérez dedicates the book to his partner of thirty years, Leyda J. Pérez, without whose love and support nothing would have been possible!

Contents

List of Tables and Figures Acknowledgments

ix xiii

1

The Challenges of Police Reform in Latin America

2

Policing and Public Security in Latin America

15

3

Chile: Too Good to Be True?

37

4

Colombia: Reform Amid Conflict and Instability

65

5

Costa Rica: Exceptionalism Under Strain

93

6

Nicaragua: A Return to Political Policing

115

7

Panama: Rebuilding Security After Invasion

143

8

Peru: The Challenges of Institutional Instability

165

9

Uruguay: Success in a Social Welfare State

193

10

What We Have Learned and Why It Matters

211

References Index About the Book

1

221 237 249

vii

Tables and Figures

Tables  2.1  3.1  3.2  4.1  5.1  6.1  6.2  6.3  7.1  7.2  7.3  8.1  8.2  8.3 10.1

Police Personnel and Public Security Resources in   Latin America Trust in Carabineros After the Estallido Social Trust in Carabineros and Attitudes Toward   Police Violence Trust in Police in Colombia Trust in the National Police in Costa Rica Comparison of Police and Public Political Attitudes in   Nicaragua Trust in Police and Willingness to Vote for Governing   Party in Nicaragua Trust in Consejos del Poder Ciudadano and Gabinetes de   la Familia, Comunidad y Vida Determinants of Trust in the National Police in   Panama Perception of Security When Near a Police Officer in Panama Level of Familiarity with and Trust of Police Known by   Respondents in Panama Fear of Crime in Peru Punitive Populism and Trust in the Police in Peru Main Result of Closer Police–Community Relationships in   Peru Support for Punitive Versus Preventive Public Security   Policies in Latin America ix

18 59 61 88 111 122 135 136 159 160 161 178 184 190 216

x

Tables and Figures

Figures 2.1 2.2 2.3 2.4 2.5 2.6 2.7 3.1 3.2 3.3 3.4 3.5 3.6 4.1 4.2 4.3 4.4 5.1 5.2 5.3 5.4 6.1 6.2 6.3 6.4 6.5 6.6 7.1 7.2 7.3 7.4 8.1 8.2 8.3 8.4 8.5 8.6 8.7

Latin American Homicide Rates per 100,000 People Latin American Homicide Rates per 100,000 People by   Country Self-Reported Victimization Rates by Country Public Fear of Crime by Country Average Trust in Police by Country Relative Trust in Police by Country Police Bribery by Country Homicide Rates per 100,000 People in Chile Fear of Crime in Chile Trust in Police and Other Political Institutions in Chile Police Bribery in Chile Trust in Carabineros and Other Institutions in Chile Crime Victims’ Evaluations of Crime Reporting Process   in Chile Homicide Rates in Colombia Trust in Police in Colombia Fear of Crime in Colombia Police Bribery in Colombia Homicide Rates in Costa Rica Fear of Crime in Costa Rica Trust in Police and Other Institutions in Costa Rica Police Bribery in Costa Rica Police Attitudes Toward Protests in Nicaragua Homicide Rates per 100,000 People in Nicaragua Fear of Crime in Nicaragua Trust in Police in Nicaragua Police Bribery in Nicaragua Average Trust in Police and Support for President Ortega Homicide Rates per 100,000 People in Panama Fear of Crime in Panama Trust in National Police in Panama Police Bribery in Panama Police Bribery in Peru Homicide Rates per 100,000 People in Peru Fear of Crime by Area in Peru Trust in Police in Peru Best Way to Fight Crime in Peru Police Perceptions of Public Security Policies in Peru Perceived Effectiveness of Police Patrols in Peru

25 27 28 30 32 33 34 50 52 54 55 57 58 69 83 85 87 101 103 109 110 123 127 129 130 133 137 155 156 157 158 173 174 177 180 182 187 188

Tables and Figures

8.8 9.1 9.2 9.3 9.4 9.5

Police Perceptions of Community Role in Public   Security in Peru Homicide Rates in Uruguay Reported Armed Robberies in Uruguay Fear of Crime in Uruguay Trust in Police in Uruguay Police Bribery in Uruguay

xi 189 201 203 204 205 207

Acknowledgments

ONE INCURS MANY DEBTS WHILE WORKING ON A BOOK; SOME ARE hard to repay, but all are worthy of acknowledgement. We are particularly grateful to Moorea Corrigan and Lynne Rienner for their patience, support, and encouragement, which helped us to navigate not just the typical research hurdles, but also the complications of the Covid-19 pandemic. We are also grateful for the hard work of our graduate assistants, particularly Patrick Baga, Maria Carrasquillo, Marina Cardoso-Vianna-Vaz, Michael Cole, Kayleigh Mitchell, and Heather Austin, who provided vital research and writing support. *  *  * Mary Malone: I would like to thank my colleagues at the University of

New Hampshire (UNH) for their support over the years. The Provost’s Office, the College of Liberal Arts, the Carsey School of Public Policy, and the Center for the Humanities provided funding at critical stages of this project. Many faculty also supported this project, particularly an amazing group of working moms who, despite their own constant juggling of work and family, never hesitate to lend a helping hand. UNH is the work home of many incredible, talented women who are dedicated to lifting each other up: Christy Beemer, Jen Borda, Marla Brettschneider, Jenni Cook, Molly Girard-Dorsey, Joan Glutting, Meghan Howey, Mary Lemcke-Stampone, Julia Rodriguez, Siobhan Senier, and Melissa Wells. In the Department of Political Science and International Affairs, I am especially inspired by the camaraderie, support, and kindness of some of the best scholars I have met: Emily Baer, Betsy Carter, Madhavi Devasher, Jeannie Sowers, and Jen Spindel. Our Friday morning outside coffee sessions were xiii

xiv

Acknowledgments

the highlight of the pandemic, even in January in New Hampshire. Jeannie Sowers has also been my comrade in arms ever since we joined the UNH faculty together twenty years ago; with Jeannie by my side, it seems much easier to take on the world. In the broader political science world, I am grateful for the ongoing support of many talented scholars, especially Dinorah Azpuru, Lisa Baglione, Abby Cordova, Jon Hiskey, Jon Hurwitz, Diana Orcés, and Stacy VanDeveer. I am also thankful for the tireless work and dedication of my coauthors, Lucía Dammert and Orlando J. Pérez. Despite many delays, they overcame obstacles and met each new challenge the pandemic threw our way. Lucía and I published our first manuscript together twenty years ago, and these twenty years of collaboration have reinforced one of my core beliefs: that work (and life) are better when surrounded by kind, generous, and creative people. This book would not have been possible without the tireless work of everyone at the Latin American Public Opinion Project (LAPOP), particularly Liz Zechmeister. LAPOP’s ongoing dedication to collecting and widely disseminating quality survey data provided the foundation for the book, as well as for the study of public opinion in Latin America as a whole. Not least, I would like to thank my family, especially my husband, Sascha Barth, for his unwavering support of this project, even when it took much longer than anticipated. I am very grateful to you all! Lucía Dammert: I would like to thank my colleagues across the University

of Santiago, Chile, and especially those in the Department of Political Studies; you have been a source of constant support in the development of all my professional and academic initiatives. I also thank and recognize the valuable contributions of the many friends who indirectly participated in the writing of this book, providing their knowledge and suggestions in each of the country cases on which they are experts. Almost twenty years of working together with Mary Malone demonstrates the importance of the continuous exchange of ideas and learning. Without the dedicated work of my coauthors, Mary and Orlando, in this latest adventure, this book would have remained unfinished. I thank you both for your patience and support. Orlando J. Pérez: I first wish to thank my coauthors, Mary Malone and

Lucía Dammert, for the invitation to travel this journey with them. The work, while arduous, was interesting and enlightening. I learned a great deal and found both Mary and Lucía to be brilliant scholars and even better human beings. Writing a book while being a full-time administrator is not easy! I thank my institutional homes for the past nine years, Millersville University of Pennsylvania and the University of North Texas at Dallas, for their support and flexibility in providing the time needed to work on this project.

1 The Challenges of Police Reform in Latin America

HEADLINES THROUGHOUT LATIN AMERICA HIGHLIGHT THE PROBlems of police abuse and misconduct. In April 2018, the Nicaraguan National Police openly attacked demonstrators, and overt, systematic police repression against dissidents and demonstrators has continued from 2018 to the present day.1 In October 2019, the Carabineros de Chile unleashed excessive force against demonstrators, leading to a month of widespread reports of police abuse and violence against protesters and detainees.2 United Nations investigators denounced the “fundamentally repressive manner” of the Carabineros’ reactions to demonstrations and called for prosecution of the army and police for human rights violations, which included unlawful killings and torture (United Nations Human Rights, Office of the High Commissioner 2019). In April 2021, the Colombian National Police targeted protesters with excessive force, leading to beatings, mass detentions, and even killings of demonstrators (Human Rights Watch 2021). Journalists chronicling the cycles of protest and police violence drew parallels to Colombia’s prior crackdowns on insurgents and paramilitaries, noting that the police treated “civilian protesters as battlefield enemies” (Turkewitz and Villamil 2021). The tragedy of these clashes between police and protesters was not just the toll they took on human rights. These episodes also undermined the legitimacy of police forces that had previously earned praise for their reforms. The Chilean and Nicaraguan forces in particular had been commended for building police–community ties, respecting citizen rights, and engaging with local communities. The Colombian police earned accolades for increasing levels of professionalism and efficiency and initiating some community-oriented initiatives in major cities. Historically, Latin American 1

2

Making Police Reform Matter in Latin America

police forces have been regarded as corrupt, inefficient, and even abusive, but over the past thirty years, police forces like those of Colombia, Nicaragua, and Chile had shed much of this reputation. Despite being once dismissed as hopeless, police reform had appeared to gain at least some traction in Latin America. This improvement came as welcome news. In the 1980s and 1990s, most Latin American countries had discarded dictatorships in favor of democracies and implemented sweeping economic reforms. Amid these transformations, observers lamented that one institution stubbornly remained mired in the past: the police (González 2020). Democratic reformers bemoaned the fact that despite efforts to modernize and professionalize the police and transform it into a fundamental part of maintaining order and serving citizens in a democracy, the police remained tied to their authoritarian traditions and structures. Historically, policing in Latin America had a political dimension, as police served to support the power of incumbent governments and control citizens. Especially in the 1970s and 1980s, military dictatorships incorporated police institutions as fuerzas de choque (repressive forces), as police (alongside the military) were often the actors who repressed political opponents and the public more broadly. Under authoritarian rule, the primary objective of the police was typically protecting the regime and its major stakeholders, and the police often relied heavily on the indiscriminate use of force to achieve this goal (Huggins 1998). Furthermore, networks of party-affiliated activists permeated police ranks, and appointments to sensitive posts were often based on political affiliation or personal affinity. In many cases, high-level political actors determined police-related decisions, including strategic and operational issues. In the 1980s and 1990s, democratization brought a change to policing practices, or at least it was supposed to (Sozzo 2016). Democratic governments ostensibly shifted the primary role of the police, limiting their use for political purposes (e.g., to repress political opponents) and deploying police primarily to ensure security, control crime, and develop public security policies (Dammert 2019). Indeed, prior to police crackdowns on protestors, Chile, Colombia, and Nicaragua had distinguished themselves for implementing major changes. Before 2019, the Chilean police had been highly regarded domestically and internationally, despite their complicity in the Pinochet dictatorship of 1973–1990 (Dammert 2012b). Chileans consistently reported that their police forces were far less likely to solicit bribes or abuse citizens’ rights compared with their counterparts in other Latin American countries (Latin American Public Opinion Project 2010). National crime rates have remained low in Chile, and homicide rates fall below the US average, on par with European countries. Chile also pioneered the inclusion of the community in its crime control policies, launching a series of initiatives to include citizens’ input in designing neighborhood-based security strategies (Frühling 2009).

The Challenges of Police Reform in Latin America

3

Nicaragua also fundamentally overhauled its institutions to orient police toward local communities. Prior to 2018, Nicaragua had spurned the mano dura (iron fist) policies of the Northern Triangle (El Salvador, Guatemala, and Honduras) and implemented holistic public security policies that sought to address the root causes of crime. Indeed, Nicaragua’s reforms and crime trends were perhaps the most surprising exception to regional problems with policing and crime. Nicaraguan history is marked by endemic poverty and political violence, just like its neighbors of the Northern Triangle; however, its crime rates are comparable to its more peaceful and prosperous neighbor to the south, Costa Rica. Colombia reported successes in professionalizing its police, and some observers even touted the Colombian model as a candidate for exportation. Advocates of the Colombian public security model pointed to the dramatic security transformations in the country over twenty years, particularly in the capital city of Bogotá (Moncada 2009). In the mid-1990s, Colombia was infamous as the murder capital of the world, and police and justice institutions were widely disparaged as hopelessly corrupt (Esparza 2022). By 2008, public security in Colombia had improved so much that the tourism board launched a new campaign with the ad “Colombia, the only risk is wanting to stay.” In 2017, Colombia reported its lowest homicide rate in forty years. For those looking for a quick fix to the contemporary Central American security crisis, the Colombian public security model seemed to hold a lot of promise (Shifter 2012). In other countries, smaller reforms reshaped the police. Ecuador launched a series of technological advances, such as the creation of ECU911, which led to the decrease in the incidence of crime. Uruguay experienced some institutional transformation and modernization of police work, and Costa Rica initiated a new system of community police formation. In Honduras, a country with some of the highest violent crime rates in the world, the government launched a new system of police education, which aimed to professionalize the police force and use other crime-fighting tools besides strict repression (Ungar 2015). Panama invited Chilean consultants to inform the implementation of a civilian community policing model in cities, aiming to integrate police officers more cohesively into the communities they serve. Panamanian reformers also sought to increase professional standards among police by raising salaries and improving training. Civil society groups tie these reforms to improved professionalism on the streets and a reduction in bribe solicitation from average citizens in urban areas.3 Although there have been plenty of setbacks, such as the excessive use of violence against protesters, it is undeniable that the police institutions of the 2020s are very different than their predecessors of the 1990s. Institutional and administrative transformations have professionalized police careers, raised salaries, and developed systems of social protection for officers. These improvements are matched with raised educational prerequisites

4

Making Police Reform Matter in Latin America

for candidates entering the police force, as well as ongoing institutionalized processes of evaluation, both of which help promote transparency and accountability (Tello 2012). Reformers developed diverse strategies to improve the design and implementation of police work (both preventative and of control) and evidence-based crime control strategies (Sherman 2013). There are some mechanisms for a public accounting of police action to citizens (González 2016; Muggah 2017; Ungar 2019). Perhaps most important, there has been widespread acknowledgment that the police should establish (and deepen) a strong, constant, active, and positive relationship with citizens. Most police institutions in the region have created community policing programs, and while the results are often mixed, such efforts are a step forward (Dammert 2019). Despite these gains, persistent and overt police misconduct, particularly in countries that had appeared to engage in serious reforms, gives observers pause. The recent high-profile examples of police abuse seemed to erase prior gains and raised questions about the sustainability of police reform. What does it take for successful reforms to be implemented, and how can they be sustained? To evaluate the emergence and sustainability of police reforms, we trace the process of reform and police–society relationships in several cases that had reported some successes by the early twenty-first century: Chile, Colombia, Nicaragua, Costa Rica, Panama, and Uruguay. We also assess reform efforts in Peru, where best practices were introduced but never gained traction. Peru is an understudied case for analyses of public security, and we can learn from some of the stifled reform efforts. The turn of the millennium witnessed the emergence of best policing practices, and we can learn a great deal from some initial successes as well as the inevitable pitfalls. We aim to synthesize what we can learn from the emergence of these best practices and their setbacks. To do so, we follow three steps. First, we identify and describe the best practices of police reform that emerged in the region since the 1990s. Second, we trace the process of creating and implementing these best practices in individual cases, noting the contextual factors (particularly levels of crime and social welfare) that facilitated success and those that thwarted reform efforts. We focus heavily on police–society relations and evaluate reform efforts at least partially on their ability to boost public trust in the police. Finally, we assess the sustainability of these best practices. The Colombian, Nicaraguan, and Chilean cases all illustrate the importance of sustainability and the tragic consequences for human rights when hard-earned gains are undermined. Other cases illustrate the problems of sustainability in more subtle ways. For example, Costa Rica was once highlighted as a public security success story, but contemporary trends have tarnished this reputation. Public trust in the police and in the justice system more broadly dipped sharply in the 2010s, and Costa Rican officials themselves have lamented that current crime trends

The Challenges of Police Reform in Latin America

5

have left their justice system sobre cargado (overburdened). Costa Ricans register high levels of public insecurity in national surveys, data corroborated by very visible investments in private security throughout society. To date, it is not clear whether these recent developments will undermine the Costa Rican success story. Do these trends reveal previously overlooked flaws? Or are they merely small setbacks in a longer history of success? This book examines the process of police reform and police–society relations in two understudied cases: Uruguay and Panama. Police reforms in these countries have received less fanfare, despite the fact that the reforms have registered steady and tangible improvements in terms of professionalism and linkages to local communities. In a final case study, we highlight another understudied case, Peru, in which best practices were introduced but failed to flourish.

How Do Best Practices Emerge, and How Can They Be Sustained? How can we explain why some Latin American police forces have successfully reformed, while others have not? How do we determine whether these best practices are sustainable in the long run? We draw on insights from the literatures of path dependency and police legitimacy to examine the institutional and micro-level factors that can shape the outcomes of the reform process. We argue that a modified path dependency best explains the institutional process by which policy reforms are adopted, perpetuated, and revised. At the micro level, we argue that these reforms can increase the legitimacy of the police when they improve police effectiveness, reduce corruption, and uphold human rights. When police reforms increase the legitimacy of police in the eyes of the public, public security policies that are more community-oriented, less repressive, and more preventative become more viable. If the public and major stakeholders perceive public security policies as successful, they become more firmly entrenched and provide fertile ground for additional reforms. The (Modified) Path Dependency of Police Reforms: Understanding the Institutional Challenges In most regimes, structural and historical conditions determine policy choices regarding policing and crime prevention. The type of regime (electoral democracies versus military-led governments), crime levels, economic development, and inequality (among other factors) lay the foundations for what is possible in the policy arena. Path dependency argues that institutional choices are mediated by structural conditions until a “critical juncture” (or

6

Making Police Reform Matter in Latin America

a moment of choice) forces policymakers to choose a different path, which then shapes the underlying conditions.4 The choices made during critical junctures establish the institutional patterns that endure over time. As those patterns then shape policymakers’ decisions and behavior, they influence political and social outcomes, which in turn could develop into new critical junctures that undermine the stability of the prevailing institutional patterns, potentially precipitating further policy changes. Authoritarianism and militarism shaped the origins and evolution of police institutions in the cases we analyze. In each country, police institutions emerged out of democratization processes, which were shaped by key critical junctures such as civil wars, military invasions, high rates of violent crime, electoral defeats, or geopolitical pressures. Fragile and often weak democracies led to underresourced, corrupt, and weak police institutions, which led to increasing crime and lack of institutional legitimacy. Crime and weak legitimacy precipitated an additional period of change in which policymakers were forced to adopt institutional changes to strengthen policing practices by increasing funding, cracking down on corruption, separating military and police functions, and adopting community policing strategies that aimed to reduce the social gap between police officers and citizens. At the institutional level, we argue that two key factors explain how successful police reforms emerge after these critical junctures: institutional stability and sustained political will. Institutional stability is critical, as the reform process requires continuity to see proposed changes through the design phase to implementation and evaluation. Sustained political will, preferably across political parties, is also essential given the scope and complexity of the reform process. Typically, reforms encounter significant political and institutional resistance. As González (2020) explains, politicians are often hesitant to engage in police reform because they fear it can be electorally disadvantageous, and they do not want to antagonize the powerful police bureaucracy. Building political consensus on the need for police reforms is essential for overcoming such obstacles. If a critical body of political actors does not advocate for reforms, or worse impedes them, even the best-designed reforms will wither and face the risk of backsliding. Police Legitimacy at the Micro Level We argue that when police reforms lead to improvements in police performance (e.g., effectiveness, fairness, levels of corruption, respect for human rights and civil liberties), the public finds reason to trust the police, creating incentives for public security policies that emphasize prevention and community orientation over repression and militarization. Alternatively, if critical junctures and institutional choices lead away from effective reforms and fail to improve on problems of effectiveness, corruption, and abuse, public trust will remain low.

The Challenges of Police Reform in Latin America

7

The literature on police legitimacy illuminates the relationship between police performance and legitimacy. As Lum and Nagin (2017) argue, in democracies citizen trust in police matters in its own right and is a critical performance metric. Police must ensure public safety, but they must also garner citizen trust. Public trust is essential, for if citizens view the police as corrupt, inefficient, or abusive, they will be reluctant to turn to police for protection or to solve problems in their communities or homes (Sunshine and Tyler 2003; Tankebe 2013).5 As Tyler (2004) emphasizes, the police need the support and cooperation of the public to fulfill their central role of maintaining order, and this support and cooperation are contingent on whether the public views the police as a legitimate actor. Many studies on police legitimacy highlight the importance of procedural legitimacy, focusing heavily on how fairly people perceive they are treated (e.g., Tyler 2004). According to this line of research, if the police treat citizens fairly and equally, citizens will trust the police and support them, even if they disagree with the ultimate outcome. More important than the results of an individual case, if people perceive the process and the treatment they received as fair and according to the law, they have a greater tendency to accept the outcome (Tyler 1990). Although most of this procedural justice literature focuses on the United States and Europe, which experience very different levels and types of crime, the few studies conducted in developing countries have found support for this school of thought in cases as disparate as El Salvador (Cruz 2015), Argentina (Haas et al. 2015), India (Madan and Nalla 2015), Nigeria (Akinlabi and Murphy 2017), and the Caribbean (Johnson, Maguire, and Kuhns 2014). Even after taking into account the particularities of these different cases, especially in terms of levels of insecurity and corruption, there are indications that procedural justice matters a great deal for police legitimacy. Another strand of literature highlights the importance of results— especially in countries facing a crisis in security. This perspective argues that what matters most to people, especially those facing unprecedented levels of personal insecurity, is the effectiveness of police—that is, their ability to achieve goals such as controlling crime and apprehending suspected criminals. For example, Tankebe’s (2009) work in Ghana finds that public cooperation with the police depends heavily on their perceptions of police effectiveness in fighting crime. Indeed, several studies have noted that crime victimization can reduce public trust in police (Corbacho, Philipp and Ruiz-Vega 2015; Malone and Dammert 2020). In Latin America, there is often limited confidence that the police will prevent or control crime. This perception is not far from reality, as several indicators highlight ongoing challenges in this area. For example, if a suspect is not detained during the commission of a crime, the crime often goes unpunished. Criminal investigation capacities of police personnel are very limited, and although there is increasing investment in technology and training, results

8

Making Police Reform Matter in Latin America

are still modest. Citizens often do not report crimes because they feel the police (and the general criminal justice system) will not do anything. Police presence on the streets is limited, and in most countries police ride in cars instead of patrol on foot, further eroding the possibility of establishing a relationship with the community. Highly centralized institutions are often a bureaucratic nightmare, leaving comisarias (precincts) with limited infrastructure and a general sense of precariousness. The literature focusing specifically on the legitimacy of Latin American police forces emphasizes the impact of corruption and police misconduct on police legitimacy. Indeed, systemic corruption has been a leading factor weakening public trust in police and inhibiting reform efforts. Corrupt practices have taken a toll on public trust in police. These corrupt practices range from daily instances of soliciting bribes from citizens to structural institutional decisions tainted by irregular practices. Tankebe (2010) finds that even vicarious experiences of police corruption, such as witnessing the police solicit a bribe from other citizens, can shape individual perceptions of police legitimacy. In some countries, corruption is a daily element of police work and almost recognized as a regular “bureaucratic” practice. To improve the standing of police, reformers have long aimed to stamp out acts of petty and institutional corruption. For example, in Peru, during the government of Alberto Fujimori (1990–2000), the transit police underwent a complete overhaul, transforming into an almost all-woman force to disrupt deeply ingrained, regular corrupt practices. In Argentina, the police of the city of Buenos Aires was created in 2016 to ensure quality of service and better police capacity, but also to limit corrupt practices of other police institutions, such as the Federal Police and the police of the province of Buenos Aires. The police on the street are not the only “bad apples,” of course; there are structural problems that have consolidated a culture of illegality. Institutional corruption is widespread and includes political actors that take a piece of general contracts, especially those with links to infrastructural needs, such as cars, uniforms, or gasoline. Even those cases considered by the literature as the very best practices have shown severe problems of corruption. As we explain in the case studies of Colombia and Chile, even when street-level bribes are less common, institutional corruption can still flourish. Unfortunately, corruption is a problem to which no police force is immune in Latin America. As a response, purge of police personnel has become a regular practice in countries as disparate as Colombia and Honduras, the latter of which expelled almost half the National Police in 2017– 2018. In Mexico, one of the constant elements of police reform initiatives is implementing “trust measures” that include all types of tests to ensure proper police conduct; for example, all police officers were obliged to take a polygraph test starting with the Calderón administration (2006–2012).

The Challenges of Police Reform in Latin America

9

In addition to corruption, police respect of human rights and civil liberties is a key determinant of police legitimacy and an area targeted by many reform efforts. Several Latin American police forces have earned infamous reputations for excessive use of violence; in the world of social media, these abuses are widely disseminated and potentially have an even greater impact on citizens. Even though Latin Americans clamor for more control, more police presence, and more punishment for criminals, these highly visible signs of police excess are not welcome. In fact, many policerelated scandals are rooted in regular police practices in which the use of violence is the norm rather than the exception.6 The Brazilian case is perhaps the most notable, since more than 33,000 people were killed in police shootings between 2010 and 2020, and more than 9,000 in Rio de Janeiro during this same period (HRW 2020b). As this trend continues to increase, civil society organizations and international nongovernmental institutions have pressed for changes and reforms. Brazil may be the most notorious case, but by no means is it an isolated one, as there are judicial processes against police for their participation in homicides, injuries, and torture in several countries. In sum, the literature on police legitimacy identifies several factors that shape levels of public trust in police. This trust is often linked to perceptions that the police treat citizens fairly and equally (procedural justice), are effective in controlling crime and responding to citizens’ needs, do not engage in bribery, and respect citizens’ human rights and civil liberties. To be sure, these perspectives are not necessarily mutually exclusive, as police legitimacy can depend on all these factors. Socioeconomic Context Modified path dependency can help us understand the institutional environment in which the reform process unfolds, and the police legitimacy literature illuminates how these reforms can culminate into more positive attitudes toward police at the micro level. Police reforms do not occur in a vacuum, of course, so it is important to consider the larger socioeconomic and historical context in which these reforms have unfolded. In areas where racial, ethnic, and or other socioeconomic divides are sharply pronounced, the population might want different things from the police (González 2020). Indeed, in many Latin American countries, police often act as “border guards,” shielding the middle and upper classes of society from the lower classes. As community-oriented policing reforms were proliferating around the region, Chevigny (1999) cautioned that particularly in poor neighborhoods, police officers often had a “shoot first, ask questions later” approach that led to the arbitrary use of excessive and deadly violence. In countries with high levels of inequality, police frequently do not view themselves as

10

Making Police Reform Matter in Latin America

part of the communities they protect and are trained to subdue citizens rather than protect them. Police force is yielded almost exclusively against the lower classes. Such misconduct is reinforced by the courts, which often decline to prosecute misuses of power and even ignore evidence of police misbehavior when pursuing convictions against defendants (Ungar 2002, 84). Because excessive uses of force are usually targeted against citizens of lower socioeconomic status, there is acceptance of these practices by both officials and citizens from other socioeconomic classes. Indeed, this acceptance has been heightened by rising levels of fear of crime (Rosen, Cutrona, and Lindquist 2023). In countries with sharp racial or ethnic divisions, members of the dominant group might expect the police to act as border guards between racial groups, thinking of calls to the police in the same vein as “calling the manager.” In these settings, people in the racial minority might be slow to call the police, if they do so at all, anticipating that the arrival of the police will have a high potential to lead to force, violence, or other sanctions against them. In such cases, people of a racial, ethnic, or economically marginalized group might only call the police as a last resort, when something occurs beyond the capacity of informal networks to handle. Such stark racial and ethnic divisions, which are almost always accompanied by socioeconomic divides, can lead police reforms to replicate prior patterns of racial and ethnic conflicts. Amar (2010) notes that even when police reforms prioritize diversity and include members from historically marginalized groups in policing and public security debates, they still run the risk of reproducing earlier trends of conflict and police repression if they do not simultaneously address structural inequalities, which are often reinforced by racial, ethnic, and economic divisions. If reforms merely paper over historical legacies of racial, ethnic, and income inequality, they are unlikely to fundamentally alter how police interact with citizens from all classes and racial groups. In contrast, in areas in which sharp socioeconomic, racial, or ethnic cleavages are absent or minimal, there is likely more uniformity in what the public expects of police. Lower levels of inequality can lead the public to have similar expectations of police and make similar demands, and they may facilitate police responses that engage the community as a whole rather than privilege one group over another. In a comparative analysis of local police reforms in Brazil and Honduras, Arias and Ungar (2009) explain the importance of socioeconomic context, noting that community-oriented policing reforms tend to succeed when societal incorporation is matched with political commitment and police cooperation. They argue that community-oriented initiatives need to engage citizens and must create the space and conditions for citizens to mediate with local officials and the police. High levels of inequality and discrimination can marginalize some parts of the citizenry and keep them from having a voice in public security policy. In an analysis of community-

The Challenges of Police Reform in Latin America

11

oriented policing programs at the local level, Arias and Ungar (2009, 425) find that one key reason some programs ran more smoothly and gained traction was due to their ability to combine policing reforms with investments in social programs. When sustained in the long run, such social programs can begin to ameliorate socioeconomic cleavages and allow citizens to have a voice in public policy on more equal footing. In the short run, investment in social programs can translate to more resources for underprivileged groups and reduce dependence on illicit activities for basic livelihoods. We explore how this broader socioeconomic context has shaped the reform process in our cases. We provide a qualitative historical overview of the reform process in each country and note the ways the reform process overcame (or was stymied by) historical legacies of inequality. For example, in our studies of Costa Rica and Uruguay, we note that the processes of police reform were set against the backdrop of a social welfare state. In Nicaragua and at the subnational level of Colombia, the reform process was buttressed by cross-cutting investments in social welfare programs. We argue that in countries with higher levels of inequality and/or racial/ethnic cleavages, police reforms have a greater likelihood of success when they are accompanied by these investments in social welfare and human capital. When investments in justice reform are matched with those of social protection, these reforms can bolster one another and provide a more holistic response to addressing crime. Public Security Policy Options and Preferences Modified path dependency helps explain why different institutional configurations of reform emerge, and the literature on police legitimacy helps us understand how these reforms can shape the micro-level perceptions of citizens. Both institutional reforms and public perceptions condition public security policy options. When police performance is poor and legitimacy is low, the number of viable public security policy options narrows considerably. That is, if police forces are poorly trained, are poorly paid, have little oversight, and are viewed as illegitimate in the eyes of the public, it is difficult to expect such forces to implement a holistic community-oriented policing strategy. Instead, lackluster police performance and public mistrust can create incentives among politicians and citizens to embrace repressive, mano dura policies, particularly when crime rates are high. When such repressive policies fail (as they overwhelmingly do), lack of trust in the police leads the public and politicians to seek alternative actors (e.g., private security forces or the military) or endorse increasingly draconian crime-fighting measures. Extralegal options can become more appealing, leading citizens to turn to other actors (ranging from private security guards to vigilante groups) to mete out justice or provide adequate protection. Furthermore, when political elites doubt the efficacy of their own police forces, they tend

12

Making Police Reform Matter in Latin America

to turn to alternative actors.7 Policymakers justify these choices by arguing that citizens prioritize results: the public wants to confront criminality effectively and decisively and is less interested in how these results are achieved. Despite the frequency of such rhetoric, in most cases these alternative actors have proven to be no better than police in controlling crime, and they create new problems because they lack adequate training to interact with civilian populations and they elude legal mechanisms for accountability and oversight (Ungar 2007; Malone 2012). Under this scenario, it is not uncommon for countries to experience spirals of increasing militarization and deteriorating public security. When police reforms flounder and do not lead to better outcomes in these areas or fail to garner public trust, critical junctures can emerge where countries can “fall out” of this virtuous circle and become mired in an undesirable status quo/equilibrium. Thus, to avoid the problems posed by military, extralegal, or private actors, it is imperative to identify how professional civilian police forces can connect with the communities they are supposed to serve. In contrast, when public trust in the police increases, public security policies that are more community-oriented, less repressive, and more preventative become more viable. When the police are regarded as legitimate actors, policymakers have more public security options; higher levels of trust create incentives for different types of public security policies. If these policies are viewed as successful in the eyes of the public and major stakeholders, they become more firmly entrenched and provide fertile ground for additional reforms to promote further improvement. Bailey (2009) articulates a similar dynamic between crime and democratic governance. He argues that crime often weakens the institutional and societal foundations of democratic governance, and weaker democracies find themselves ill equipped to maintain security in their borders. Increasing crime and violence can induce a moment of inflection for policymakers in which critical choices can either break or perpetuate the vicious circle. If the cycle becomes self-reinforcing, new democracies become ensnared in a security trap. We argue that when reforms fail to rectify poor police performance (e.g., effectiveness, levels of corruption, respect for human rights and civil liberties), this leads to low levels of public trust, and low levels of public trust in turn lead to limited public security policy options. In this book, we identify the institutional and structural factors that shaped policing and examine the key policy choices made by elites during critical moments of regime stress. In particular, we focus on the importance of institutional stability, sustained political will, and cross-cutting investments in human capital. We identify the places in which best practices emerged and trace the process of their creation and implementation. We then turn to the micro level and explain how successful police reforms can foster public trust, particularly when the public perceives police as effective, noncorrupt, and respectful of their human rights. When police reforms

The Challenges of Police Reform in Latin America

13

succeed at both the institutional and micro levels, nonrepressive public security policies become more viable. In sum, we argue that successful police reforms foster higher levels of public trust, which in turn are less conducive to repressive public security policies (i.e., mano dura).

Plan of the Book In the pages that follow, we study the context in which national police institutions emerged, analyze the critical junctures that shaped early policy choices, discuss the outcomes of such decisions, identify best practices, trace their trajectories of creation and implementation, and then assess their sustainability. We focus on Chile, Colombia, Costa Rica, Nicaragua, Panama, and Uruguay. We also highlight the case of Peru, in which best practices were introduced but failed to flourish. To contextualize these case studies, Chapter 2 provides an overview of the public security sector in Latin America, as well as the high crime context in which the police sector operates and these policing practices and reforms unfolded. We identify and examine cases in which police reform made critical advances and highlight a series of best practices that have emerged. We analyze how these best practices took root and transformed authoritarian policing practices and police–society relationships. We then assess long-term sustainability. Indeed, it is important to recognize that police behavior is far from static. As Wilson (1968) explains, the way police engage in their work can change quite rapidly. Major events can quickly erode public trust in police, as was the case in Chile in 2017, when corruption scandals quickly eroded prior widespread support for the Carabineros. In these country case studies, some best practices have emerged that have improved police performance and strengthened police–community relationships. However, as crackdowns against protesters illustrate, in some countries there remain serious questions about the sustainability of these practices. Repression and human rights abuses often serve as policy choices during critical junctures in which civil society challenges state authority. The decision to repress undermines police trust and legitimacy and thus the sustainability of best practices. This book addresses such questions of sustainability to determine what factors can lead police reforms to backslide. We identify the challenges that have emerged to keep hard-won reforms in place and discuss implications for sustainability. In our concluding chapter we empirically test the linkage between trust in police and public support for different types of public security practices. We examine how low levels of trust in police can shape public willingness to support punitive public security measures over those prioritizing prevention. Finally, it is important to note that most of the research for this book was conducted prior to the Covid-19 pandemic. Our longitudinal analyses offer insight into durable regional and national trends; however, we recognize

14

Making Police Reform Matter in Latin America

that the extreme circumstances of the pandemic altered the performance of all state institutions, including the police. The pandemic also affected the linkages between state and society, in some cases culminating in political crises (e.g., Peru). As of early 2023, it is difficult to discern which pandemicinduced changes are temporary and which will endure. This book provides a comprehensive overview of key developments and trends from the 1990s through 2021; the postpandemic years will offer additional opportunities to reexamine the sustainability of police reforms against a backdrop of such an extreme public health crisis.

Notes 1. According to the Comisión Interamericana de Derechos Humanos (CIDH) and the Asociación Nicaragüense Pro-Derechos Humanos (ANPDH), approximately 300 people were killed in violent clashes with the police between April and August 2018, more than 2,000 people were injured, and more than 2,000 arrested (many in arbitrary fashion or kidnapped by security forces) (Baldizon 2018; CIDH 2018) 2. Human Rights Watch (HRW) estimated that over 11,000 Chileans were injured during police clashes with protesters between October 18 and November 22, 2020, and reported that the attorney general’s office in Chile investigated 26 deaths. The Carabineros arrested more than 15,000 people, and there were widespread reports of police abuse of detainees (HRW 2019b). 3. Author interviews with leaders of community organizations and civic groups in Panama City, July 2014. Civil society groups noted that while community-oriented public security practices had taken root in urban areas, in rural zones (particularly in areas with high rates of illicit trade), policing practices were increasingly militarized, and human rights abuses continue to be prevalent. 4. For examples of the application of path dependency, see Collier and Collier (1991), Mahoney and Snyder (1999), Mahoney (2000, 2001), and Boas (2007). 5. Jackson et al. (2013) argue that police legitimacy can also shape public attitudes on the acceptability of violence. In their survey of young men from ethnic minority communities in London, they find that procedural justice is linked to police legitimacy and higher levels of police legitimacy are correlated with negative views on the use of violence. They argue that when the police are perceived as rightfully monopolizing force, their legitimacy “crowds out” the acceptability of the use of violence by other actors. 6. The degree of scandal and public outcry tend to depend quite heavily on the socioeconomic status of the victims of police violence. Fernández Roich (2017), for example describes media and public outrage when police use extensive violence against members of the middle class, in contrast to those from other socioeconomic groups. 7. For example, Mexican leaders cited police corruption and inefficiency as reasons to enlist the military in their campaign against organized crime, and El Salvador has dispatched its military to patrol the streets despite peace accords forbidding this practice. In Guatemala, paramilitary groups are linked to political and economic elites.

2 Policing and Public Security in Latin America

TO CONTEXTUALIZE THE COUNTRY CASE STUDIES, THIS CHAPTER provides a regional overview of the public security sector in Latin America, as well as a review of the high crime context in which this sector operates and police reforms have unfolded. We focus on one particularly important development: the design and implementation of community-oriented policing practices. Since the 1990s, most countries have implemented at least some reforms that have aimed to integrate police into the communities they serve, with the goal of fostering better community–police relationships. In this chapter, we provide a regional overview of public attitudes toward the police and introduce several of the survey questions we use to assess public perceptions of police performance. Although we consult multiple sources of public opinion data, our analyses rely heavily on the Latin American Public Opinion Project (LAPOP), which has collected high-quality public opinion data throughout the region since the early 1990s. LAPOP’s surveys include questions related to trust in police, experiences with and perceptions of police work, and perceptions and experiences with crime. The identical wording of questions across countries allows for ready comparisons among our cases and over time.1

The Latin American Public Security Sector The public security sector in Latin America consists of a wide variety of police institutions with some common traits but multiple elements of differentiation. The most obvious difference is that four countries (Argentina, Brazil, Venezuela, Mexico) have federal systems of government and organize 15

16

Making Police Reform Matter in Latin America

their police forces at the national, regional, and in some cases local levels. In those countries, the mandate of national police institutions has traditionally targeted complex crimes, such as drug trafficking and money laundering. For much of the 1990s, these national forces were small in personnel and had limited resources.2 The other Latin American countries are unitary states and have one national police force or multiple national forces with areas of specialization. For example, in Chile, two police forces have operated throughout the country, one in charge of prevention (the Carabineros) and the other in charge of criminal investigation (Policía de Investigaciones). There are also differences in the missions of police forces. Some are in charge of both prevention and criminal investigation, such as the national police forces of Honduras, Nicaragua, Peru, and Uruguay. In other countries, such as Chile and Mexico, specialized police forces focus on criminal investigation. There are also institutions with more specific missions, such as the Policía Bancaria e Industrial (Banking and Industrial Police) in Mexico, which provides protection for private companies through services such as personal bodyguards and secure transit of assets and merchandise. These differences aside, there are three common elements in the public security sector in Latin America: militarization of public security, new actors involved in providing security, and increasing budgets. The militarization of public security has intensified as crime rates have soared, and politicians and citizens have often regarded their current police forces as inadequate responders to the security crisis. Pressure for more militarized approaches to policing is most pronounced in countries affected by the drug trade, and the corresponding US-inspired war on drugs.3 There is a strongly entrenched narrative of police dysfunctionality and ineffectiveness, paving the way for military solutions to police problems (Pion-Berlin and Carreras 2017). The role of military personnel in public security activities has increased substantially, becoming the norm in several countries in the region (Flores-Macías and Zarkin 2021). Mexico is perhaps the most obvious case of military involvement in domestic security, as the military engages in police-type patrolling in areas where criminal organizations have a strong presence. Mexico is not the only country that readily dispatches its military to patrol the streets, however. Several countries have involved military personnel in areas such as street patrolling, responses to specific threats, and border control (e.g., Peru, Argentina, Guatemala, Brazil, Paraguay). Although this strategy is widely employed, it is not clear that the military are more capable of curbing crime, and when the military engages in such missions, a trail of human rights violations tends to follow them (Brewer 2009). The second shared element among police institutions is the presence of new actors on the public security scene. Besides the involvement of armed forces, all countries have witnessed exponential growth in the private secu-

Policing and Public Security in Latin America

17

rity industry, which often has more personnel and even weapons than the police. In Guatemala, Honduras, and Peru, private security personnel can carry weapons to protect citizens or businesses. In Argentina, Uruguay, and Peru, police officers can work as private security officers for part of the week. The lack of proper state regulation allows private companies to engage with situations more appropriately addressed by governmental institutions, without adequate oversight (Ungar 2007). Other actors in the public security panorama are local governments keen to increase their involvement in crime prevention (e.g., Chile, Colombia, Uruguay, Costa Rica, Guatemala) and crime control mechanisms (e.g., Peru, Bolivia, Brazil, Mexico). There is ample evidence that crime prevention is best implemented at the local level, especially where protection of public spaces is involved. However, there is a general temptation to view “municipal police” institutions as quick fixes to fill the void of legitimacy left by their national counterparts.4 Finally, the third common element is increasing budgets. Police officers are no longer simply precarious public servants with low salaries and limited social protection benefits. Although there are differences among countries, there is a clear trend toward more investment in police personnel and infrastructure (Casas, González, and Mesías 2018). As the country chapters will show, this investment is not always wisely spent, which is an indicator of the limited transparency and accountability of the sector rather than a limited state commitment to invest. Table 2.1 provides an overview of police personnel and resources in the region and includes data from the United States and Canada for comparative purposes. As Table 2.1 indicates, there is a considerable amount of variation in terms of resources available for policing by country. Overall, the number of police personnel per 100,000 is highest in Argentina and Uruguay, but most countries retain a police force ranging between 250 and 400 per 100,000 people. Compared with the United States and Canada, most Latin American countries employ far more police personnel. Only in Guatemala, Honduras, and Nicaragua does the rate fall below 200 police per 100,000. More personnel do not necessarily translate into more police on the streets, however, because of the increasing size of police bureaucracies. Table 2.1 also illustrates that there is a wide range of resources devoted to domestic policing in each country. Panama reports the highest percentage of GDP invested in policing and domestic security, while Bolivia had the smallest budget as a percentage of GDP. In terms of average monthly salaries, there is a great deal of regional variation. Table 2.1 shows average monthly salaries for all police personnel, although in some cases the average is inflated by higher salaries for higher ranking officers (e.g., Ecuador). In Mexico, average police data can be misleading, as subnational entities vary tremendously in their levels of

18 Table 2.1

Country Argentina Bolivia Brazil Chile Canada Colombia Costa Rica Ecuador El Salvador Guatemala Honduras Mexico Nicaragua Panama Paraguay Peru United States Uruguay Venezuela

Police Personnel and Public Security Resources in Latin America Police Personnel per 100,000a

Percentage Female Policeb

Average Monthly Salary in US$c

810 352 266 315 188 368 278 273 365 194 160 356 168 442 243 383 206 681 Data not available

Data not available 11.1 14.9 18.3 21.4 9.4 15.6 10.1 12.1 13.2 18.6 19.6 27.2 13.4 8.0 15.1 12.5 23.5 Data not available

801 285 860 874 $3,355 486 481 1,002 424 509 472 715 120 690 Data not available 962 $5,450 $849 $7e

Police/Domestic Security Budget as % of GDPd 0.34 0.0003 0.1 1.7 0.8 0.7 1.8 0.8 1.8 1.5 1.5 0.8 1 2.4 1.2 0.2 1.3 1.4% Data not available

Notes: The table reports the most recent estimates available for police personnel at the time of writing: 2017 (Honduras, Mexico, United States, Canada, Chile, Colombia, Peru), 2015 (Argentina, Costa Rica, El Salvador, Paraguay, Uruguay); 2014 (Guatemala and Ecuador), 2013 (Panama and Brazil), 2011 (Bolivia), and 2010 (Nicaragua). a. Data on police personnel and percentage of women police are from the United Nations Crime Trends Survey, https://dataunodc.un.org/data/crime/Police%20personnel. b. Data on the percentage of women police are from the same year as the data on overall police personnel, except in the case of Peru and Paraguay, where the data are from 2015 and 2017, respectively. c. For the Central American countries, average salary and security budget data are from RESDAL’s Public Security Index (2013), https://www.resdal.org/libro-seg-2013/resdal-index-2013.pdf. For the remaining countries, average salary data are from the Konrad Adenauer Foundation, https://revista .dialogopolitico.org/wp-content/uploads/2020/05/08-Ruiz-Vazquez.pdf. We report the most recent estimates available: 2014 (Bolivia, Ecuador), 2015 (Chile), 2016 (Argentina, Canada, Nicaragua, Panama), 2018 (Brazil), and 2019 (El Salvador, Guatemala, Mexico, Peru, Uruguay, Venezuela, Colombia). For the United States, average salary data are from the US Bureau of Labor Statistics. d. For the Central American countries, budgetary data are from RESDAL’s Public Security Index (2013), https://www.resdal.org/libro-seg-2013/resdal-index-2013.pdf. For Brazil, Chile, Colombia, and Mexico, data are from the National Institute of Statistics and Geography (2018), https://www .inegi.org.mx/contenidos/productos/prod_serv/contenidos/espanol/bvinegi/productos/nueva_estruc/70 2825107260.pdf. Data for Ecuador are from the Ministry of the Economy and Finances (2020), https://www.finanzas.gob.ec/wp-content/uploads/downloads/2020/09/28092020-Informe-semestral -ejecucion-2020-CNv6-RL.pdf. Data for Argentina, Bolivia, Paraguay, Peru, and Uruguay are based on national budgetary reports and adjusted by GDP data from the World Bank’s World Development Indicators. Data for the United States are from the Organisation from the Economic Co-operation and Development (2019), https://data.oecd.org/gga/central-government-spending.htm#indicator-chart. Canadian data are from the International Monetary Fund (2016), https://www.imf.org/en/Data. e. Estimate gathered during time of hyperinflation.

Policing and Public Security in Latin America

19

compensation. As low police salaries have been cited as a reason police often collude with organized criminal groups, the Mexican federal government conducted a 2021 study on police compensation.5 This report determined that the average monthly salary of police should be US$680 per month, but it also noted that only fourteen states paid police a living wage or higher. The remaining eighteen states (including Mexico City) paid police far less, creating incentives for corruption. Data on police salaries can be unreliable, but generally speaking, police tend to be better compensated in Argentina, Brazil, Chile, Ecuador, Peru, and Uruguay. Comparatively speaking, police earn more modest salaries in Colombia, Costa Rica, El Salvador, Guatemala, Honduras, and Panama; they remain very poorly paid in Bolivia and Nicaragua. In most cases, women represent only a small percentage of police officers. Nicaragua and Uruguay are notable exceptions, as concerted efforts to recruit women to police ranks have led to comparatively high percentages on the force (27.5 percent and 23.5 percent, respectively). However, it is important to note that the most recent data from Nicaragua are from 2010, and as we discuss in Chapter 5, these earlier attempts at including women in police ranks have been undermined by lack of investment and the politicization of police forces. Brazil, Chile, Costa Rica, Honduras, Mexico, and Peru trail a bit behind, with roughly 15–20 percent women police officers. The remaining countries report fewer than 15 percent women in police ranks, on par with the United States. This overview provides a backdrop against which regional reforms have unfolded. Comparative analyses of reform processes highlight just how complex it is to transform police institutions, given the multitude of groups, missions, and problems that face each country (Dammert 2019). Despite these challenges, some countries have clearly made progress and bear little resemblance to their past. We turn to examine the ways policing practices have evolved since the 1990s, noting in particular the many reforms for orienting police toward local communities and improving community–police relations. Community Policing Practices Since the 1990s, most Latin American countries either created completely new civilian police forces or extensively reformed existing ones.6 In most cases these institutional changes were tied to broader political transformations as countries moved to democratic rule. Countries that faced internal conflict (Guatemala, Nicaragua, and El Salvador) implemented the most extensive structural changes, which included creating new civilian police forces and a new definition of police missions. Democratic regimes with longevity also overhauled their policing practices. In the mid-1990s, Costa Rica began to implement a series of gradual reforms to change its policing practices, even though its police forces have long enjoyed a reputation for

20

Making Police Reform Matter in Latin America

professionalism and respect for human rights. Costa Rican elites led the process and aimed to winnow out clientelist practices in the public security sector, create better working conditions for police officers, and centralize a fragmented system (Eijkman 2007). Many initial reform efforts have confronted obstacles. In some countries data collection has not been considered key for designing and implementing policing strategies. Especially at the beginning of the reform process, community involvement in public security is often minimized, as citizens are viewed as a source of information regarding crime-related activities but not as active participants in creating and informing public security practices. At worst, community involvement can even be considered a threat, as police can be leery of civil society oversight. Despite these pitfalls, policing reforms have continued to gain traction, particularly as high crime rates (and public fear of crime) have given public security a premier role in public policy and political debates. Indeed, since the early 2000s, citizens across the region have identified insecurity as a top priority and, in some countries, rank insecurity as the most pressing national problem. To calm a fearful public, some reformers have turned to communityoriented policing reforms, promoting them as a means for improving police performance and increasing ties between police officers and the communities they serve, reducing reliance on repression to maintain order. Community policing is one strategy credited with successfully improving ties with citizens and reducing public feelings of insecurity at the local level in the United States and Europe (Gill, Weisburd, and Telep 2014). Although there are multiple definitions, a commonly accepted one defines community policing as: “A philosophy of policing that promotes community-based problem-solving strategies to address the underlying causes of crime and disorder and fear of crime . . . crime control is shared, or co-produced, with the public . . . thus enhancing quality of life for local communities and building police legitimacy” (Virta 2013, 55). Community policing models seek to strengthen police commitment to service to the community, engage the community in developing local crime prevention strategies, and foster public trust in police forces. Community policing emphasizes improving the relationship between the police and the community so the entities can act as allies in addressing crime at the local level, as well as other problems faced by the community. The conceptual pillars of community policing focus on nontraditional crimefighting activities, such as maintaining order, reducing fear, and resolving community conflicts (Gill, Weisburd, and Telep 2014). In Latin America, Frühling (2012) analyzed the wide variety of initiatives that fall under the name of community policing and found that the only thing they had in common was this name, given that their activities, tasks, and strategic definitions varied significantly. Still, community policing models put their emphasis squarely on strengthening ties with citizens.

Policing and Public Security in Latin America

21

Since the 1990s, community policing models have migrated and settled in Latin America with different results (Dammert 2012a; Mota-Prado, Trebilcock, and Hartford 2012; Hathazy 2013). All Latin American police forces have at least one office ostensibly devoted to community policing, but hanging a sign on a door does not necessarily mean that community policing models have been successfully transplanted. Many countries inherited repressive and corrupt police forces from their authoritarian predecessors, and there is a great deal of variation in how well countries have shed authoritarian practices and gone beyond mere lip service in promoting communityoriented policing practices. The adoption of a community policing model generally does not require changes in the organic laws or regulations of the police forces. The most relevant transformations can be found at three different levels. First, at the philosophical level, security policy is reframed to include the citizen as a coproducer of security and a proactive participant (Neild 1998). This conceptualization of security gives police functions far more breadth, expanding their work to include things such as collaborating with other institutions and actors to face the most relevant problems in a community. Public trust in police is a critical component of this transformation, as it envisions the community as an active collaborator working with police on the ground. Second, at a strategic level, community policing models prioritize crime prevention and territorial concentration of police work. Community policing decentralizes command levels to facilitate identifying and implementing targeted solutions for specific spaces and times. This decentralization provides more flexibility for patrol techniques, allowing approaches that highlight community involvement (Greene 2000). Community policing models extoll the benefits of police patrolling on foot, which allows direct contact with the community daily (Neild 1998, 6) and raises the profile of police presence on the streets (Frühling 2003, 19; Labra 2011, 54). Community policing requires reforms (particularly those that permit flexibility) and resources to provide education and specialized training to police officers. To promote transparency and accountability, different management parameters are needed to monitor the daily operations of police work; traditional benchmarks of police performance need to be overhauled to include commensurate indicators of efficiency and effectiveness (Labra 2011, 60; Lum and Nagin 2017). Police on the street could be perceived with caution by the community if their routine work is considered violent, corrupt, or inefficient. To consolidate changes at this strategic level, the public must trust that in their daily routines, the police uphold the law equally for all citizens, limit their use of force, and respect citizens’ rights. Third, at a tactical level, community policing models recognize the importance of linking the institution of the police with various local actors (including those from municipalities, neighborhood associations, and social organizations) to strengthen the results of preventive work. According to

22

Making Police Reform Matter in Latin America

Skogan (2004), the community policing model is based on an adaptive strategy that responds to the demands of a constantly changing environment. For this adaptive strategy to work well, police information systems need to gather and analyze accurate data and use it to inform patrolling techniques and decisionmaking in the community. Fundamentally, community policing reconceptualizes the police–community relationship, and this new perspective carries several assumptions. Community policing models focus on neighborhoods or small communities to delimit the space of police organization and operations, in contrast to a traditional, broader territorial focus. Police work takes place at the community level, with specific goals and methods that should target the intrinsic (or even historical) characteristics of a community. Community policing models recognize that each community faces problems that could be unrelated to crime but would benefit from police presence and action. These parallel problems include community disorder or increasing levels of incivilities, which if left unchecked could lead to heightened violence. Given this focus, for community policing models to work, the police must earn the trust of community members and other local organizations. Community policing requires organizational changes to reform the institutional structure of police and give them the flexibility, education, and training to respond to local problems. Community policing reforms should establish management parameters that recognize the importance of transparency and accountability in officers’ daily lives and develop indicators of efficiency and effectiveness of police work to replace traditional ones that emphasize arrests and detentions (Labra 2011, 60). Although most Latin American police forces have at least one unit named policia comunitaria (community police), in practice these units vary greatly in terms of aims, scope, and magnitude. In many cases, the philosophical changes needed to reconceptualize public security and fully implement the community policing model have not happened at the national and local levels. Most police institutions persist in perceiving the community as an external actor that could assist in specific activities or actions, especially at the local level, but is far from co-producing security. Concentration of power has been the dominant trend in police work in Latin America, and decentralization is often seen as a possible threat to corruption and entrenched ineffectiveness. High levels of autonomy of those in the higher ranks has embedded corrupt practices in institutional structures, even in countries where the police are widely regarded as professional (e.g., the police corruption scandals in Colombia and Chile in 2017–2018). Furthermore, new indicators of police performance have not materialized to capture the value of prevention strategies; instead, traditional indicators of effectiveness, such as the number of detainees and drug seizures, continue to be norm. Because of these limitations, community policing tends to be a secondary strategy, often developed by a specific

Policing and Public Security in Latin America

23

unit whose main objective is to talk to the community and implement specific activities in the public space (Dammert 2019). Even so, we do see some noticeable changes. Most police forces have increased links with local governments and civil society organizations. Although the results of such initiatives are not immediately clear, there is increasing evidence of police collaborative work on specific initiatives throughout Latin America (Dammert 2019). For example, Colombia and Chile both inaugurated Plan Cuadrante (Quadrant Plans) to consolidate collaborative work with local governments, ranging from data gathering to implementing initiatives. In Uruguay, the National Police developed a strategic reform that includes an evidence-based problem-oriented policing strategy to enhance the police–community relationship. In Panama, the National Police created a Community Preventive Unit in 2012 with the goal of fostering better police–community relations in high-crime areas. In Peru, the National Police partnered with the Ministry of Women and Vulnerable Populations to build more than fifty “Women Emergency Centers” to bring state resources to victims of violence and their families. Although there is no comprehensive record to gauge how exactly community policing strategies have shaped Latin American police forces, there are multiple data points on police work that we can use to initially define its importance in each country. Since almost all police institutions have had at least some type of community-oriented program, we reviewed institutional strategic statements, implementation of national programs, and the development of specific indicators to value preventive practices and police– community relations. We relied on extensive local and country-level case studies and interviews with actors from government agencies, nongovernmental organizations, civil society groups, and academic researchers. Clearly community policing is best implemented at the local level, but because only a handful of countries have local police forces (Brazil, Venezuela, Argentina, and Mexico), we place countries on this continuum based on the practices of national police institutions. Although we cannot unilaterally declare whether a country has uniformly implemented a community policing model, we can say that Uruguay and Costa Rica shared institutional definitions as well as national programs that aimed to develop community-oriented policing strategies. These institutional changes demonstrate a commitment toward a more community-oriented perspective to policing practices. In many countries, community-oriented reforms have been implemented, but these initiatives have been more like pilot projects (with shortterm periods of implementation), or initial reforms have been subject to substantial reversal. In Ecuador, the narrative of community policing has changed definitions of national police missions, but there is no clear evidence of real changes in everyday police work. In Chile, Colombia, and

24

Making Police Reform Matter in Latin America

Nicaragua, community-oriented reforms witnessed a dramatic reversal when the police responded to protesters with extreme repression. In Nicaragua, this repressive political orientation has steadily undermined two decades of community-oriented police practices, leading some observers to speculate that the Nicaraguan model will completely revert to its authoritarian and repressive incarnation under the Somoza family (in power 1933–1979). Finally, several countries remain firmly rooted in traditional policing practices. In the context of Latin America, this means that police work is closely linked to repressive strategies to control crime and public disorder. Although some community policing initiatives may have been developed, as was the case in Argentina and Brazil, such initiatives are sporadic, underfunded, based on broad territories rather than local communities, or involve a limited number of police officers.

Crime and the Security Crisis Police reforms have been implemented in a specific context of crime and insecurity, in several cases marked by extraordinarily high rates of violent crime. The advent of democratization coincided with a sharp uptick in violence and crime. Throughout the early 2000s and 2010s, El Salvador and Honduras alternated as the most violent country in the world, and even countries such as Costa Rica, which had long enjoyed a reputation for peace and stability, saw murder rates increase sharply. More recently, Venezuela has reported some of the highest homicide rates in the region, as murder rates have surged against a backdrop of economic crisis and political instability. Police are on the front lines of this security crisis, and processes of police reform have been complicated by the fact that these reforms have unfolded against such a violent backdrop. High rates of violent crime have made it difficult for many countries to jettison authoritarian and repressive police practices. Opponents often label reformers as “soft on crime,” and argue that reforms favor the interests of criminals over those of crime victims. Many countries, such as Chile or Colombia, have been tempted to abandon (or at least dilute) reform efforts in favor of mano dura (iron fist) crackdowns (Rosen and Cutrona 2021). In cases such as El Salvador and Brazil, high crime levels have been used as a reason to use mano dura tactics, whereby the police (and typically military) confront crime with an iron fist, an approach that relies heavily on repression and very little on prevention and frequently involves mass arrests, incarcerations, and human rights violations (Wolf 2021). Typically, mano dura campaigns are accompanied by charges of police abuse and excessive force (Chevigny 2003). As Ungar (2009) notes, “mano dura is often just a continuation of pre-democratic practices and a justification for the dividing line drawn by many officials between ‘public order,’ associated

Policing and Public Security in Latin America

25

with a strong state, and ’human rights,’ associated with delinquency” (95). Indeed, as Rios, Prieto, and Ibarra (2020) note, mano dura practices can even creep into community-oriented initiatives, as when police officers tried to foster community trust while at the same time engaging in extensive stop-and-frisk practices. In Latin America, it is not uncommon for community-oriented efforts to take place against a broader backdrop of punitive mano dura practices, undermining reform efforts and resulting in a paradoxical approach to policing. While rates of violent crime have abated somewhat in 2019–2020, as Figure 2.1 illustrates, the regional average of homicide rates in Latin America is far higher than the global average.7 After peaking in 2009–2012, regional homicide rates started to decline slightly, and by 2020, the regional

Figure 2.1

Latin American Homicide Rates per 100,000 People (2005–2020)

Notes: The World Bank reports reliable homicide data for most countries through 2020. When the World Bank does not report data, InSight Crime is a reliable organization that gathers homicide data from regional police agencies and observatories (https://insightcrime.org/). The World Bank did not report homicide rates for El Salvador (2020), Nicaragua (2020), Bolivia (2020), Peru (2019, 2020), and Venezuela (2019, 2020), so this figure relies on InSight Crime’s homicide estimates for these cases. Given the political instability in Venezuela, homicide data are less reliable and should be treated with caution.

26

Making Police Reform Matter in Latin America

average reached its lowest rate in 15 years (16.4 homicides per 100,000 people). Still, Latin American homicide rates in 2020 were almost three times the global average. Of course, these regional averages mask a great deal of variation among countries. Figure 2.2 reports homicide rates by country in 2020. Only three countries—Chile, Argentina, and Nicaragua—have homicide rates below the global average. Paraguay, Ecuador, Peru, and Uruguay reported homicide rates slightly higher than the world average, and in Costa Rica and Panama, homicide rates were approximately double the global average. In contrast, homicide rates in Venezuela and Honduras were among the highest in the world. In Mexico, Colombia, Guatemala, El Salvador, and Brazil, homicide rates are also quite high and reflect the presence of large, organized criminal groups. Homicide data allow us to examine trends in the most serious and violent of crimes in the region, but it is also important to consider the prevalence of other types of crimes. To measure trends in these other types, we turn to self-reported victimization rates. LAPOP has consistently included questions related to policing, public security, and victimization in its AmericasBarometer surveys for nearly fifteen years.8 To begin, LAPOP asked respondents, “Have you been a victim of any type of crime in the past twelve months? That is, have you been a victim of robbery, burglary, assault, fraud, blackmail, extortion, violent threats or any other type of crime in the past twelve months?” Figure 2.3 depicts the percentage of respondents in each country who indicated they had been victimized by some type of crime in the previous year. We see that many countries with low homicide rates still struggle to contain other types of crimes. Peru tops the chart with over a third of the population indicating they had been victimized by some type of crime in the past year, even though its homicide rate was low for the region. Uruguay and Argentina, which also had comparatively low homicide rates, registered high rates of victimization as well. Once again, it is important to view crime rates over time, not just as a cross-section. When we compare self-reported victimization rates in 2018/19 to earlier rates in 2008, we see that victimization rates increased in every country between the 2008 and 2018/19 waves of the survey, and this difference was statistically significant for more than half the countries in the region. Although homicide rates have been on the decline in the region, victimization from other types of crime has increased or remained the same. Public Fear of Crime Homicide rates and self-reported victimization, when interpreted with caution, help gauge the actual occurrence of crime. Although these more objective indicators are very useful, subjective perceptions of crime also help

Latin American Homicide Rates per 100,000 People by Country (2020)

Source: Data for El Salvador, Nicaragua, Bolivia, Peru, and Venezuela are from InSight Crime. Remaining data are from the World Bank’s World Development Indicators.

Figure 2.2

27

Self-Reported Victimization Rates by Country (2019)

Source: LAPOP’s 2018/2019 AmericasBarometer.

Figure 2.3

28

Policing and Public Security in Latin America

29

shape security policy. In many countries, fear of crime has become a more critical political issue than the occurrence of crime itself. Fear of crime is related to crime’s actual incidence, but this overlap is not complete. Public fear of crime is tied to a variety of factors, including patterns of urbanization, social identity characteristics, civic participation, media influence, trust in justice institutions, and insecurity about other issues such as economic livelihood (Dammert and Malone 2006; Bautista Lara 2008; Krause 2014; Singer et al., 2019; Hernández, Dammert, and Kanashiro 2020; Singer et al. 2020). Garland (2002) ties fear of crime to anxiety over sweeping societal change, particularly when these changes reconfigure racial, class, and gender power relations. Indeed, preexisting racial and/or ethnic divides can politicize crime, making it easier for societies to create archetypes of fearful criminals by labeling lawbreakers as “dangerous members of distinct racial and social groups which bear little resemblance to ‘us’” (Garland 1996, 461). Thus, crime becomes a tangible scapegoat for all types of insecurities. Fear of crime has increased steadily throughout Latin America, dominating public discourse and social interactions, even in countries widely considered to be safe (Malone and Dammert 2020). For example, Cruz (2009) notes that Argentina and Chile, two of the safest countries in the region, often register the highest levels of public insecurity. Scholars and practitioners agree that reducing fear of crime is almost as important as reducing crime itself. Public perceptions of crime often have political ramifications, as citizens’ feelings of insecurity make their way to the ballot box and to the ears of public officials. Scholars and practitioners have noted the impact of public insecurity, linking higher levels of fear of crime to lower levels of support for democracy and its norms (Peréz 2015; Kessler and Otamendi 2020; Singer et al., 2020). To measure fear of crime, we rely on a LAPOP question designed to gauge citizens’ fear of victimization in their own neighborhoods: “Speaking of the neighborhood in which you live and thinking about the possibility of being the victim of an assault or robbery, do you feel (1) very safe, (2) somewhat safe, (3) somewhat unsafe, or (4) very unsafe?” This measure is a cognitive one, tapping into respondents’ personal assessments of the likelihood of their future victimization in the neighborhood. As Figure 2.4 illustrates, several countries with low levels of homicides but high levels of self-reported victimization register high levels of public fear of crime. Peru, for example, ranks significantly higher than Mexico and Colombia, despite the fact that murder rates are substantially higher in the latter two countries. While fear of crime is subjective, it is nevertheless important given that perceptions are what fuel citizen behavior and policy preferences. Thus far, we have provided an overview of the Latin American public security sector, community policing initiatives, and the larger backdrop of

30

Making Police Reform Matter in Latin America

Figure 2.4

Public Fear of Crime by Country (2019)

Source: LAPOP’s 2018/2019 AmericasBarometer.

crime against which these initiatives unfolded. We turn now to examine public perceptions of police. Here, we provide a regional overview of attitudes toward and experiences with police. In the country case studies, we empirically assess the linkages among police performance, crime, and socioeconomic factors to public trust in police.

Public Perceptions of Police How has the public regarded the police amid these transformations? As Lum and Nagin (2017) argue, in democracies citizen trust in police matters in its own right and is a critical performance metric. Police must ensure public safety, but they must also garner citizen trust. Public trust in the police is essential, for if citizens view their police as corrupt, inefficient, or abusive, they will be reluctant to turn to them for protection or for solving problems in their communities or homes (Sunshine and Tyler 2003; Tankebe 2013).

Policing and Public Security in Latin America

31

Furthermore, police are typically the state actor with which citizens have the most contact. As Moncada (2009) notes, since “the police are a highly tangible manifestation of the state on the streets, police accountability affects not only citizen security but also the broader dynamics of state-society relations” (431). Under dictatorships, a key grievance was lack of respect for human rights and civil liberties, typically at the hands of police (and the military). Recent research indicates that improving police performance in these areas matters not just for police but also for the legitimacy of the political system as a whole. When police treat the public poorly, for example, by demanding bribes or engaging in verbal or physical abuse, such misconduct erodes public support for the overall democratic system (Cruz 2015). Therefore, to legitimize themselves in the eyes of the public, democratic governments feel pressure to demonstrate that police abuse is a relic of the past. To examine public perceptions of the police, we turn to recent survey data. In 2018/19, LAPOP included the following question in its regional AmericasBarometer survey: “To what extent do you trust the (name of national police force)? (1) not at all–(7) very much.” This item was part of a battery of questions measuring trust in a wide range of institutions (e.g., the legislature, courts, elections).9 Figure 2.5 displays average levels of trust in police by country. As we can see, many of the countries that we have highlighted as engaging in community-oriented practices rank higher in the region, although there are some exceptions (e.g., Brazil). Although these average levels of trust illuminate national trends, it is important to remember that public trust in the police tends to be highly correlated with trust in other national institutions (Goldsmith 2005; Bridenball and Jesilow 2008; Andreescu and Keling 2012). Furthermore, indices created by such survey batteries are further correlated with perceptions of economic prosperity and other national trends. We attempt to isolate trust in the police from trust in these other institutions. To this end, we created a composite variable gauging average trust in the key political institutions in each country. We computed average levels of respondents’ trust in the judicial system, legislature, political parties, president, municipal government, and elections.10 We subtracted each respondent’s trust in the police from the average trust in other political institutions.11 The resulting variable, relative trust in police, measures the extent to which respondents trusted police more than other political institutions. Positive values represent more trust in the police, and negative values indicated that respondents trusted the police less than other political institutions. Figure 2.6 compares trust in police with average levels of trust in other political institutions (trust in the national legislature, political parties, the president, the national high court, the municipality, and elections).

32

Making Police Reform Matter in Latin America

Figure 2.5

Average Trust in Police by Country (2019)

Source: LAPOP’s 2018/19 AmericasBarometer.

As the figure illustrates, in most countries trust in police was higher than average trust in other political institutions. In Guatemala and Nicaragua, relative trust in police hovers around zero, indicating that respondents register similar levels of trust in police and other institutions. Only in Paraguay, Mexico, and Bolivia was there more trust in other political institutions than in the police. Mexico’s and Bolivia’s lower rankings reflect that in these countries, respondents report much higher levels of police bribery. LAPOP included one question that directly gauged respondents’ personal experiences with police, asking, “Has a police officer asked you for a bribe in the last twelve months? (1) yes (0) no.” Figure 2.7 reports the percentage of respondents by country who reported that police had solicited a bribe personally from them. As we can see, bribery was very low in Costa Rica, Uruguay, and Chile. In most Latin American countries, 6 percent to 15 percent of respondents indicated that police had solicited a bribe from them in the past year. Mexico and Bolivia are at the high end of the spectrum, with significantly

Relative Trust in Police by Country (2019)

Source: LAPOP’s 2018/19 AmericasBarometer.

Figure 2.6

33

Police Bribery by Country (2019)

Source: LAPOP’s 2018/2019 AmericasBarometer.

Figure 2.7

34

Policing and Public Security in Latin America

35

much higher rates of bribery than all other countries in the region. Roughly one fourth of respondents in these countries indicated that police had asked for a bribe, a very high percentage.

Conclusion This regional overview highlights important trends in policing and public security. As we can see on multiple measures, the countries we selected for our case studies tend to perform better than regional trends on at least one indicator. For example, Nicaragua reports very low levels of homicide and victimization, despite a historical legacy of political violence, poverty, and inequality. On average, trust in police is highest in Uruguay and above the regional average in Costa Rica, Chile, and Panama. Costa Rica, Chile, and Uruguay report some of the lowest levels of police bribery. Colombia was one of the first countries to follow Chile’s lead in pioneering community-oriented policing programs. Against this backdrop, we turn to examine the evolution of police reforms in each country. We trace the institutional process by which these reforms unfolded and gauge public attitudes toward police at various stages of the reform process. We strive to identify the factors that led to the implementation of successful police reforms and assess their sustainability.

Notes 1. For more information on LAPOP’s surveys, see www.vanderbilt.edu/lapop. We thank LAPOP and its major supporters (the US Agency for International Development, the Inter-American Development Bank, and Vanderbilt University) for making the data available. 2. For example, in Mexico, debates surrounding the Agencia Federal de Investigación (created in 2002), the Federal Police (created in 2009), and the National Guard (founded in 2018), have emphasized the need for stronger police institutions that could lead the battle against organized crime. 3. In the early twenty-first century, drug trafficking routes shifted to run through Central America, and leaders in the area took up the mantle of the war on drugs and its corresponding militarization of certain political groups whose infrastructure, formation, and strategy has been linked to the use of violence (Mota Prado, Trebilcock, and Hartford 2012). 4. We explore this phenomenon in greater depth in our case studies of Peru, Colombia, and Chile. 5. This report, titled Salario Digno para Policías en México, is available at https://www.gob.mx/sesnsp/documentos/salario-digno-para-policias-en-mexico. 6. For an overview of police reforms in Latin America in the 1990s, see Uildriks (2009), Dammert (2019), and Saín (2010). Sieder, Ansolabehere, and Alfonso (2019) provide a broader, comprehensive overview of reforms to Latin American justice systems.

36

Making Police Reform Matter in Latin America

7. Studies of policing and crime rely on homicide rates as a measure of overall crime in a country. As the United Nations Office on Drugs and Crime (UNODC) explains, “intentional homicide is one of the most measurable and comparable indicators for monitoring violent deaths. Because of its lethal outcome, homicide is particularly amenable to temporal (longitudinal) and cross-national (geographic) comparisons: it tends to have greater definitional specificity than other crimes in different historical and national contexts. Homicide is an act that meets with virtually universal condemnation, and homicide statistics are accordingly considered to be relatively reliable and valid—both at the national level and for longitudinal and cross-national comparisons. As a readily measurable indicator, homicide is both a reasonable proxy for violent crime and a robust indicator of levels of violence within States” (UNODC 2019, 7). 8. Cross-national studies of crime cannot rely on official crime statistics due to inherent problems of measurement reliability and validity. Differences in definitions of crime, crime-reporting rates, and data collection make national reports on crime difficult to compare country to country. Given these difficulties, comparative scholars tend to rely on homicide rates to gauge the actual occurrence of crime, as homicides tend not to have the same problem with underreporting and can be more easily compared cross-nationally than other crimes, where legal definitions may vary (Lafree and Tseloni 2006). A strict reliance on homicide rates can mask problems with other types of crime, such as burglary or robbery, however. Homicides are the most serious of crimes, but other crimes also matter to citizens. To measure the prevalence of these other types of crimes, researchers turn to self-reported rates of victimization in public opinion surveys. Victimization surveys themselves can also be problematic, as respondents might self-report crimes outside the specified time range, or might dismiss crimes if they think of them as unimportant (Bergman 2006). Victimization surveys tend to overreport crimes involving property, so wealthier countries often report higher levels of victimization. Despite these recognized limitations, victimization surveys provide valuable insight into respondents’ self-reported experiences with less violent types of crime, provided such data are used to document general trends as opposed to absolute rates of victimization. 9. The full 2018–2019 LAPOP questionnaire is available at https://www .vanderbilt.edu/lapop/ab2018/AB2018-V12.0-Eng_190131_W.pdf. 10. These survey items were part of the battery of questions that measured trust in police. The exact text of this battery read: “On this card there is a ladder with steps numbered 1 to 7, where 1 is the lowest step and means NOT AT ALL and 7 the highest and means A LOT. For example, if I asked you to what extent do you like watching television, if you don’t like watching it at all, you would choose a score of 1, and if, in contrast, you like watching television a lot, you would indicate the number 7 to me. If your opinion is between not at all and a lot, you would choose an intermediate score. So, to what extent do you like watching television? Read me the number.” Respondents were then asked to indicate their levels of trust in a variety of institutions, including the police, the national high court, national legislature, political parties, the president, municipal government, and elections. To compare trust in police to trust in these other institutions, we created an additive index using a means formula that calculated average responses across these items, provided respondents answered four out of the six questions. The Cronbach’s alpha for this index is 0.845. 11. Relative trust in police = trust in police – average trust in other political institutions.

3 Chile: Too Good to Be True?

UNTIL RECENTLY, WHEN ONE MENTIONED POLICING SUCCESS stories in Latin America, Chile was typically the first country that came to mind. Unlike their regional neighbors, the Chilean Carabineros have been highly regarded, despite their complicity in the dictatorship of 1973–1990 (Dammert 2020). Chileans report that their police forces are far less likely to solicit bribes or abuse citizens’ rights compared with their counterparts in other Latin American countries. Chile has also pioneered the inclusion of the community in its crime control policies, launching a series of initiatives to include citizens’ input in designing neighborhood-based security strategies (Frühling 2009; Lunecke 2018). Indeed, many governments in the region have sought to replicate the Chilean model in their attempts to reform their own struggling police forces. These policing tactics are often credited, at least partially, with maintaining comparatively low crime levels in the country.1 Violent crime rates historically have been low in Chile, on par with European countries. Although the Carabineros had been renowned for their professionalism, training, orientation toward local communities, and low levels of petty bribery, their institutional reputation began to deteriorate in 2017, as reports emerged of acts of corruption at the highest institutional levels, violations of human rights, and a lack of efficiency. The Carabineros’ community-oriented reputation was further tarnished in October of 2019, when they unleashed excessive force against demonstrators, leading to a period of widespread reports of police abuse and violence against protesters and detainees. United Nations investigators denounced the “fundamentally repressive manner” of the Carabineros’ reactions to demonstrations and called for prosecution of the army and police for human rights violations, which 37

38

Making Police Reform Matter in Latin America

included unlawful killings and torture that lasted until March 2020. Some Chileans responded to the police’s misconduct with violence of their own, escalating clashes between civilians and Carabineros in the streets. The use of violence against demonstrators led to public condemnation and questions about the role of the police in Chilean democracy and the sustainability of community-oriented reforms. In this chapter, we trace the development of police reforms since the return of democracy in 1990, which marked a significant juncture that led to deep changes in Chile’s security apparatus. We explain how the Carabineros responded to democratization with a series of community-oriented reforms to rebrand themselves after their complicity in Pinochet’s dictatorship. We highlight the importance of political stability for the success of police reforms, as the return to democracy ushered in thirty consecutive years of governance under the same center-left coalition, the Concertación de Partidos por la Democracia (Coalition of Parties for Democracy), which invested in professionalizing the Carabineros and increased their budget and remuneration. We examine public evaluations of the Carabineros after these reforms, noting the comparatively high levels of public trust and positive evaluations in the early 2000s, which are linked to perceptions of the Carabineros’ effectiveness in controlling crime and general reluctance to engage in petty bribery. We note how the institutional autonomy of the Carabineros inhibited mechanisms of accountability, however, ultimately culminating in a series of scandals that squandered public trust. We conclude with a discussion of public evaluations of Carabineros after the 2019 social uprising and accompanying police abuse.

The Return of Democracy and the Rebranding of the Carabineros For much of its history, Chile enjoyed a reputation of exceptionalism, sidestepping many of the problems plaguing other South American countries. Prior to Augusto Pinochet’s right-wing military coup in 1973, Chile had boasted a strong democratic tradition. In keeping with this tradition, the country upheld the rule of law, particularly through well-trained and professional public servants such as the national police force, the Carabineros of Chile. In a thorough historical analysis, Esparza (2022) traces the development of the modern Carabineros throughout the nineteenth and twentieth centuries, documenting the ways successive governments sustained a commitment to professionalize and modernize the police. There was political consensus that Chile needed a centralized police system to confront banditry in the countryside and crime in the cities, and successive governments increased funding to provide police with higher salaries, more training, and a social

Chile

39

safety net (e.g., healthcare, subsidized housing) (Esparza 2022, 43). By 1932, Chilean democracy was firmly established, and the Carabineros largely assumed an apolitical role, as disputing parties tended to resolve their differences through political channels rather than through repression, and widespread protection of civil liberties and political rights created space for competing interests and visions. The 1973 military coup interrupted this democratic tradition for seventeen years, replacing democracy and the rule of law with repression, arbitrary detention, torture, and murder. Military rule also tarnished the reputation of the Carabineros, as the police actively participated in the coup and were the only police institution in Latin America to form part of a military junta. In 1989, democratic elections put an end to military rule and ushered in a new era for Chilean democracy. At this critical juncture, Chile discarded authoritarian practices and institutions in favor of democratic ones in the 1990s, and reformers targeted the justice system as a crucial part of their efforts. Still, the constellation of actors in the postauthoritarian environment, and their institutional setting, slowed the reform process. Since Pinochet and his allies were not completely vanquished and discredited, former authoritarian officials wielded a substantial amount of power in the early years of the return to democracy, and Chilean political institutions were designed to preserve this power.2 The constitution that governed Chile for the first thirty years of democratic governance was authored by Pinochet himself in 1980, and only its most egregious authoritarian elements were removed with the transition to democracy. Although right-wing parties and leaders garnered the support of only a third of Chilean citizens, institutions inherited from authoritarian rule magnified their power and allowed them to block or dilute reform efforts they did not support. For the police, this meant that the reform process was gradual. The Carabineros retained their highly centralized, hierarchical, military structure. The return of democracy did not completely sever the Carabineros’ tie with the military. Resisting reformers’ calls that the Carabineros report to the Ministry of the Interior, the national police remained under the jurisdiction of the Ministry of Defense up to 2011, guaranteeing the Carabineros a great deal of institutional autonomy. Because of these high levels of autonomy, the transformation of the Carabineros under democracy has been promoted and developed from within the institution. There are few instances of reforms generated from civil society, experts, or government proposals, given that the space for these actors to participate in police matters is minimal. Pinochet’s constitution left multiple safeguards that impeded the reform of educational bodies, personnel distribution, and even the format for evaluating the income of applicants. Still, it is fair to say that under democracy, the other political parties did not do much to change these dictatorial “locks” either. On important issues, such as the distribution of police personnel, there is still a

40

Making Police Reform Matter in Latin America

strong informational bias. Even in the present day, the mayors of Chilean communities claim not to have sufficient endowments or real capacity to define their public security priorities. Similarly, calls to reform the system of military justice for Carabineros waited for more than two decades, as changing this system of accountability required constitutional reform, and parties across the political center and left have been hesitant to challenge their counterparts on the right on this issue. High levels of citizen trust also insulated the Carabineros from calls to reform. At least partially due to the very positive evaluations citizens bestowed on the police, most democratic governments continued to grant the Carabineros a great deal of autonomy, allowing them to self-regulate their institutional changes and strategic decisions. Most initiatives designed and implemented since the 1990s tend to be driven by forces inside the Carabineros, as officials populating the higher institutional ranks have basically determined the areas they want to prioritize. Despite lingering authoritarian tendencies, in many ways this process of gradual reform served the Carabineros well, at least initially. The police rebuilt their ties to Chilean society, and until recently, were typically commended for their professionalism, ability to maintain public order, and respect for citizens’ rights. This acclaim was not universal, however, as social groups at the bottom of the socioeconomic ladder registered significantly less favorable evaluations of police performance and reported far more personal experiences with police misconduct (Lunecke 2018; Dammert 2020). To rebuild their ties to Chilean society as part of the transition to democracy, the Carabineros launched a series of initiatives to engage the community into policing strategies. To distance themselves from their legacy under authoritarian rule, the police gradually opened the door to community-oriented policing initiatives and established themselves as pioneers in this approach to public security. Along with some knee-jerk mano dura decrees, Chilean policymakers sought to address crime increases and public insecurity with a series of municipal and national policies that emphasized the Carabineros’ preventive role in maintaining public security and the critical role the community can play in supporting police activities. Engaging the Community in Public Security Initiatives In the late 1990s, reforms altered the internal structure of the Carabineros, increasing the number of officers engaged in operational activities. These officers were then redeployed under new programs designed to integrate the police into the neighborhoods they served. For example, in 1994 the Programa Puertas Abiertas (Open Door Program) aimed to make the police more visible in the public’s eye. Under this program, Carabineros initiated a series of dialogues with neighborhood organizations and leaders, intend-

Chile

41

ing to learn about local concerns and disseminate basic information such as emergency telephone numbers. The Programa Seguridad Compartida (Program for Shared Security) followed close on its heels and also strived to publicize the importance of public collaboration with the police to confront crime, as well as train Carabineros to work with community organizations. In 1998, mobile police stations emerged to increase police presence at the local level and liaise with communities, developing the grounds for future community-oriented policing programs. Most community-oriented programs emerged at the behest of domestic political officials; in some cases even after their implementation, citizens were not aware of community policing initiatives in their own neighborhoods. In addition to the initiatives of elected and appointed leaders, some domestic nongovernmental organizations have been important actors in addressing issues of public security more broadly. Perhaps most prominently, the Fundación Paz Ciudadana (Citizen Peace Foundation) (FPC) emerged as a major policy institute concerned with the issue of public safety and has collaborated with the government to design policies that promote citizen engagement with public security policies. In 1999, the official rhetoric extolling community involvement in crime control policy became more concrete with the implementation of Plan Cuadrante (Quadrant Plan). Plan Cuadrante aimed to increase police street presence by dividing the national territory into zones (cuadrantes) and assigning a police station to each zone, along with the necessary vehicle and personnel resources. Appointed personnel have been responsible for thoroughly personalizing themselves with the communities in their zone. To facilitate this personalization, each zone has a community relations division to work with residents and civic organizations, as well as designate officials responsible for dialoguing with residents to identify security concerns and solve community problems. Each cuadrante official carries a cell phone to receive calls from community members, facilitating direct communication between citizens and their local police officers. The Carabineros described this program as one of “preventive patrolling practices in coordination with the community” (Candina 2006, 91). Plan Cuadrante initiated a series of activities that engaged the community in the policing strategies of their neighborhoods. For example, the Carabineros held public hearings to explain their policing practices to neighborhood leaders and solicited feedback from the communities. When the Ministry of Finance commissioned an evaluation of Plan Cuadrante in 2007, however, it found that the plan was effective in terms of managing financial and personnel resources but did not adequately incorporate the community in identifying and prioritizing security concerns (Dirección de Presupuestos 2007). The findings persisted in evaluations in subsequent years (Fundación Paz Ciudadana 2012; Dirección de Presupuestos 2013).

42

Making Police Reform Matter in Latin America

In 2000, the government and the police designated community participation in crime prevention as a critical pillar of public security initiatives. This community engagement occurs on two levels, which ideally are complementary. On one level are policies that emphasize citizen involvement in prevention by forming community councils, installing working groups, and developing grant competitions. Such policies create space for organized citizens to access public funds to develop prevention initiatives that are mainly situational. On the second level are initiatives developed by the Carabineros to improve relations with the community and enhance their ability to prevent crime. Policies on the first level are designed and implemented by the national government, and the policies on the second level are developed by the Carabineros. In the beginning of the twenty-first century, several ambitious programs aimed to build on police–community relationships to formulate public safety policies. For example, Comuna Segura Compromiso 100 (Safer Cities Program 100) came to fruition in the mid-2000s from collaboration between the Interior Ministry and the FPC. This program encouraged community groups to diagnose crime-related problems in their neighborhoods and develop a local agenda to prioritize the most pressing issues, for which they could request government financing. Comuna Segura Compromiso 100 expanded over four phases of implementation and eventually included fiftysix communities, more than half of the national population. Barrios Vulnerables (At-Risk Neighborhoods) emerged in roughly the same time period, targeting high-crime neighborhoods with drug trafficking problems in metropolitan Santiago. Barrios Vulnerables aimed to address the many facets of crime plaguing these neighborhoods by investing in social programs targeting the communities’ youth. It also encouraged communities to work with the police to provide intelligence to disrupt the drug trade networks. While Chile’s geography has isolated it from the major international drug trading routes, and it has not experienced as sharp an increase in internal drug consumption as other Latin American countries have, drug use has increasingly become a problem and caused an uptick in crime in some neighborhoods. To increase public security in at-risk neighborhoods, Barrios Vulnerables sought to combine community–police collaboration with investment in social programs. Evaluation mechanisms for the program are still in development, but hopes remain high for a program that combines investment in social welfare with police disruption of drug trafficking. Although concrete data on the effectiveness of these interventions in reducing crime is hard to come by, in the twenty-first century subsequent governments have maintained investments in these types of programs.3 Although there are name changes, the programs share some substantive elements, such as encouraging community groups to diagnose crime-related problems in their neighborhoods and develop a local agenda to prioritize the most pressing issues. Once the community prioritizes its security needs,

Chile

43

it designs initiatives and requests government financing for its security program. These initiatives were also a political tool that the national government used to establish a working relationship with municipal governments receiving specific funding for crime prevention initiatives. Listening and Responding to Community Feedback The Carabineros aimed to systematically incorporate the voice of the community. To collect data on citizens’ security needs and police performance, the Carabineros developed a series of new performance indicators in the early 2000s. In 2003, for example, the Ministry of the Interior conducted the Estudio Percepción y Evaluación de la Labor de Carabineros de Chile (Study of Perceptions and Evaluations of the Work of the Carabineros of Chile), a survey to assess public perceptions and evaluations of specific police duties. Among the many performance measures, the survey asked respondents to rank the duties they considered to be most important and then evaluate how well the Carabineros fulfilled them. Although concerns and priorities did vary depending on socioeconomic groups, on average respondents ranked the sale of illegal drugs as their first concern, followed by minors’ consumption of alcohol and burglaries. When asked to evaluate how well the Carabineros responded to these concerns, responses revealed a gap between public expectations and performance. For example, the public prioritized concern over the illegal drug trade and evaluated police performance in this area negatively. Such surveys are valuable tools for tailoring police performance to meet the needs of local communities. This emphasis on performance and responsiveness to citizens intensified in 2006, when the Carabineros launched the Quality Improvement Program for Police Services to the Community, which assessed citizens’ satisfaction with police performance through a series of surveys of those who had visited police stations or called the police. The survey asked how promptly the Carabineros responded to their concerns, how well the police treated them, and how they evaluated the police’s response to their concern. After a sixmonth period of data collection, the Carabineros’ Office of Planning and Research used the survey data to develop performance targets, aiming to use community feedback to improve police performance. Through measures like these, Chilean policing strategies have sought to respond to communities’ needs. Of course, policing strategies cannot merely bend to the will of the majority, but soliciting community feedback and subsequently developing performance goals encourage a positive feedback loop between police and the communities they serve. Indeed, scholars have found that surveys of citizens after their contact with police can improve police handling of interactions with the public, and positive encounters with individual officers can boost trust in the institution as a whole (Bradford, Jackson, and Stanko 2009).

44

Making Police Reform Matter in Latin America

While Chile was a pioneer of community policing programs, some external assessments of programs such as the Quadrant Plan identified areas for improvement. FPC conducted an independent evaluation of the Quadrant Plan and concluded that it was difficult to find evidence that communities with the plan reported better outcomes than communities without it, and they argued that this indicated “that the presence of resources would be more relevant than the management model implied by the Quadrant Plan in the communes” (FPC 2012). In addition, this independent evaluation concluded that “of the 117 barracks studied, 85% of the supply reductions are explained by a decrease in human resources. If it is considered that this occurs in a context of growing demand, then it may reflect inefficient human resource management” (FPC 2012, 18). This evaluation was not published until 2017, and the Carabineros responded by objecting to the study’s methodology. Another evaluation established that “the current design of the Program does not present actions aimed at a community policing approach (an approach that inspired the creation of the program) and the general design has basically reinforced a greater endowment of police resources in those communes that participate in the Quadrant Plan” (DIPRES 2007, 17). Even after these evaluations, the Carabineros did not alter the budget of the Quadrant Program; on the contrary, the program received steady investments to continue and strengthen it. The Carabineros did make some changes, however. In 2011, the police inaugurated the Modelo de Integración Carabineros-Comunidad (MICC), or the Model of Community–Police Integration. Inspired by the models of community-oriented policing and problem-oriented policing that emerged in Uruguay (see Chapter 9),4 MICC aimed to increase community patrols and dialogue with neighbors, collaborate with local businesses to improve surveillance, train the community to prevent crime, and identify the main problems of each community through surveys and community meetings. MICC also prioritized addressing minor offenses that were of particular annoyance to neighbors and engaging in nonviolent conflict resolution. Under the MICC, citizen participation has increased through working groups consisting of two representatives from each community organization, using the quadrant division already established by the Quadrant Plan. This marks a distinction between direct beneficiaries (the organized community that participates) and indirect beneficiaries (the rest of the community). The MICC makes it easier for citizens who are already organized to participate, thus opening the possibility of leaving out community members who do not typically participate in formal organizations but who might wish to do so in matters of public security. Building on previous community organizations and networks obviously has advantages but tends to replicate existing power dynamics in the community and does not facilitate the participation of new actors. The indirect beneficiaries have the oppor-

Chile

45

tunity to provide some input, however, as they tend to be consulted through regularly conducted surveys. Thus, the indirect community can have a voice, but not direct participation in MICC initiatives (RAIS and Instituto de Sociología UC 2012). The MICC model has undergone multiple evaluations that have proven quite positive (GFK-Adimark 2013). An early study identified a boost to community participation and cooperation, as well as enhanced trust and credibility in the Carabineros. However, the community does not participate in the evaluation of police action, nor of the MICC model itself. Still, the MICC has been credited with better targeting patrols and surveillance in territories. In general, inhabitants of communities operating under the auspices of MICC policing strategies tend to voice strong support for several actions they think would improve security in their neighborhoods: increased Carabinero surveillance, development and implementation of school programs to prevent violence, improved lighting in streets and public spaces, and initiatives to increase youth employment (GFK-Adimark 2013). Such calls echo our observation in Chapter 1, where we noted that investments in social protection programs can address not just criminality but the root causes of criminal careers, as these cross-cutting reforms can buttress one another and provide a more holistic response to addressing crime.

Curating the Carabineros’ Image Community-oriented policing emerged as part of the Carabineros’ efforts to refurbish their image after the end of Pinochet’s dictatorship. Chile has invested not just in the resources of the Carabineros, but also in how the they present themselves to the public. Indeed, Bonner (2013) conducted interviews with fifty police experts in Chile and found they often attributed the positive public image of the Carabineros more to effective communication strategies instead of the results of reform. The Carabineros are meticulous in cultivating the image they portray to society. Some images are cartoonish, targeting younger audiences in particular, and aiming to present the police as approachable and friendly, even under the banner of orden y patria (order and country). Scholars have noted the important role of such legitimizing images in cultivating public support for political institutions. For example, in a study of public support for high courts around the world, Gibson, Caldeira, and Baird found that “Exposure to legitimizing judicial symbols reinforces the message that ‘courts are different’ and owing to such difference, courts are worthy of more respect, deference, and obedience—in short, legitimacy” (1998, 44). Applying this theory to the Chilean case, one could argue that high levels of trust in the Carabineros could be tied to this consistent use of

46

Making Police Reform Matter in Latin America

legitimizing symbols. Bonner (2018) finds support for this argument, as she analyzed the ways the Carabineros present themselves to the public and found that the police designed and implemented effective communication strategies to showcase their achievements, portraying themselves as heroes protecting the public and occasionally as victims of a system that sent them to fight crime practically unarmed. In addition to a well-financed police academy, Chile maintains a Historical Museum of the Carabineros, which introduces police recruits and the public to the evolution of policing practices over time. The exhibits use state-of-the art technology to show the historical development of the Carabineros over time, emphasizing the close contact that they have maintained with the communities they serve, tracing contemporary policing practices to those of the early nineteenth century. The exhibit glosses over the Carabineros’ role during Pinochet’s dictatorship, noting only that the police “attempted to deal with troubling times.” Adjacent to this museum is an exhibition hall, equipped to host cultural events for the community. Of course, legitimizing images alone are not sufficient to generate high levels of public support. Personal interactions with police, perceptions of performance, and evaluations of the public security situation are also important. For three decades of democratic rule, the Carabineros largely reinforced their self-curated public image with such positive interactions. Starting in 2017, a series of scandals steadily eroded their image.

The Police’s Political Crisis While Chile has been lauded for its community-oriented initiatives, other issues such as military justice and accountability remained highly problematic. These issues have persisted across practically all the center-left governments that have governed Chile since the return of democracy. Although governments have recognized the problems military justice poses for accountability, they have largely been unable (or unwilling) to tackle it once in power. In 2010, Sebastián Piñera won the presidency, leading the first right-wing coalition since the restoration of democracy. In his first year in office, Piñera introduced a legal modification to challenge military justice for Carabineros. December 30, 2010, marked the passage of Law 20,477, which excluded the trial of civilians (including Carabineros) by military justice tribunals. Despite this legal change, substantive accountability challenges remain. Until recently the justice system has been reluctant to sanction Carabineros accused of misconduct or abuse of power. The strong institutional autonomy of the Carabineros has insulated it from many reform efforts. Indeed, major initiatives to strengthen community– Carabinero ties overwhelmingly emerge from the police institution itself, rather than at the behest of the incumbent government or civil society. Even

Chile

47

changes resulting from the Quadrant Plan originated from the Carabineros, as changes in management, evaluation, and especially distribution of performance indicators all came from within the institution. To confront increases in insecurity, the Carabineros formulated their own proposals for internal redefinition and community engagement, not conferring with the incumbent government or civil society until the very end of this process, when there were only a few details of definition left to elaborate. In contrast, in other police forces in Latin American, changes are typically led by external forces concerned about corruption scandals, the use of force, or the general perception of inefficiency. These high levels of autonomy can also lead to structural inefficiencies. For example, the Carabineros created a criminal analysis department in November 2011 to study and analyze criminological variables and satisfy demands for strategic and operational information. This new department was consolidated internally, without any efforts to link its work with that of the government agencies charged with the same tasks in the Ministry of the Interior and Public Security. Problems have arisen with these high levels of autonomy, as well as with the government’s reluctance to establish specific control mechanisms for the Carabineros. The saying “too good to be true” became a reality during the second government of Michelle Bachelet (2014–2018), when several scandals began to show the negative repercussions of the government’s neglect of supervising police work. Although there are multiple incidents, we highlight four to understand the legitimacy crisis of the Chilean Carabineros. First, a young university protester was left in a coma due to a blow directly to the head from a police water cannon. Despite public outrage and initial government annoyance, the Carabineros did not provide further information about what happened, illustrating the limitations on transparency and accountability that have lingered. Various nongovernmental organizations had called attention to the excessive use of police force, especially in the context of student protests, but this case put a face to a practice that had long been criticized. Second, in September 2017, the Carabineros launched Operation Hurricane, an ostensibly delicate intelligence operation that resulted in the arrest of eight Mapuche community members who were allegedly involved in an illicit terrorist association with international connections. The Carabineros claimed they gathered incriminating intelligence from social networks and telephone messages. Later scrutiny of this operation uncovered a montage of disturbing details orchestrated by the Police Intelligence Directorate of the Carabineros. In early 2018, the Public Ministry recognized that the intelligence was false, and the institution had fabricated the messages and lied to authorities. In the words of the national prosecutor, it was an “irregular, very serious and unprecedented” situation.5 Still, the government of President Bachelet refused to remove the police officers.

48

Making Police Reform Matter in Latin America

The second government of President Piñera (2018–2022) was quick to accept the resignation of the general director of the Carabineros, a move widely viewed to signal seriousness and strength. However, at the end of 2018, police action in southern Chile ended with Camilo Catrillanca, a young Mapuche, shot in the back. The Carabineros initially called the incident a confrontation, but a confusing series of statements showed that it was murder without a confrontation. These incidents of excessive force against Mapuche people eroded citizen support for the government and revealed the problems of excessive use of police force and lack of institutional accountability. The accusations against the officers involved in this incident highlighted how the police targeted some groups of Chilean society, such as the Mapuche, with discriminatory and violent treatment. Third, while the Carabineros have shown zero tolerance for minor acts of corruption involving routine police work in the streets, the higher command benefited from another type of treatment. In 2015–2017, the so-called Pacogate investigation uncovered an illicit structure consisting of more than one hundred top-ranking Carabineros officials who fraudulently withdrew more than US$40 million from institutional accounts for their own benefit. Investigations unearthed elaborate mechanisms to defraud the treasury and evade institutional controls for over a decade. Indeed, the institutional accounts seemed healthy and in compliance with regulations, leading the Carabineros to request larger budgets each year. However, more in-depth audits revealed multiple mechanisms for fraud and embezzlement. A report from the Chamber of Deputies verified the facts of the case and criticized the limited control mechanisms of the Ministry of the Interior, as well as the inattention to the comptroller general’s reports warning of irregular transactions. This process destroyed the image of institutional probity and professionalism at the highest levels of the Carabineros and generated internal institutional cracks that disrupted the previously lauded police discipline. However, it was not immediately clear in the aftermath of this scandal how these highly publicized corruption crimes at the top levels of police command would affect public perceptions of police patrolling the streets or the actors with whom most citizens have daily interactions. The final crisis for the Carabineros’ reputation came during the estallido social (social uprising) of October 2019, when strong social mobilizations and protests erupted and turned the streets into a space for venting pent-up frustration (Araujo 2020). While most protesters were peaceful, some did engage in violence by burning metro stations and targeting police officers. Police responded to protester violence with heightened levels of repression, which escalated the level of violence and led to condemnation from human rights groups. To respond to growing protests, the Carabineros unleashed excessive force against demonstrators, leading to widespread reports of police abuse and violence against protesters and detainees. United Nations investigators denounced the “fundamentally repressive

Chile

49

manner” of the Carabineros’ reactions to demonstrations and called for prosecution of the army and police for human rights violations.6 Although there are multiple interpretations of the causes of the social uprising, as well as the acts of violence that were unleashed and sustained for nearly four months, one element is clear: police action served to increase the spiral of violence (Dammert 2020). The human rights violations detailed by the reports of the National Institute of Human Rights and Human Rights Watch acknowledge violent, humiliating, and often outrageous police action, and found the police ill-prepared to fulfill their central roles of maintaining order and respecting citizens’ rights (HRW 2020b). The images and apparent lack of institutional control shocked the country, resulting in sharp declines in public trust in the police and leading many citizens to question the legitimacy of the Carabineros. Police response to the social uprising led to another critical juncture, as evidence of the police crisis generated—for the first time since the return of democracy—widespread consensus across the political spectrum on the need for police reform. Progress has been tepid, as the levels of autonomy and political power of the police institution are strong and transversal, but one thing is certain: the current police model, which requires fundamental changes, has been broken. Calls for an overhaul of the Carabineros became part of a broader political movement toward social, economic, and political change. In May 2021, Chileans elected delegates for a constitutional assembly, which designed a new constitution to replace the one written by Pinochet in 1980. Following a failed referendum on a proposed constitution in September 2022, Chilean lawmakers announced plans in December 2022 to convene a new commission of constitutional advisers and experts. If these efforts result in a new constitution, this would mark a new critical juncture for police reform. We turn to examine public perceptions of the Carabineros against this backdrop of reform and institutional crisis. While the police faced a crisis in public confidence in the aftermath of the 2019 social uprising, the institution has still benefited from Chile’s reputation as the safest country in Latin America. The public has expressed grave misgivings about police misconduct and excessive use of force, but these sentiments have not necessarily jeopardized public support for the long-standing public security initiatives that prioritize prevention and community involvement.

Crime Trends and Public Perceptions The transition toward more community-oriented policing strategies occurred against a backdrop of very low crime rates. Chile is widely considered one of the safest places in Latin America, as homicide rates have remained well below regional trends. As Figure 3.1 illustrates, while there was a slight

50

Making Police Reform Matter in Latin America

increase in homicide rates in 2014 and 2020, these rates remain far below regional and even international trends. This context is important for understanding the relative success of Chile’s early implementation of communityoriented and preventive public security policies. Because levels of violent crime have historically been low, it created more space for such public security policies, particularly compared with other countries in the region where homicide rates are among the highest in the world. Indeed, it is one thing to emphasize prevention and community orientation when crime rates are low and another when they are high and police face well-organized and armed opponents, such as the maras (gangs) in El Salvador. That is not to say that community-oriented policies can only work when crime rates are low, but the Chilean context certainly facilitated their adoption and posed fewer obstacles to their achieving successful outcomes. Although homicide rates have been low, rates for other types of victimization have recently been closer to the regional average. In LAPOP surveys conducted in 2006–2014, rates of self-reported victimization declined steadily and significantly, reaching their lowest point of 11.4 percent in 2014. However, in 2016 and 2018–2019, rates increased sharply, nearing the regional average. Figure 3.1

Homicide Rates per 100,000 People in Chile (2005–2020)

Source: World Bank World Development Indicators.

Chile

51

National citizen security surveys in Chile echo these findings. The Encuesta Nacional Urbana de Seguridad Ciudadana (ENUSC; National Urban Survey of Citizen Security) measured the percentage of households reporting victimization in the past year (in contrast to the previous LAPOP victimization measure that focuses only on whether a survey respondent experienced victimization). According to ENUSC’s surveys, overall household victimization rates decreased between 2008 and 2014 (from 31.9 percent to 23.5 percent), before rebounding a bit to 26.4 percent in 2015. From 2015 to 2019, roughly one quarter of households reported victimization of at least one crime. Still, the highest percentage of these crimes are thefts and do not include the use of violence. When looking at these national statistics, it is important to keep in mind that Chile’s high levels of inequality can keep many citizens from identifying with national averages. For example, a communal survey demonstrated that in the peripheral communes of Santiago, more than 69 percent of the population reported hearing or witnessing shootings always or almost always (ENUSC 2016). In contrast, in the more affluent communes of the city, such as Las Condes, the percentage does not exceed 4 percent. Even though respondents lived less than thirty kilometers away from one another, they experienced profoundly different realities in terms of public security. The concentration of criminal activity in some neighborhoods is not new and not unique to Chile, of course. However, Chile’s national reputation as a security oasis and policing success story tends to hide the problems of these vulnerable neighborhoods, rendering them almost invisible. Furthermore, even though many parts of Chile conform to national trends of low crime, perceptions of insecurity tend to be high. Indeed, these perceptions have been widely studied in Chile precisely because of the apparent gap that exists between lower levels of crime and high levels of fear of crime (Dammert and Malone 2003, 2006). Even when victimization decreased between 2008 and 2014, feelings of insecurity remained quite high. As Figure 3.2 illustrates, rates of public fear of crime decreased between 2008 and 2014, matching the drop in victimization rates. However, this decrease is substantively quite small, and overall, Chileans tend to report feelings of insecurity that are on par with the regional average, despite the fact that crime rates (particularly violent crime) are comparatively much lower.7 In 2008, Chilean fear of crime was slightly higher than the regional average of 2.3, meaning that Chileans were slightly more afraid of potential victimization than their counterparts in far more violent circumstances, on average. From 2010 to 2019, the average levels of fear of crime in Chile were just slightly below the regional averages, which is unusual given Chile’s very different security context. Scholars have generated various hypotheses to try to explain this gap between perceptions and reality (Dammert and Malone 2003). The role of

52

Making Police Reform Matter in Latin America

Figure 3.2

Fear of Crime in Chile (2006–2019)

Source: LAPOP’s 2006–2019 AmericasBarometer.

the media has come under scrutiny, as the media often sensationalizes crime coverage and portrays a generalized sense of impunity for criminals, which leads many citizens to view the justice system as incapable of holding criminals accountable. Scholars have also noted that crime can act as a proxy for other types of insecurities that Chileans face in their daily lives (Dammert and Malone 2003, 2006). Although this transference of insecurities from other parts of life to fear of crime occurs in other countries (e.g., Garland 1996), given the high levels of individualism and low levels of trust that pervade Chilean society, such trends are exacerbated, and crime becomes even more of a scapegoat. Furthermore, the discrepancies among communities that we noted earlier can heighten feelings of insecurity, as citizens in secure communities can feel threatened by levels of violence nearby and view such insecurities as evidence of social deterioration and threats to the Chilean way of life. These high levels of fear of crime have driven a clear political agenda in Chilean public policy, as throughout the decades following the return of democracy, citizens have ranked crime among the top three areas of concern. Indeed, for more than half of this time period, crime ranked as the first concern.8

Chile

53

Possibly a very distinctive element of the Chilean case is that citizens’ demands for better security are not framed as demands for a better police force, as is often the case in other Latin American countries. Instead, these demands tend to emphasize the limitations police face to do their work and apprehend criminals. Citizen criticisms tend not to fault the police for security failures, but the alleged garantismo of the justice system, which critics argue privileges the rights of criminals over the well-being of crime victims and society more broadly (Dammert 2019). Thus, in Chile, fear of crime often goes hand in hand with high levels of public trust in the police. Despite high levels of trust in police, Chileans retain very pessimistic views about the ability of the justice system to apprehend and prosecute criminals. On average, according to LAPOP’s 2006–2019 AmericasBarometer, Chileans reported that they had “just a little” trust that the justice system will punish guilty parties; from 2006 through 2012, levels of trust never rise above 2.2 on a 4-point scale. Indeed, levels of trust that the justice system would punish the guilty party decline further from 2012 through 2019, to a low of 1.8. In the only year that LAPOP asked if the police would apprehend the guilty party, the police scored on average slightly higher than the justice system overall (2.4 compared with 2.1).

Perceptions of Police Although it is difficult to disentangle causality and link Chilean reforms to levels of trust, it is clear that trust in the police has been historically high. As Chapter 2 demonstrated, Chile has one of the highest levels of trust in the police in the region, and Figure 3.3 illustrates that this trust has remained much higher than trust in other political institutions from 2006 through 2019. Figure 3.3 carries an important caveat, however, as the last round of LAPOP survey data was collected between January 19 and March 28, 2019, prior to the excessive police response to the estallido of October 2019. While trust in the Carabineros remains higher than that of other Chilean institutions and is high when compared to trust in police forces in the region, Figure 3.3 demonstrates that these levels of trust decline significantly in 2016–2017 and again in 2019. LAPOP conducted the 2016–2017 wave of its survey in Chile between March 17 and May 30, 2017, so these declines in public trust in police correspond to the breaking of the Pacogate embezzlement scandal. The 2018–2019 survey wave registers the further erosion of trust in Carabineros that followed on the heels of the Operation Hurricane scandal and the use of excessive force against student protesters. However, the 2018–2019 data point occurred prior to the more extensive and widely publicized police abuse of power that transpired during the

54

Making Police Reform Matter in Latin America

Figure 3.3

Trust in Police and Other Political Institutions in Chile (2006–2019)

Source: LAPOP’s 2006–2019 AmericasBarometer.

estallido of October 2019–March 2020. Still, we see that earlier scandals corresponded to declining levels of trust in the Carabineros. Additional surveys expand on these findings from LAPOP. According to monthly survey data from the Chilean polling firm CADEM, there was a decisive decline in public approval of the Carabineros from February through August 2017, when media coverage of Pacogate was quite extensive.9 Support of the Carabineros dropped thirty-two points during this time frame (from 72 percent to 45 percent), before rebounding in June 2018 (to 63 percent). Trends in public approval of the Carabineros remained quite erratic throughout 2018 and 2019, indicating the toll the many publicized scandals took on the institution. Each time support for the Carabineros recovered, it faced another setback. In the aftermath of the estallido and corresponding police brutality, in December 2019 public approval of the Carabineros reached its nadir, registering only 35 percent. Interestingly enough, approval of the Carabineros began to improve in March 2020, and this uptick could be linked to optimism that the proposed constitutional and police reforms could redress the problems of 2015–2020, as well as a rally-

Chile

55

around-the-flag affect due to early reports that Chile was managing the Covid-19 pandemic well (although these initial positive responses to Covid19 were not sustained as the pandemic dragged on). Part of the reason that trust in the Carabineros has been historically high is because of their professional conduct and interactions with citizens (Malone and Dammert 2020). Chapter 2 documented that levels of police bribery were significantly lower in Chile (and Uruguay) than other Latin American countries. Figure 3.4 examines police corruption over time and illustrates that these low levels of bribe solicitation further declined between 2006 and 2014. These levels could help the police recover from the toll of recent scandals, as consistently low levels of corruption on the streets gave citizens another perspective by which to evaluate the work of the Carabineros. Other indicators reaffirm the positive perceptions the public has held of the Carabineros. Prior to the scandals of 2015–2020, when asked about the primary role of the police, nearly two thirds of Chileans reported that the police served to protect communities from crime. Indeed, in 2008, only 21 percent of citizens reported the police were primarily involved in crime, and this number decreased to 16 percent in 2012.

Figure 3.4

Police Bribery in Chile (2006–2019)

Source: LAPOP’s 2006–2019 AmericasBarometer.

56

Making Police Reform Matter in Latin America

Respondents also tended to register satisfaction with police performance in their immediate neighborhoods, although there are differences between people who had been victimized by crime and those who had not. In 2014, LAPOP measured respondents’ satisfaction with police performance in their own neighborhoods, ranging from (1), very unsatisfied, to (4), very satisfied. In 2014, 62 percent of people who had not been victimized by crime in the past year were either satisfied or very satisfied with police performance in their neighborhoods. Among those who had been victimized, however, these levels of satisfaction dropped significantly to 47 percent. When asked about the amount of time police need to respond to citizens’ calls, Chileans were quite positive. On average, respondents perceived police as quite responsive in LAPOP surveys, reporting that they would expect police to arrive roughly within thirty to sixty minutes if they called to report a burglary. This estimated response time remained quite consistent between 2014 and 2019, indicating that the Pacogate and Operation Hurricane scandals did not seem to have a dramatic effect on measures of police efficiency and responsiveness, according to this indicator. These findings signal the potential for the Carabineros to rebuild their image. If citizens register positive evaluations in the area of crime prevention and work in the community, the Carabineros can potentially use these high marks to compensate for the weakness in other areas. However, if problems continue to emerge in these other areas, it is unlikely that the Carabineros will be able to rebuild their image in the long run. As LAPOP conducted its last survey wave before the social uprising; those data preclude an examination of how the Carabineros’ actions during the social uprising shaped public attitudes toward the police. For a deeper examination of public attitudes after the social uprising, we conducted a face-to-face representative survey in the Santiago metropolitan region between December 2019 and February 2020 (n = 932).10 Although the original goal of the survey was to assess police legitimacy in Chile, given the events of October–December 2019, we added several questions related to perceptions of the social uprising and attitudes toward the Carabineros. In the 2019 survey, it was clear that the scandals and police action during the social uprising had taken a toll on public attitudes toward the Carabineros. We asked, “In your opinion, three years ago was the institutional reputation of the Carabineros better, the same, or worse than today?” An overwhelming majority—72.3 percent of respondents—reported that the Carabineros’ image was better three years ago compared with today.11 Furthermore, only 32.9 percent of respondents agreed that the Carabineros represented Chilean society, and only 31.1 percent stated that the Carabineros had the same idea of right and wrong as they did.12 Respondents also registered low levels of trust in the Carabineros in 2019. We asked respondents, “I am going to mention several institutions. On a scale

Chile

57

of 1 (not trustworthy at all) and 7 (very trustworthy), how much trust do you have in each institution?” Average levels of trust in the Carabineros did not approach the midpoint of the scale, indicating that most people regarded them as untrustworthy after the social uprising. As Figure 3.5 demonstrates, the Carabineros ranked far below the firefighters, and significantly lower than the municipality and military. Only the incumbent government and the national congress ranked significantly lower than the Carabineros – two institutions that were also widely blamed for failures to respond to citizens’ concerns during the estallido (Dammert and Malone 2020). These data from our 2019 survey indicate that as a whole, the institution of the Carabineros suffered a sharp decline in citizen trust following the 2019 estallido. In contrast, when people recalled their own interactions with the Carabineros, they overwhelmingly found them to be respectful and courteous. Few respondents described the Carabineros as indifferent in their encounters, and even fewer as violent or aggressive. While the institution as a whole has suffered from a reputational decline, it does appear that in routine interactions, people tended to have positive interactions with the Carabineros on average. We also look more specifically at the experiences of people in our survey who indicated that they had been victimized by a serious crime. Among the people who identified themselves as victims of a severe crime, 75.9 percent

Figure 3.5

Trust in Carabineros and Other Institutions in Chile (2019)

Firefigters Municipality Military Carabineros

How much trust do you have in the following institutions?

Private Security Firms Government Congress Error bars: 95% CI

not at all trustworthy

Average Levels of Trust

Source: 2019 Carabineros Legitimacy Survey.

very trustworthy

58

Making Police Reform Matter in Latin America

stated that they reported the crime to the police, a very high percentage that demonstrates a willingness to turn to the Carabineros to address problems of victimization. While a high percentage of victims reported the crimes to police, levels of satisfaction with the process varied considerably, as Figure 3.6 demonstrates. Evaluations of the treatment and attention of the Carabineros during the crime reporting process were at the midpoint of the scale; crime victims were a bit ambivalent about the type of treatment they received. Crime victims registered slightly more negative evaluations of the speed of the crime reporting process, and low levels of satisfaction with the final results. Still, these lower levels of satisfaction could be due to the nature of crime reporting itself – the experience of victimization may lead people to have more pessimistic views overall, and naturally have more negative views given the trying circumstances of the crime itself. For a more in-depth analysis of how the estallido shaped public trust in the Carabineros, we conducted a regression analysis, with our 2019 measure of trust in police as the dependent variable (as depicted in Figure 3.5). Using ordinal logistic regression, we regressed trust in the Carabineros on socio-economic variables (age, education, gender, and urbanization), to determine if trust in the Carabineros varied across socio-economic groups. Given the high levels of inequality we noted earlier, it is imperative to

Figure 3.6

Crime Victims’ Evaluations of Crime Reporting Process in Chile (2019)

Carabineros’ treatment

Carabineros’ attention After you reported the crime to police, how would you rate the following aspects of the process?

Speed of the proceeding

Results of the reporting Error bars: 95% CI terrible

average

Source: 2019 Carabineros Legitimacy Survey.

excellent

Chile

59

determine whether these inequalities lead to different experiences with the Carabineros, and subsequently different levels of trust. While we did strive to include a measure of income, our survey question asking respondents to self-report their monthly income yielded many missing values, compromising the validity of this indicator. We were able to include a measure to gauge employment status, however. We also include a measure of citizens’ perceptions of the estallido. If people viewed the estallido in negative terms, for example, that could lead them to justify police retaliation against protestors.13 We also assess the effect of respondents’ direct experiences with the Carabineros, as the literature finds that prior experiences and encounters with police can condition perceptions of procedural justice and expectations for future encounters (Belur 2010; Reisig, Mays, amd Telep 2018; Geller and Fagan 2019).14 Finally, given the long-standing ideological divides in Chile, and the role that ideology can play in shaping levels of trust in law and order institutions, we include a measure of political ideological orientation. Table 3.1 reports the results of this analysis. Our regression analysis reveals that perceptions of the estallido play a large role in determining levels of trust in the Carabineros. Those who

Table 3.1

Trust in Carabineros After the Estallido Social (2019)

Independent Variables Age Education Ideology (1) left – (10) right Perceived impact of estallido Men Urban resident Employed Direct contact with Carabineros (p21) N Nagelkerke pseudo R2

Model 1 0.029*** (0.005) 0.051 (0.037) 0.084*** (0.022) −0.701*** (0.086) −0.081 (0.151) −0.486** (0.170) 0.015 (0.156) 0.308* (0.148) 680 0.242

Notes: Table reports ordinal logistic regression estimates, with standard errors in parentheses. *p < 0.05; **p < 0.01; ***p < 0.001.

60

Making Police Reform Matter in Latin America

thought the estallido would have positive results for Chile were significantly less likely to trust the Carabineros. Residents of urban areas, which witnessed more political violence than other areas, were significantly less likely to trust the Carabineros. The rest of the socioeconomic variables performed unevenly, however. Age was significant, with older respondents registering higher levels of trust than younger ones. The remaining socioeconomic variables were insignificant, although it is important to note that there was a degree of multicollinearity in this model, and ideology had a strong and significant correlation with education (r = −0.165, p >0.01). Table 3.1 does point to one reason for optimism: people who had direct contact with the Carabineros were significantly more likely to trust them. Our survey also allows us to examine how public trust in police can shape public acceptance or rejection of police use of force against protesters. Our 2019 survey included four questions measuring public attitudes toward police use of force during the estallido. Respondents were asked if they (1) approved, (2) neither approved nor rejected, (3) rejected the following types of Carabinero violence against citizens: • “Carabineros throw tear gas canisters against citizens protesting violently.” • “Carabineros respond to citizens who verbally abused them by hitting them with their baton.” • “Carabineros respond to citizens who physically assaulted them by hitting them with their baton.” • “Carabineros make violent arrests and injure citizens.” To measure attitudes toward the Carabineros’ violence against citizens, we created an index using these four survey questions.15 The results show that tolerance for Carabinero violence against citizens is not high, as respondents overwhelmingly rejected this excessive use of force. The few who support Carabineros violence against protesters also tend to have higher levels of support for the police. Indeed, as Table 3.2 indicates, this relationship holds even when controlling for socioeconomic variables, ideology, and prior experiences with police. Those with high levels of trust in the Carabineros are significantly less likely to reject their excessive use of force against protestors.

Conclusion The Carabineros of Chile is not an institution comparable to any other police force in Latin America. For decades it consolidated a strong, professional, educated, and limited group of officials who governed the institution

Chile Table 3.2

61

Trust in Carabineros and Attitudes Toward Police Violence (2019)

Independent Variables Age Education Ideology (1) left – (10) right Perceived impact of estallido Trust in police (1) completely untrustworthy – (7) very trustworthy Men Urban resident Employed Direct contact with Carabineros (p21) N Nagelkerke pseudo R2

Rejection of Police Violence 0.010 (0.006) 0.036 (0.047) 0.088** (0.028) 0.242* (0.112) −0.331*** (0.051) −0.186 (0.186) 0.210 (0.208) 0.022 (0.191) −0.715*** (0.188) 665 0.135

Notes: Table reports ordinal logistic regression estimates, with standard errors in parentheses. *p < 0.05; **p < 0.01; ***p < 0.001.

and its processes. Institutional stability greatly facilitated the reform process, as a center-left coalition governed Chile from 1990 to 2010. Even when Piñera disrupted the coalition’s winning streak with his presidential victory in 2010, his administration continued the process of police reform, aiming to improve accountability. It is important to remember that this process of police reform occurred against a backdrop of high levels of inequality, although later community-oriented programs such as Barrios Vulnerables and MICC did seek to couple police engagement with the community with investment in social programs to address some of the root causes of crime. The Carabineros have maintained a strict hierarchy for both officials and their civilian staff and protected the autonomy of their organizational structure. For most of the postdictatorship period, incumbent governments and the legislature have been hesitant to encroach on this autonomy, and before 2017, positive citizen evaluations further insulated the Carabineros from critiques and calls for reform. Public perceptions that the Carabineros were effective at controlling crime and not as susceptible to petty corruption

62

Making Police Reform Matter in Latin America

contributed to high levels of trust in the institution. Indeed, high citizen trust in the police allowed the consolidation of spaces where the Carabineros held absolute discretion outside the world of politics. Former Minister of the Interior José Miguel Insulza illustrated this point well in 2018, when he noted, the “Carabineros command themselves” (CNN Chile 2018). These high levels of autonomy fostered police opacity, as the Carabineros were able to elude political and civil control. The scandals of the late 2010s and excessive use of force during the 2019 estallido overhauled this status quo, however. As Chileans work on rewriting their constitution, this could be another critical juncture—a moment to engage the Carabineros in real reform, particularly by establishing stronger mechanisms for horizontal accountability and civilian oversight. Although most Chileans still report positive interactions with individual Carabineros (though this varies across socioeconomic lines), evaluations of the institution were damaged by scandals and police violence against protestors.

Notes 1. To be sure, policing practices alone cannot receive all the credit for low rates of violent crime. As Orozco (2015) notes, socioeconomic structural factors can reduce social tensions and economic deprivation, which in turn lead to lower levels of violence. 2. For an overview of this period of Chilean history, see Oppenheim (2007) and Borzutzky and Oppenheim (2006). 3. Not all of these community policing programs have been successful. For example, in 1998 the national government encouraged the formation of citizen protection committees in the greater Santiago metropolitan area, designed to create community alert systems and reinforce the protection of individual homes against crime. The Secretaría General de Gobierno (General Government Secretary) trained local leaders to spearhead these committees, encouraging them to network in their communities and with the Carabineros and officials in municipal governments on preventive crime measures. After the initial boost following implementation, these committees languished, and the program subsequently dissolved. 4. As we note in Chapter 9, problem-oriented policing emerged as a public security strategy to address increased crime in urban areas of Uruguay. 5. The full text of the statement is available at: https://cooperativa.cl/noticias /pais/region-de-la-araucania/fiscalia-confirmo-que-carabineros-presento-pruebas -falsas-en-la/2018-01-25/190934.html. 6. For the full text of this statement, see: https://www.ohchr.org/es/2019/12 /un-human-rights-office-report-chile-crisis-describes-multiple-police-violations -and-calls. 7. In LAPOP’s surveys, fear of crime is measured on a four point scale, and the Latin American regional averages for fear of crime are: 2.3 (2008); 2.3 (2010); 2.1 (2012); 2.4 (2014); 2.0 (2016–2017); 2.5 (2018–2019). 8. According to the Estudio Nacional de Opinión Pública del Centro de Estudios Públicos (CEP). See https://www.cepchilde.cl/encuestaCEP. 9. Plaza Pública–CADEM. See https://cadem.cl/plaza-publica/.

Chile

63

10. The sampling frame included inhabitants over eighteen years of age from urban locations and rural areas of the metropolitan region. To carry out the stratification, clusters were created using two variables at the community level that are relevant to the study: crime rate per 100,000 inhabitants and number of Carabineros per 100,000 inhabitants. With this information, each commune was classified into one of the four possible clusters (A, B, C, D). Subsequently, districts were surveyed, considering the proportion of each cluster at the level of each province. For each community drawn, blocks were chosen at random considering the socioeconomic diversity of the community (as measured by cartography and census data (INEC 2017). The sampling error is 3.2 percent for a 95 percent confidence interval. 11. Only 16.9 percent of respondents stated that the Carabineros’ reputation was the same as three years ago, and 10.8 percent stated that it was worse three years ago. 12. The exact text of these survey questions was: “Do you think the Carabineros is an institution that represents Chilean society? (1) yes; (2) no.” “Do you think that the Carabineros have the same idea of right and wrong as you do? (1) yes; (2) no.” Sixty-seven percent of respondents answered “no” to the first question, and 68.9 percent responded with “no” to the second question. 13. To measure perceptions of the estallido, we asked respondents what they thought the long-term consequences would be: “Do you think that this movement or social uprising . . . (1) has hurt the country; (2) has neither helped nor hurt the country; (3) has favored the country.” Original survey responses were (1) has favored; (2) has hurt; (3) has neither helped nor hurt. We recoded these original responses for a more intuitive interpretation of this variable. 14. To measure direct contact with the Carabineros, we asked respondents, “Do you remember having direct contact with the Carabineros? (1) yes; (0) no.” When asked to describe their prior experiences in their own words, most respondents gave positive descriptions, with 60 percent stating that the experience was good, efficient, normal, or routine, while the remaining 40 percent provided negative feedback. 15. The Cronbach’s alpha for this scale (0.776) indicated that there was a high level of interitem reliability. We calculated this scale using a means formula: if respondents answered three of the four questions, we computed their average responses across all valid items. The resulting scale ranged from (1) approved violence in all three scenarios through (3) rejected violence in all three scenarios.

4 Colombia: Reform Amid Conflict and Instability

POLICE REFORM IN COLOMBIA HAS TAKEN PLACE AGAINST AN unusual backdrop of armed conflict and stark rural and urban inequalities. In the 1980s and 1990s, Colombia was infamous for its high homicide rate, largely due to the dominance of organized criminal groups and drug trafficking. During this time, the Policía Nacional de Colombia (PNC) were widely dismissed as corrupt, violent, and often guilty of perpetrating crimes rather than solving them. By some accounts, the PNC was close to institutional collapse in 1993 (Esparza 2022). This confluence of extraordinarily high rates of violent crime and the potential collapse of the police created a critical juncture, in which political leaders were open to extensive reforms given the dire situation. Because of its pivotal role in the international drug trade, Colombia received extensive international attention to help overhaul the PNC and public security policies. Violent crimes such as homicides began to decline in the early 2000s, but these gains in public security were offset by the advances of the Fuerzas Armadas Revolucionarias de Colombia (FARC) and other armed groups, which began to expand their territorial occupations. Once again, international actors allocated attention and resources to address security in Colombia, although such international aid has not been without controversy. Under the $10 billion Plan Colombia (2000–2015), the United States funded the largest international intervention, investing heavily in security to eradicate the cultivation of illicit drug crops, dismantle drug trafficking operations, and defeat the FARC. Colombia has transformed its policing practices and introduced some community-oriented reforms in urban areas while confronting an armed insurgency and the drug trade. Domestic and international reform efforts did lead to improvements in policing and security, and observers applauded 65

66

Making Police Reform Matter in Latin America

the transformations of the PNC, particularly the improvements in professionalization, management, and corruption (Frühling 2003). These improvements in police effectiveness at controlling crime and reduced corruption led the public to perceive police work more positively. At the subnational level, some municipal public security programs were matched with cross-cutting investments in social welfare, which we noted in Chapter 1 can increase the viability of reforms by addressing some of the root causes of criminality. While the police retained its military structure and continued to report to the Ministry of Defense, several changes strengthened the linkage between police and local communities. In urban areas, Colombia introduced some community-oriented policing practices, although heavily armed police stationed in major urban areas provide visible signs of persistent militarization. After many decades of high homicide rates, in the first quarter of 2020 Colombia registered its lowest rate since 1974. Official data report that the rates of injuries and thefts also decreased, and officials have attributed these changes to the PNC’s focus on prevention by channeling resources to areas with the greatest concentrations of people.1 Reductions in violence cannot be tied completely to improvements in policing, however. Many observers have noted that the crackdown on organized crime and the drug trade in Colombia created market incentives for drug operations to shift to other countries, which have then been plagued with the same problems of violent crime and corruption. However, domestic and international observers charted concrete improvements in policing practices in Colombia, and experts have even encouraged countries such as Mexico and Peru to adopt similar policing reforms (Frühling 2003; Shifter 2012).2 Due to the concentration of political violence in the countryside, reforms have had markedly different effects in urban and rural areas. In small towns and rural areas where guerrillas were active, police typically functioned more as a military force rather than a community force cooperating with the local citizenry. Even in urban areas such as Bogotá, which have been praised for the overhaul of their policing practices, community policing initiatives are paired with an extensive police presence on the streets, and they are heavily armed and dressed in military gear. Despite these early signs of progress, police reputation in Colombia has been marred by violent crackdowns on protesters and abuses of human rights and civil liberties. On September 9, 2020, the police killing of Javier Ordoñez, who had only nonlethal weapons in his possession and pleaded for police to stop, generated massive protests that ended in thirteen more deaths linked to police brutality, as well as hundreds of injuries. Police violence against protesters continued through 2021, and journalists have compared contemporary repression of protesters to Colombia’s prior strikes against insurgents and paramilitaries, reporting that police treat “civilian protesters as battlefield enemies” (Turkewitz and Villamil 2021, A9).

Colombia

67

Human rights organizations in Colombia and around the world have condemned the police abuse of protesters in 2021, meticulously documenting the systematic use of excessive force, arbitrary detention, sexual assault, intentional injury, and extrajudicial killings (Angelo 2021; Amnesty International 2021; HRW 2021). As in Chile (Chapter 3), widespread police abuse and misconduct has led to another critical juncture for police reform, as citizen demands and protests have steadily increased, highlighting the need for new reforms on the national scene and raising questions about the implementation, consolidation, and sustainability of these reforms.

The National Police and a Historical Legacy of Violence The PNC was created in 1891 with the advice of the French National Gendarmerie. Esparza (2022) provides a thorough historical overview of the first attempts to develop a national police force, which failed to gain effective jurisdiction over the country as a whole and resulted in the concentration of operations in Bogotá and the proliferation of other subnational and municipal police forces in the rest of the country. The institutional development of the PNC has been marked by turbulence. During the first half of the twentieth century, the PNC depended on subnational entities for its operations, a situation that lent itself to politicization, culminating in the pivotal role the police played during La Violencia (the violence). La Violencia lasted from 1948 through 1958, and this time of widespread political violence ultimately resulted in the deaths of approximately 200,000 people, as well as massive increases in internal and external migration (Leal Buitrago 1994). In 1948, during the Bogotazo (armed uprising in Bogotá), the members of the Bogotá police joined the fray, which resulted in the virtual destruction of the capital and the liquidation and institutional closure of the police. In 1950, the police were reestablished under the tutelage of the army. This institutional reconfiguration sought to depoliticize the police and block them from participating in politics. However, the police remained dependent on the army, and after a military coup, they joined the army (and some Conservative militias) to fight against Liberals, guerrillas, organized peasant groups, and loosely organized “bandits.” To end this period of political violence and turmoil, the two major parties established the National Front in 1957. Under the National Front, elites divided political power, representation, and political appointments evenly between the two major parties, the Conservatives and the Liberals. The power-sharing agreement paved the way for the return of civilian rule, but it excluded additional actors and their constituents. The central government regained control of urban areas but maintained a weak presence in rural ones. The weakness (or in some areas, complete absence) of control in rural areas

68

Making Police Reform Matter in Latin America

facilitated the proliferation of guerrilla groups, paramilitaries, and eventually organized crime. Under the National Front, the PNC subsumed various provincial and municipal entities into a national force. Despite their military dependence, the civilian nature of the PNC allowed it to advance with significant levels of autonomy and specialization. In the early 1970s, definitions of public order increasingly took on a political character, particularly as the Ministry of Government delivered special instructions to safeguard a particular type of public order (Llorente, Bulla, and Gómez 2016, 209). For example, Decree of Law 1355 in 1970 advanced the administrative and legal modernization of the PNC, but with a particular new competence: “The National Police is responsible for the preservation of internal public order, which is what results from the prevention and elimination of the disturbances of security, health and public morality” (Article 2). The constitution and legislation defined the PNC as a civil, preventive, and educational force; in the 1980s the police began to evolve into a force that prioritized the fight against drug trafficking and guerrillas. The police grew increasingly involved in both the internal armed conflict and government efforts to curtail drug cartel and drug trafficking operations. In the 1980s and early 1990s, the United States and Colombia concentrated their efforts on toppling the major drug cartels, particularly the Medellín and Cali cartels. Under the leadership of notorious drug trafficker Pablo Escobar, the Medellín cartel corrupted the state apparatus to facilitate the ready flow of drugs into the United States. Escobar undermined the police, along with judges and other government officials, through bribery and intimidation. He steadily increased his use of violence against police and other government officials, leading President Virgilio Barco Vargas to reinstate extradition to the United States for drug traffickers such as Escobar, which had been prohibited by the 1991 constitution. To avoid extradition, Escobar negotiated his surrender to authorities on the provision that he serve his sentence under house arrest with opulent amenities. This lasted a little over a year, as the Colombian government reacted with alarm to Escobar’s continued use of violence against state officials, police, and ordinary citizens. Working with the United States, which provided logistical and technological support, the Colombian authorities launched a concentrated campaign to apprehend Escobar. At this point, Escobar had ordered the murder of many police officers and other officials, and police in Medellín joined the pursuit after witnessing the death of many of their own. While police in the past had often succumbed to Escobar’s offers of bribery or threats of violence, by 1993 Escobar had ordered the murder of so many police officers and other officials that a new resolve emerged to apprehend him. Ultimately, the Colombian police killed Escobar in a shoot-out on a Medellín rooftop in 1993, the most prominent example of US and Colombian efforts to apprehend the leaders of drug cartels, a strategy dubbed Operation Kingpin.

Colombia

69

Escobar’s death fragmented the Medellín cartel, but as is often the case, the apprehension of the leader of an organized criminal group did not necessarily lead to the end of violence or diminishing drug trafficking and organized criminal activity. In the void created by Escobar’s death, the Cali cartel grew in power, and by 1995 some estimates indicated that the Cali cartel controlled approximately 80 percent of the world’s cocaine market. Once again, US and Colombian collaboration led to the apprehension of the major actors of the Cali cartel, and by 1995, its influence plummeted. As was the case with the Medellín cartel, catching the cartel leadership did not lead to a decline in drug trafficking, as other actors filled the void in this internationally lucrative enterprise. Still, as Figure 4.1 illustrates, there was a decline in homicide rates that corresponds with the end of the Medellín and Cali cartels. In 1991, homicide rates in Colombia were among the highest in the world, but dropped sharply from 1992 to 1998. Between 1998 and 2004, homicide rates rebounded before beginning a relatively steady decline between 2004 and 2018. Some evidence suggests that decreases in acts of violence in the first decade of the twenty-first century are also linked to greater police and judicial efficiency, which we discuss in the following section (Gómez et al. 2017). Figure 4.1

Homicide Rates in Colombia (1990–2020)

Source: World Bank World Development Indicators.

70

Making Police Reform Matter in Latin America

Reforms and Counterreforms 1990–2020 Just as the Colombian state was focused on dismantling drug cartels and reducing high homicide rates and violent crime, it also embarked on a series of reforms that aimed to strengthen the police and improve its levels of professionalism and transparency. Such efforts launched a process of police reforms and counter-reforms that have spanned three decades. Many of these efforts were attempts to respond to high-profile scandals involving corruption and police brutality. While reforms did increase the professionalism of the police, they have not succeeded in consolidating mechanisms to ensure accountability and transparency. Furthermore, despite concerted efforts to implement community-oriented policing practices, the militarization of police organization, education, and operations has been difficult to dislodge. In the early 1990s, the PNC were seriously undermined by high levels of corruption and daily confrontations with criminal organizations linked to drug trafficking. During the process of drafting a new constitution in 1991, reformers sought to tackle these problems. The drafting and subsequent implementation of the new constitution served as a critical juncture for subsequent police reforms. The presidency of César Gaviria (1990–1994) oversaw the ratification of the new 1991 constitution, which included a range of measures to overhaul the police. For example, under the new constitution a civilian would be appointed to lead the Ministry of Defense—a sharp departure from forty years of military leadership. Chapter 7 of the 1991 constitution is devoted to describing the Fuerza Publica (Public Force), and in eight separate articles (Articles 216–223) the constitution draws clear distinctions between the military forces (army, navy, and air force) and the police. Article 216 makes clear that these are very different organizations, and Articles 217 and 218 clarify their purposes and functions. Article 217 stipulates that the military’s primary purpose is to defend national sovereignty, independence, the integrity of national territory, and the constitutional order. In contrast, Article 218 describes the PNC as a civil force, whose primary purpose is maintaining the necessary conditions for the citizenry to exercise their rights and freedoms and to ensure that the inhabitants of Colombia can live together in peace. The constitution also stressed the importance of an apolitical character for the police and military forces. Article 219 aims to isolate the Public Force from politics and limits the political activities of members of the armed forces, prohibiting them from engaging in political debates, movements, or political parties.3 This new constitution set the groundwork for legislation to further define and regulate the work of the PNC, and just two years later, a scandal accelerated the institutional process of police reform. In 1993, the brutal rape and murder of a young girl inside a police station in Bogotá shocked the nation and laid bare the problem of police brutality. Gaviria’s govern-

Colombia

71

ment had already prioritized police reform, due to the widespread reports of corruption, excessive use of force, and the high degree of drug traffickers’ infiltration. The 1993 scandal gave a decisive push to these ongoing efforts, and even the police leadership spearheaded some reform initiatives. Subsequent police reforms focused on two fronts. First, reforms created the figure of the national police commissioner, with functions of internal control and processing and handling of complaints (Casas Dupuy 2005, 18). Major General Rosso José Serrano was the first to hold this post in 1994, and Congress gave him authorization to purge police officers who had engaged in corrupt activities or had credible abuse or misconduct allegations against them. Approximately seven thousand police officers were dismissed, including among the upper ranks, which lent credibility to anticorruption pledges (Frühling 2003). The reform process also aimed to change the structure and culture of the police, specifically by introducing better management and public administration processes and enlisting public administration experts from the Universidad de los Andes. Second, reforms established the National System of Citizen Participation, a commission to be replicated at the municipal and departmental levels. The National System of Citizen Participation aimed to encourage broad citizen participation, such as taking part in formulating preventive programs; promoting a civil, democratic, and educational ethic; and channeling complaints and claims (Casas Dupuy 2005, 16). Both pillars of the reform were developed from a process of wide participation of experts and civil society; however, police acceptance of these reforms was quite limited at the beginning of their implementation. Police discontent with President Gaviria ran high, particularly since he insisted that to prosecute crime successfully (especially crimes involving drug trafficking and organized crime), the PNC needed to undergo an extensive internal purge and implement reeducation and retraining programs. The police objected to the mass expulsions, grumbling that while they were carried out through legal means, they were discretionary and rapid. Due to this pushback, political will for reforms dampened a bit, leading President Ernesto Samper (1994–1998) to announce the rollback of external oversight measures in the mid-1990s, arguing, “Let’s let the police regulate itself.” Samper did try to address the militarization of the police with the Cultural Transformation Program, which the PNC implemented between 1995 and 1998. However, under Plan Colombia (2000–2015), the militarization of the police intensified. President Andrés Pastrana (1998–2002) had originally envisioned Plan Colombia as a Marshall Plan for rural Colombia; however, the version signed into US law in 2000 focused heavily on militarization to defeat drug traffickers and the FARC. This militarization of security intensified under President Alvaro Uribe (2002–2010), who proposed a new way of understanding security in

72

Making Police Reform Matter in Latin America

Colombia called “democratic security,” which emphasized the militarized fight against insurgency, organized crime, and extortion. Still, Uribe was pressed to enact some police reforms, particularly to respond to scandals involving the appropriation of more than two tons of cocaine that had originally been seized from drug traffickers. The misappropriation of the cocaine, together with the nontransparent handling of the problem by the police high command, triggered additional reforms in 2003. Uribe’s government created the Special Mission for the Police to propose reforms; however, this mission did not convene civil society or political leaders. Rather, it brought together four experts who prepared a document of recommendations, and only those that generated the least amount of police resistance were implemented. The result was more of an improvement plan than an institutional transformation of the police. Despite these setbacks, reforms improved the professionalization of the police forces and the management and operations of the institution as a whole. National and municipal efforts aimed to strengthen police ties with local communities. In 2010, Colombia adopted the Plan Nacional de Vigilancia Comunitaria por Cuadrantes (National Plan for Quadrant Community Surveillance) in its eight major cities. Based on the Chilean quadrant model (see Chapter 3), this plan divided cities up into smaller geographical areas or quadrants, assigned six police officers to each quadrant, and trained officers on community-oriented patrolling techniques (Mejía, García, and Ortega 2013). The plan notes that all officers must be visible, approachable, and recognizable to residents and businesses in their quadrant, which facilitates accountability (Leon 2019). Like its Chilean counterpart, Colombia’s Quadrant Plan aims to employ problem-oriented policing strategies, whereby police use data-driven methods to identify high-crime areas, diagnose the causes of the criminal activity, and tailor responses to address the acts and their root causes holistically. Geographic information systems facilitate the defining of geographic quadrants and the allocation of personnel and resources (Leon 2019). Engagement with the community is the cornerstone of Quadrant Plans, as the goal is not simply to reduce crime but to improve police–community relations. To enable citizen contact with the police, the plan has included the rollout of smartphone apps, online messaging, and neighborhood callboxes (Leon 2019). On the heels of the Quadrant Plan, the police service doctrine underwent modification in 2013, which profoundly altered the provision of police service (Bulla and Guarín 2015). Throughout this period of reform, the number of functions entrusted to the police increased steadily. This is perhaps best exemplified by the new National Police and Coexistence Code in 2017, which had not been updated since 1970. Lawmakers argued that it was imperative to revise the code to address contemporary challenges and promote “peaceful coexistence.” Police duties expanded to cover issues related to conduct on public trans-

Colombia

73

portation, pet ownership, cell phone theft, appropriate use of public spaces, and consumption of psychoactive substances and alcohol, among other issues (Castillo and Ayala 2019). This amplification of functions has generated some criticism, as it involves police in some behaviors that could in practice be contrary to peaceful coexistence. The corrective measures prescribed in the new code emphasize sanctions and could lead the code to become repressive when mixed with the logic of criminal law (Granda and Manolo 2018). For example, the 2017 code stipulates that police must not be disrespected and levies fines on any person who attacks, disrespects, or challenges a police officer. Citizens face fines for the failure to carry their identification cards and clean up after their pets in public. Citizens can download an app to familiarize themselves with the new police code and the fines for a wide range of infractions.4 Then President Juan Manuel Santos championed the new code, arguing, “We had a code that had not been adjusted for more than 45 years . . . the new code is an important step to ensure a more peaceful, harmonious, and peaceful coexistence in our daily lives” (Wade 2017). The modernized code does contain provisions to reflect societal changes, such as the protection of LGBTQ rights. Still, some have expressed concern that the language of the new code is too broad and gives police too much latitude in areas such as entering homes without warrants. After two decades of police reforms, it is undeniable that the PNC transformed from an ineffective, corrupt, and illegitimate organization to a more professional and effective force, which began to earn local and regional recognition and legitimacy. Despite this progress, problems of corruption, accountability, and respect for citizens’ rights lingered. In 2016, the “community of the ring” scandal highlighted persistent problems of institutional corruption, as reports uncovered the presence of a network of police corruption and prostitution embroiling police officers and lawmakers. According to revelations, some police paid legislators with sexual favors to promote themselves in their institution from 2006 to 2010. This scandal led to the resignation of the director of the Police and the vice minister of the Interior and led to a new institutional reform commission. To combat corruption, the commission proposed improvements to the educational formation of police and to the institution’s management, budget, and operations. It advocated improvements to the quadrant model and an increase in the number of women police officers. International Actors and Police Reform International actors have played a large role in the police reform process. Many targeted their aid and loans to address Colombia’s role in the drug trade and the power of organized criminal groups, which often led to prioritizing short-term counternarcotics objectives and the militarization of civilian

74

Making Police Reform Matter in Latin America

police forces (Moncada 2009, 435). Most notably, this has been a key criticism of Plan Colombia, which targeted police reforms as part of its aid package to combat drug trafficking and left-wing guerrilla forces. The United States originally budgeted $115.6 million to the PNC for helicopters, agricultural spray, aircraft, communications equipment, ammunition, spare parts, training, and logistical support (US Department of State 2001). A much smaller amount, $25 million, was budgeted to establish human rights units in the PNC and the attorney general’s office. While Plan Colombia was conceptualized as a multidimensional strategy to tackle problems as diverse as the peace process, justice reform, and economic stabilization, most of the aid was military in nature. The wartime context prioritized a militarized response to the ongoing conflict, and many politicians tended to assuage citizens’ anxiety over public security with militarized responses. With his vision of “democratic security,” President Uribe intensified the militarization of the state’s response to drug trafficking and insurgency and consequently the militarization of the police as well. Other actors strived to have a more sustained and holistic approach. The Inter-American Development Bank (IDB) provided loans to support the police reform process in Bogotá and other Colombian cities and advocated an integrated approach to address “the social and situational cause of crime and violence through activities that typically spanned different areas of public policy” (IDB 2013, 8). These different areas of public policy included education, social protection, and gender diversity and resulted in IDB loans distributed across different government ministries. The IDB invested US$33.7 million into Colombia’s program Support for Peaceful Coexistence and Citizen Security, which aimed to reduce crime and violence with investments in social prevention, situational prevention, information systems, production and use of data on violence, alternative dispute reconciliation/conciliation, houses of justice, and police strengthening (IDB 2013, 27). With this approach, the IDB intended to provide long-term funding to citizen security projects, as well as loans to “other sectors that have specific components to prevent violence (IDB 2013, 7). Decentralization and Subnational Variations Reforms in the 1990s decentralized power, making cities important actors when insecurity was the most dominant issue in the country. While the Colombian police force remained national, cities were given the option to design public security policies tailored to meet their unique circumstances, and police forces reported to the national Ministry of Defense as well as mayors. At the local level, important changes took place in the police, and several mayors became “stars” of crime prevention in Latin America (Hoelscher and Nussio 2016).

Colombia

75

Colombia’s three largest cities, Bogotá, Medellín, and Cali, stand out for innovations in policing and public security. In the 1990s, these cities were infamous for high rates of violent crime such as homicides, coupled with high levels of police corruption, misconduct, and abuse of power. By 2020, despite persistent, lingering problems with police transparency, accountability, and conduct, public security in these cities had improved dramatically. Moncada (2016) outlines how cities became pioneers in confronting violent crime, particularly as patterns of the global drug trade and organized crime caused violence to surge in the 1990s. It is important to note that the security situations in these cities in the 1990s created a critical juncture and may have facilitated state and societal acceptance of sweeping reforms. Given how critical the security situation was, politicians and the public may have been willing to accept any reforms that held promise to remedy the situation. In Bogotá, police reform began in earnest in 1995 and continued for two decades across the administration of three different mayors. At the subnational level, we see that sustained political will across municipal administrations enabled successful reforms. Some mayors hailed from very different places on the political spectrum, demonstrating that there was a great deal of political will for sustaining the reform process. Financial investments underscored this commitment, as the budget for policing in Bogotá increased dramatically across these administrations. The police budget roughly quadrupled between 1992 and 1997, and increased an additional 125 percent between 1997 and 2003 (Bello Montes 2004). Moncada (2009, 431) provides a detailed account of how these mayors pursued a strategy of lateral reform, a strategic sequence of policies that altered the societal context in which institutional reforms unfolded over time. Institutional reforms were matched with investments in social and economic policies, increasing the public’s ability to participate and engage in the process and ultimately leading to greater levels of public support for institutional reform and trust in the police as a whole. Institutional reforms prioritized police accountability and transparency. Up to the late 1990s, the Bogotá police were infamous for their abuse and misconduct and the inability to hold police officers accountable. In 1996, Mayor Antanas Mockus overhauled the police code of conduct, rendering them more accountable to state officials. The public played a vital role in this effort, as reformers organized a series of meetings between police and local communities to solicit feedback on police conduct and areas for improvement. Moncada (2009, 440) provides a detailed overview of this process, noting that four hundred community meetings “brought together over 18,000 citizens to discuss how the police could be more accountable to city residents,” and attendees included “prostitutes, human rights organizations, and other sectors of society that had been historically marginalized

76

Making Police Reform Matter in Latin America

in discussions of citizen security.” These marginalized populations had historically been estranged from (and sometimes brutalized by) the police. These forums created opportunities for citizens to have a voice in police reforms, and this broad participatory process helped shield reform efforts from resistance and downturns in political will. Societal mobilization kept pressure on mayors and other political actors to sustain the reform process. Particularly in the aftermath of police brutality scandals in 1993, societal pressure from below gave public officials incentives to continue to strengthen the state’s ability to hold police accountable through mechanisms such as investigative oversight committees and judicial proceedings (Moncada 2009). Societal mobilization was even more effective when the police systematized the collection of crime statistics and disseminated this information to the public. In 1995, Mayor Mockus inaugurated the Sistema Unificado de Información de Violencia y Delincuencia (Unified System on Violence and Crime), which collected comparable data on criminal activity and regularly shared these data with the press and public. Armed with information, citizens were able to press public officials to maintain investments in police reforms, particularly when crime rates started to rise (Moncada 2009). Institutional reforms in Bogotá were matched with extensive community outreach. A 1996 external evaluation of policing in Bogotá stressed the need to strengthen police presence at the local level and improve forms of organization, association, and communication with local communities (Bello Montes 2004). To address these weaknesses, Bogotá initiated “proximity policing,” which emphasized the need to integrate police work more cohesively into local communities under the oversight of local officials. This emphasis on oversight was critical, as it highlighted the importance of increasing police accountability and transparency, which had been sorely lacking. Indeed, it would be counterproductive to increase the presence of police in local communities without improving accountability mechanisms, as a larger police presence without accountability would be a recipe for more police misconduct and human rights abuses. Community outreach was institutionalized through the creation of Frentes de Seguridad Local (Local Security Fronts) and Escuelas de Seguridad Ciudadana (Schools for Citizen Security) in 1995. The Frentes de Seguridad Local were community organizations headed by the PNC, which aimed to address local security problems with prevention, primarily by increasing citizens’ abilities to engage with their legal authorities. The goal of these organizations was to promote collaboration between civic leaders and police, so that civic leaders could transmit vital knowledge on local criminal activity. Such information helps police target problem areas with lights, alarms, and motion-activated cameras activated, for example (Bello Montes 2004). By 2004, there were more than nine thousand Frentes

Colombia

77

de Seguridad Local, and approximately half of the residents in Bogotá lived in areas served by community police (Bello Montes 2004, 68). Later, the Frentes de Seguridad Local enabled other community-oriented policing programs, such as the attempted transplantation of Chile’s Quadrant Plan (discussed in Chapter 3). As already mentioned, in 2010, the PNC launched the Plan Nacional de Vigilancia Comunitaria por Cuadrantes in Bogotá and seven other cities, dividing the cities into smaller geographical areas and assigning specially trained police to patrol and establish strong relationships with local communities. Over nine thousand police officers were trained to partake in this community-oriented program, and the implementation included clear mechanisms for accountability (Mejía, García, and Ortega 2013). Police were trained to approach crime holistically and consider not just the prevalence of criminal acts but the main concerns specific to each community, such as the quality of social services and local infrastructure. The Escuelas de Seguridad Ciudadana consisted of weekend classes at public universities to educate citizens on the functions and responsibilities of the police, as well as how they could work to increase public security (Moncada 2009). In the first decade of their implementation, over 37,000 citizens attended these workshops (Bello Montes 2004). These workshops aimed to redefine traditional means of providing security and emphasized that successful police work needed community collaboration. The workshops approached security from a holistic and preventive perspective, addressing not just crime but related issues, such as conflict resolution, alcoholism, and drug addiction. Reformers also implemented a series of community workshops to educate citizens on conflict resolution and alternative dispute resolution and arranged for mental health specialists to work with citizens who had been victimized by violent crime. Other measures were more practical and targeted. Relying on data that linked alcohol to crime and transit fatalities, in 1995 Mayor Mockus banned alcohol sales between 1 and 6 a.m. (Moncada 2009, 438). Reforms modernized the automatic dispatch centers to reduce the police response time for resident calls. Centros de Atención Inmediata (Immediate Attention Centers) were reoriented to decentralize the work of police stations and become more accessible to local communities, particularly those in busier parts of the city or areas with higher crime rates (Policia Nacional de Colombia 2009). Officials also worked with private actors, such as the Bogotá Chamber of Commerce, to collect data on public awareness and perceptions of community policing in Bogotá. Police reform in Bogotá occurred against a backdrop of extraordinarily high rates of violent crimes, such as homicides. In 1993, homicide rates in Bogotá were among the highest in the world, reaching a historical peak of 81.2 per 100,000. By 2004, homicide rates had declined by approximately

78

Making Police Reform Matter in Latin America

72 percent, to a low of 22.6 per 100,000. Police attributed much of this decline to the reform process, although as mentioned, this decline could reflect the fact that criminal activity associated with the drug trade shifted to other locations. Even so, the correlation between police reforms and better public security outcomes succeeded in increasing public trust in police in Bogotá (Bello Montes 2004; Moncada 2009). Even with these positive developments, police reform in Bogotá also highlights the limits of reform. Even with investments in social and economic policies, police reforms were still embedded in a larger context of inequality and its subsequent power relations, and police accountability and treatment of citizens often varies by socioeconomic status (Moncada 2009). Furthermore, when faced with large-scale protests, police quickly reverted to prior patterns of repression and abuse of power. This extensive and holistic reform process did not culminate in police respect for demonstrators, as evidenced by police attacks on protesters in 2019–2021 (Turkewitz and Villamil 2021). Furthermore, despite these efforts to improve police–community relations, the Bogotá police have retained much of their militarized structure and presence. Colombia’s second largest city, Medellín, also witnessed sharp reductions in violent crime and homicides in the early 2000s. Like Bogotá, Medellín recorded some of the highest homicide rates in the world during the 1980s and early 1990s, and its name was synonymous with organized criminal groups and high rates of violent crime. Pablo Escobar centered his drug trafficking operations in Medellín, leading homicide rates to reach a historic high of 381 per 100,000 people by 1991. When faced with extradition to the United States, Escobar unleashed a murderous rampage, assassinating political candidates, detonating car bombs, and offering a $2,000 bounty for the murder of any police officer (Maclean 2015). Due to historical patterns of inequality, poverty, and social exclusion, Escobar readily found recruits for his enterprise in Medellín’s poorest neighborhoods. Even after Escobar was killed in 1993, Medellín continued to face problems of violent crime and insecurity, as fragmented criminal organizations battled for the remnants of his criminal empire and the Colombian civil war increasingly became urbanized (Moncada 2016). As in other parts of Colombia, in Medellín the lines between legitimate state actors and other forces has been unclear; state forces, for example, often rely on paramilitaries in their fights against guerrilla forces (Maclean 2015). Homicide rates did steadily drop from the historic highs of the early 1990s; however, by 2000 the homicide rate was still high by international standards, at 160 per 100,000 (Dolan 2018). To address these high rates of violent crime, Medellín relied less on changes to policing practices and more on socioeconomic investment and political change. The national government had sponsored projects to target

Colombia

79

crime and violent criminal activity in Medellín in the 1990s, but these efforts stalled in the face of local political opposition (Moncada 2016).5 In 2004, newly elected Mayor Sergio Farjado reenergized these efforts under a policy of social urbanism. Farjado’s social urbanism prioritized investing in poor and marginalized areas and working with business elites and civil society leaders to develop new public spaces, schools, libraries, high-tech educational facilities, and parks. Improvements to public transportation, including the metro system and cable cars, transformed how Medellín’s residents traveled around the city. Investments in public transportation integrated many of the peripheral parts of the city with the center and dramatically cut commuting times, particularly for Medellín’s poorest citizens. Political change accompanied these investments and encouraged participation from Medellín’s historically marginalized citizens. The inauguration of participatory planning and budgeting programs encouraged civic engagement and gave communities a greater voice in public spending and safety projects. To combat corruption, Fajardo increased transparency and made city budgets, contracts, and spending data publicly available (Moncada 2016, 240). These political changes, coupled with sustained investments in Medellín’s most vulnerable communities, drastically reduced violent crime rates. Homicide rates dropped sharply between 2002 and 2007, from 184 to 33.8 per 100,000 people. By 2015, Medellín’s homicide rate reached a historical low of 20.2 per 100,000, a rate on par with US cities like Indianapolis and Atlanta (Held and Robbins 2019). Medellín has received a series of international awards attesting to the success of its transformation, such as the 2013 World’s Most Innovative City Award (awarded by the nonprofit Urban Land Institute). As violent crime rates abated, tourism, foreign investment, and civic association participation increased. Rather than being the murder capital of the world, Medellín was now famous for its “Medellín miracle,” and observers contemplated the viability of exporting this model to other cities struggling to address violent crime. Medellín’s transformation deserves its applause, but it is important to note that these transformations are also attributable to the actions of some unsavory characters. Socioeconomic investments and political inclusion were critical parts of the Medellín Miracle, but these occurred against a backdrop of important changes in the criminal landscape. The 1990s were marked by the rise and fall of the Medellín cartel and subsequent fighting over the spoils of Escobar’s empire. Bandas, milícias, combos, and death squads battled for territorial control as Colombia’s civil war increasingly entered urban areas (Moncada 2016, 232). By the early 2000s, Adolfo Paz, better known as Don Berna, consolidated control over the Oficina de Envigado criminal organization and the Bloque Cacique Nutibara paramilitary group (Moncada 2016). Even after the paramilitary group was ostensibly

80

Making Police Reform Matter in Latin America

disbanded, Don Berna maintained tight territorial control over his illicit enterprises, and with this monopoly, he was able to control levels of violence. Local officials seeking to work in historically marginalized communities typically had to coordinate with existing criminal organizations that controlled these areas. While public security has improved since the 1990s, policing in Medellín can still entail extreme force. Before Farjado’s social urbanism, the militarization of public security culminated in the highly controversial Operation Orion, launched by President Uribe in 2002. The military targeted the Comuna 13 neighborhood of Medellín with extreme force, and a four-day siege led to the death, injury, and displacement of hundreds of civilians. Once again, the line between legal and illegal actors was blurred, as the military and police collaborated with paramilitaries to rout left-wing guerrilla groups. Critics have argued that the military succeeded only in replacing the guerrillas with paramilitaries (Yagoub 2017b). The InterAmerican Court of Human Rights condemned the militarized operation in a densely populated urban setting, as well as the arbitrary detention of residents (Corte Interamericana de Derechos Humanos 2017). More recently, positive images of Medellín’s urban renewal are juxtaposed with photos and videos of police forcefully evicting residents of informal housing communities during the global pandemic (El Tiempo 2020). Particularly for poorer residents, the face of the police is the Escuadrón Móvil Antidisturbios (ESMAD; Mobile Anti-Disturbance Squad) not an officer trained in community-oriented policing. Police reforms in Colombia’s third largest city, Cali, highlight the problem of sustainability. Like Bogotá and Medellín, Cali reported extraordinarily high rates of violent crime and homicide in the 1990s from the presence of drug-trafficking and organized criminal groups. International donors and domestic reforms focused on Cali’s public security crisis; for example, in 1992 Cali became the first Colombian city to implement reforms to improve police accountability. Rates of violent crime and homicide improved in Cali; homicide rates fell from a high of 124 in 1993 to 47.8 in 2018 (Guerrero-Velasco et al. 2021). However, many reforms have not been sustained. Moncada (2010) provides a thorough overview of police reform in Cali, noting that international donors prioritized technologically driven reforms, such as crime mapping. Because Cali is home to the largest urban population of Afro-Colombians, crime mapping originally aimed to determine when socioeconomic and political inequalities “mapped onto longstanding racial fault lines” (Moncada 2010, 697). Unfortunately, crime mapping became detached from the broader goals of addressing socioeconomic, racial, and political exclusion and devolved into an exercise of “counting bodies” (Moncada 2010, 697). By simply counting criminal

Colombia

81

activities without following up with targeted investments in reducing socioeconomic, political, and racial exclusion, Cali’s crime mapping served mainly to stigmatize Afro-Colombian neighborhoods and deploy an additional police presence without addressing the root causes of criminal activity. Even as the city sought to transform its public security policies, police officers disproportionately targeted Afro-Colombian men with harassment and even disappearance and disparaged and impeded outreach efforts to Afro-Colombian youths. In contrast to the private sector in Medellín, which spearheaded efforts to invest in initiatives to address the root causes of crime, private sector elites in Cali have tended to favor investments in “building the coercive capacity of the police, increased coordination between the city and the national military and combating kidnappings of business owners” (Moncada 2010, 711). As in other Colombian cities, Cali police have been charged with human rights violations for their attacks against protesters. In May 2021, the UN High Commissioner for Human Rights, Michelle Bachelet, denounced the violent response to strikes and social mobilization in Cali, which included reports of extrajudicial killings at the hands of off-duty police, as well as attacks against police. Reports of police abuse of force, including extrajudicial killings with live ammunition, were widespread across Colombia in 2021; according to Human Rights Watch, the toll was highest in Cali (HRW 2021).

Contemporary Challenges Police reform in Colombia faces two main contemporary challenges. The first is related to the implementation of the 2016 peace accords. The Final Agreement for the Termination of the Conflict and the Construction of a Stable and Lasting Peace highlights the need to review the role of the public force, especially in terms of incorporation criteria, education, and personnel administration. Still, the postconflict process takes place in a context of “grays” (Llorente, Bulla, and Gómez 2016) and implementation has been quite uneven (Turkewitz and Villamil 2021). Since 2016, security has rapidly deteriorated in the rural countryside, as criminal groups fight for control of areas vacated by the demobilized FARC. The United Nations documented eleven massacres in 2017, and this number rose to thirty-three in 2020. To halt this backslide, the government must safeguard rural areas traditionally abandoned by the police service (Llorente, Bulla, and Gómez 2016). Furthermore, there is an urgent need to review the leadership capacities of the institutional command to face situations related to limited efficiency (Mendez 2016), corruption (Väsquez 2013), or violence against special groups of the population (Moncada 2010).

82

Making Police Reform Matter in Latin America

The second challenge relates to police violence against protesters. Domestic and international headlines have raised alarms over police abuse of human rights during demonstrations in 2019, 2020, and 2021. However, these latest manifestations of police violence are part of a much larger trajectory. According to the Directorate of Criminal Investigation, between 2005 and 2015, 9,048 national police officers were reported for acts of violence (Bonilla Ovallos 2016, 115). Although several reforms have sought to demilitarize the police, the organizational structure, training, and role that police assumed in the face of the armed conflict set them on the opposite path (Casas Dupuy 2005). In 2020, the Supreme Court mandated several police reforms to curb police abuse of protesters, such as allowing human rights groups to verify detentions and ordering the Ombudsperson’s Office to monitor ESMAD abuses against protesters (HRW 2021, 26). However, President Iván Duque’s government (2018–2022) has ignored some of these mandates and only superficially enacted others. After the wave of police violence against protesters in 2021, multiple protests and demands have clamored for a true, deep, and necessary police reform. Duque was not especially proactive at this critical juncture, however, preferring hard-line security policies and superficial reforms. Duque relied heavily on the antiriot squadron ESMAD to repress protesters, favored inviting the military to patrol the streets, and compared protesters and vandals to terrorists (Angelo 2021). In some cases, prosecutors have supported these efforts, filing “disproportionate chargers of ‘terrorism’ against some demonstrators who allegedly engaged in vandalism” (HRW 2021, 21). To date, the militarization and lack of oversight of the PNC have not been adequately addressed.

Perceptions of Police How have public perceptions of police fluctuated amid these broad institutional changes and some of their reversals? Although public opinion data is not yet available to examine public trust in police after the widespread abuses of 2020 and 2021, we can use LAPOP data to examine public attitudes toward police during the latter period of reform, from 2004 to 2018.6 As Figure 4.2 illustrates, from 2004 through 2018, average trust in police was significantly higher than average trust in other political institutions. However, average trust in police did decline steadily over this time, from a high of 4.6 in 2005 to a low of 3.7 in 2016. As 2016 was the year of major human rights and corruption scandals that led to the mass dismissal of officers, this drop in trust is not surprising. Average trust in police rebounded slightly in 2018 to 4.0, the midpoint of the scale and just slightly above the regional average. In 2016 and 2019, the survey included an additional question to measure how much the police respected human rights of Colombians. When asked specif-

Colombia

83

ically about the police’s respect for human rights, Colombians ranked police significantly lower (by approximately half a point each year).

Crime Trends and Public Perceptions of Crime As Figure 4.1 illustrated, rates of homicides began to drop sharply in 2004 and reached historical lows in 2014–2020. Other indicators of violent crime also improved. The number of massacre victims dropped from 680 in 2002 to 147 in 2009, and kidnappings declined from over 2,800 in 2002 to 213 in 2009 (Rosen 2021). Although police reforms garnered positive attention because they coincided with a drop in homicides, as we noted earlier, it is not clear that reforms were the only factor in reducing crime rates. Furthermore, critics charge that there is a lack of investment in human and material resources dedicated to criminal investigations, as antidrug efforts tend to claim the lion’s share of investments due to their higher profile (Frühling 2003). While indicators of the most violent crimes decreased over time, rates of self-reported victimization rose steadily. In 2010, according Figure 4.2

Trust in Police in Colombia (2004–2018)

Source: LAPOP’s 2004–2019 National Surveys.

84

Making Police Reform Matter in Latin America

to LAPOP’s 2004–2019 National Surveys, the percentage of respondents reporting victimization in the past year increased from 16.9 percent to 20.5 percent, and it increased to 23.8 percent in 2018. Murder rates declined, but other types of less violent victimization began to increase; self-reported victimization rates were 65 percent higher in 2018 compared with 2004. Additional surveys on victimization and perceptions of safety explore victimization trends in more detail. According to a 2018 survey conducted by the Departamento Administrativo Nacional de Estadística (DANE) on coexistence and citizen security, victimization was highest in the capital city, as 26.5 percent of Bogotá respondents reported that they had been victimized by crime, slightly more than the 25.3 percent recorded in 2016. Of those who reported victimization, crimes involving the theft of a vehicle or vehicle parts (8.7 percent) and theft from people (7.3 percent) were most common. The Bogotá Chamber of Commerce also conducts surveys on crime and insecurity and in 2019 identified theft as the most common crime in the city. In the DANE survey, 28.7 percent of victims reported the incident to the police, whereas in the Chamber of Commerce survey this rate was around 44 percent. The Chamber of Commerce survey also reported that only 16 percent of respondents indicated they had gone to the police in the previous six months. Still, Colombians who interacted with the police rated police service higher than those who reported no contact with the institution. Ratings of police service have declined more recently. In 2008, 35 percent of the sample rated police service as good; by 2019, this percentage had declined to 18 percent. Among those who had contact with the police, the percentage declined from 48 percent to 29 percent in the same period. Not surprisingly, higher rates of self-reported victimization corresponded to increases in public fear of crime. Indeed, fear that crime was increasing was a major factor in Duque’s 2018 presidential victory, along with public discontent over negotiations with the FARC (Rosen 2021). As Figure 4.3 illustrates, average fear of victimization increased steadily from 2012 through 2018. The increase was sharper for rural residents compared with their urban counterparts. By 2018, fear of crime in rural areas was on par with urban ones, and both were higher than the regional average. The 2018 DANE survey also found insecurity to be high, as 44.6 percent feel insecurity in their city, mainly because of the presence of common crime. In addition, 84 percent of those surveyed expressed fear of crime. In the 2019 Chamber of Commerce survey, 58 percent of those surveyed in the city felt fearful, and 41 percent felt fearful in the neighborhoods where they live. In 2004, LAPOP included a series of questions related to perceptions of security, police performance, and President Uribe’s Citizen Security Plan. After two years in office, Uribe’s public security policy earned positive evaluations: 61 percent of LAPOP respondents indicated that this pol-

Colombia Figure 4.3

85

Fear of Crime in Colombia (2004–2019)

Source: LAPOP’s 2004–2019 National Surveys.

icy had led to a decrease in human rights violations, compared with only 24.7 percent who stated that human rights violations had increased under this plan, and 14.4 percent who stated that some types of human rights violations had increased while others had decreased. Furthermore, that year 48.6 percent of respondents indicated that access to justice in the comiserias (precincts) was either good or very good. When asked about the levels of satisfaction with various entities, the police ranked statistically the same as the alcaldías, fiscalías, and tribunals (mayors, prosecutors, and courts, respectively) (average score 2.7 on a scale of 1–4). In 2004, a variety of indicators seemed to point to at least a fair amount of public satisfaction with police and public security policy. One of the primary objectives of the reforms initiated in the 1990s was to improve police professionalism. Although we do not have data on rates of bribery from the early 1990s, before such reforms were initiated, we do have data from 2004, a decade after the reform process began. By that time, observers were touting the successes of the reforms, particularly by improving personnel benefits and recruiting and vetting officers. In 2004, the percentage

86

Making Police Reform Matter in Latin America

of respondents who indicated that the police had solicited a bribe was 3.8 percent, one of the lowest rates of the region (Seligson 2004).7 As Figure 4.4 indicates, this percentage increased slightly between 2004 and 2008 but rose sharply in 2009 to 8.8 percent. Despite some fluctuations, the overall trend in bribe solicitation was one of increase, reaching a peak of 12.9 percent in 2016. Although Colombian respondents reported some of the lowest levels of police bribery in the region in 2004, by 2016–2017, bribe solicitation was above the regional average, and Colombia ranked on par with Honduras (Cohen, Lupu, and Zechmeister 2017). Despite a decline in 2018, police bribe solicitation remained high at 9.5 percent—two and a half times higher than reported in 2004. By many accounts, police corruption decreased from its highs in the early 1990s, but during the period covered by LAPOP’s survey data (2004–2018), instances of petty corruption increased. Other indicators of police performance started to worsen in the 2010s. The percentage of respondents who thought that the police served to protect the community declined steadily from 2008 through 2013, from 56.5 percent to 45.8 percent. By 2013, over a third of respondents indicated that they thought that police were involved in crime. Likewise, when asked whether they thought that the justice system would punish the guilty party if they were victimized by robbery or assault, respondents registered some trust that the justice system would hold suspected criminals accountable, but not a great deal. Average responses on this item were largely stable between 2004 and 2011 but began to decline slightly. In 2018, only 32.4 percent of respondents indicated that they trusted the justice system to sanction criminals “somewhat” or “a lot.” In 2014, LAPOP asked respondents to indicate their level of satisfaction with police in their own neighborhoods. Overall, most respondents were not very satisfied with police performance in their immediate neighborhoods. In Bogotá, only 41.1 percent were either “satisfied” or “very satisfied” with police performance, whereas in other urban areas and rural ones, half of residents indicated at least some degree of satisfaction. When asked about police responsiveness, perceptions of response time have remained consistent over time. From 2014 to 2018, on average respondents thought that police would respond to a home burglary between thirty and sixty minutes. However, bivariate analyses detected significant differences in perceived response times according to level of urbanization and respondents’ race. According to Pearson’s correlation coefficients, residents of rural areas reported perceived response times significantly longer than those of urban residents (r = −0.209, p < 0.01), although this is somewhat understandable given the longer distances of rural areas. Respondents who identified as Afro-Colombian, Indigenous, or of mixed race thought the police would take significantly longer to arrive than respondents who identified as white or mestizo (r = −0.061, p < 0.01).

Colombia Figure 4.4

87

Police Bribery in Colombia (2004–2019)

Source: LAPOP’s 2004–2019 National Surveys.

In addition to charting longitudinal trends in public attitudes toward police, we aim to examine the degree to which these levels of trust vary according to key cleavages in Colombia. The urban/rural divide has featured most prominently, and race has been a dividing line as well, particularly in cities with larger Afro-Colombian populations. Given the police’s disproportionate response to protesters beginning in 2019, we examine whether there are earlier indicators that protesters were more wary of police. To determine the effect of these factors on trust in police, we conducted a regression with key demographic and socioeconomic variables; perceptions and experiences with crime; participation in protests; and political ideology.8 As police reforms and public security policy has been subject to important subnational variations, we include dummy variables for Bogotá, Medellín, and Cali. Table 4.1 reports the results for these ordinary least squares regressions.9 Our analysis finds that perceptions of crime have a significant effect on trust in police, as those who fear victimization in their neighborhoods register significantly lower levels of trust in police on average, holding all other independent variables constant. Personal victimization is not significant in

88

Making Police Reform Matter in Latin America

Table 4.1; however, this is mainly due to problems of multicollinearity, as victimization is significantly correlated with fear of crime (r = 0.127, p < 0.01). When victimization is included in the model without the fear of crime variable, it has a significant impact on trust in police (see Model 2). Our model included measures of ideology and personal experience with police bribery. Both variables are significant and positive: respondents who identify with the ideological right report more trust in police, and those who indicate the police solicited a bribe from them in the past year reported significantly less trust in police.

Table 4.1

Trust in Police in Colombia (2018) Model 1

Constant Participation in protest Ideology Police solicitation of bribe Crime victimization Fear of crime Gender   (1 = women) Income Race   (1 = non-White) Education Age Bogotá Medellín Cali Adjusted R2 N

4.137** (0.269) −0.302* (0.148) 0.076*** (0.018) −1.092*** (0.159) −0.146 (0.110) −0.223*** (0.049) 0.095 (0.095) −0.222 (0.254) −0.157 (0.117) −0.002 (0.014) 0.069*** (0.016) −0.521*** (0.128) −0.280 (0.203) −0.602** (0.211) 0.117 1,453

Model 2 (Fear of Crime Omitted) 3.511*** (0.228) −0.281^ (0.149) 0.081*** (0.018) −1.125*** (0.160) −0.217* (0.109) — — 0.064 (0.095) −0.212 (0.255) −0.158 (0.117) 0.000 (0.014) 0.071*** (0.017) −0.595*** (0.127) −0.259 (0.203) −0.653** (0.212) 0.106 1,463

Notes: Coefficients are unstandardized, with standard errors in parentheses. p < 0.06; *p < 0.05; **p < 0.01; ***p < 0.001.

Colombia

89

Age was the only demographic variable that was significant, as older respondents had more trust in the police. We find differences among the city dummy variables. Residents of Bogotá and Cali were significantly less trusting of police than residents of smaller cities and rural areas, whereas residents of Medellín did not have significantly different levels of trust in police. People who reported that they had participated in a protest in the past year had significantly lower levels of trust in police, holding all other independent variables constant. Thus, even before the major crackdowns on protesters began in 2019, people who engaged in protest were already leery of the police. The significance of the protest participation variable does decrease to p < 0.06 in Model 2, which aims to address the problem of multicollinearity. All in all, we find that trust in police hinges heavily on perceptions of security, but protesters were also less likely to trust the police even before the well-publicized incidents of excessive use of force from 2019 to 2021.

Conclusion The Colombian case offers important insights for those seeking to improve police performance. First, the reforms occurred against a backdrop of extraordinarily high rates of violent crime and organized criminal activity. This is critically important, as some observers have questioned whether community-oriented reforms can take place in such a context. Indeed, this has been the rationale that some have used to resist more sweeping police reforms in high-crime countries such as El Salvador. While high rates of violent crime had the potential to hinder reform efforts in Colombia, in many ways this context enabled the consideration of sweeping reforms, as it created a critical juncture in which national and local actors were willing to try a range of measures. In the aftermath of these reforms, public safety undoubtedly improved, as crime in Colombia today is on a completely different scale and magnitude than it was during the 1980s and 1990s. Still, it is not clear whether improvements in public security can be linked completely to reform efforts, as other factors can shape security outcomes (e.g., shift in criminal activity to other countries, coordination among criminal actors to minimize violence). The Colombian case also highlights the importance of cross-cutting investments in social services and preventive measures to address the root causes of crime, although this occurred primarily at the subnational level. Particularly at the local level, reforms were more likely to succeed in reducing violent crime rates and fostering public security when they were holistic, addressed socioeconomic deprivation, and created opportunities

90

Making Police Reform Matter in Latin America

for political participation for historically marginalized groups. In Medellín, such investments even eclipsed police reforms and are largely credited with dramatically improving public security. Police reform in Bogotá highlights the importance of sustained political will, as the process spanned several different municipal administrations. While these mayors hailed from different parts of the political spectrum, they agreed that police reform was critical to improve public safety and sustained political consensus to carry out the reforms, even when inherited from prior administrations. Several have argued that nowhere in Latin America has the process to democratize the police been as extensive and complex as in Colombia (Moncada 2010, 703). However, the Colombian case shows the limitations to reform efforts and highlights problems of sustainability. While the professionalization of the national police improved, the PNC failed to replace the repressive, militarized, centralized, and politicized police model that was its mark for much of the twentieth century (Bonilla 2016, 118). Even when community-oriented reforms were implemented, these often coexisted with more militarized policing practices and structures. As people continued to take to the streets in 2021 to protest police violence and demand accountability, it became clear that demilitarizing the police is the critical next step. If the police continue to pertain to the Ministry of Defense, and accountability occurs only in military courts, it is highly unlikely that the police will transform into a force that truly respects the rights of the citizens they are supposed to serve.

Notes 1. Press release from the Office of the President, “Colombia registró en el primer semestre del 2020 la tasa de homicidios más baja de los últimos 46 años: Policía Nacional,” Bogotá 2020, /id.presidencia.gov.co/Paginas/prensa/2020/Colombia -registro-en-primer-semestre-del-2020-tasa-de-homicidios-mas-baja-de-los-ultimos -46-anios-Policia-Nacional-200701.aspx. 2. Such advocacy is often part of a broader package of reforms and international aid, modeled after Plan Colombia. There have been numerous calls for a similar plan for Mexico and Central America; however, critics caution that the successes of Plan Colombia are often overstated (Rosen 2014), and that the ability to replicate it in other countries is questionable (Bailey 2011). 3. The full text of the 1991 Colombian constitution, and its subsequent reforms through 2009, can be found at Georgetown University’s Political Database of the Americas, https://pdba.georgetown.edu/. 4. The app is called Código de Policía de Colombia, and it is available on the Google Play store. 5. For an in-depth discussion of 1990s efforts to address violent crime and organized criminal activity in Medellín, see Moncada (2016) and Maclean (2015).

Colombia

91

6. Average trust in other institutions relied on a scale that included trust in the legislature, political parties, local mayors, and courts. 7. In 2004, police bribery in Colombia was on par with that of Costa Rica and Nicaragua. In contrast, bribery in Mexico was almost five times higher than that reported by respondents in Colombia. 8. The variables included in this analysis are slightly different from those of the Chilean regression analysis (Chapter 3) and reflect the different factors important in the Colombian context. The control variables included the traditional survey research socioeconomic and demographic variables: age, gender, education, income, and race. Age was measured in cohorts: (1) 18–25; (2) 26–30; (3) 31–35; (4) 36–40; (5) 41–45; (6) 46–50; (7) 51–55; (8) 56–60; (9) 61–65; (10) 65 and over. Education was measured as the number of years of formal education completed, and race was measured dichotomously (0 = White or mestizo and 1 = non-White). The income scale was calculated based on answers to the following survey items: Do you or any member of your household have any of the following possessions? TV; car; refrigerator; telephone; cell phone; computer; microwave oven; washing machine; drinking water; sewage system. Responses were coded as (1) yes and (0) no. We created an index of personal income using a means formula that included a case if there were valid responses to at least eight of the ten items. Crime victims were coded as (0) not a victim; (1) victim of crime in past twelve months. The question regarding protest participation asked respondents to recall the past twelve months and indicate dichotomously if they had participated in protest or not. We have recoded these responses, so that respondents are coded (1) if they had participated in a protest and (0) if they had not. 9. We report the ordinary least squares regression results for ease of interpretation. We replicated these results using ordinal logistic regression, and the variables had the same signs and levels of significance. Results of ordinal logistic regression are available on request.

5 Costa Rica: Exceptionalism Under Strain

COSTA RICA HAS LONG BEEN REGARDED AS THE EXCEPTION IN Central America, and its exceptionalism is typically cited as the reason the country has not struggled with crime to the same extent as other Central American countries. Costa Rica differs from its Central American neighbors on a variety of fronts: public health, sanitation, democratic governance, and educational achievement. It stands to reason, many argue, that it would also do better in crime prevention and maintenance. Costa Rica has had the resources and political will to develop justice institutions that are less corrupt, more professional, and more integrated into civil society. In this context, it is not surprising that its crime control policies tend to stress the importance of balancing punitive and preventive tactics and developing strong ties with citizens. Costa Rica’s reputation as a peaceful oasis has come under strain, however. In the mid-2000s, murder rates reached historical highs, public fear of crime peaked, and trust in justice institutions dropped. Costa Rican officials lamented that their justice system was sobre cargado (overburdened). Public fear of crime became linked to anti-immigrant sentiment and more punitive security policy preferences. To date, it is not clear whether these recent developments will undermine the Costa Rican success story. Do these trends reveal previously overlooked flaws? Or are they merely setbacks in a longer history of success? Is Costa Rica undergoing a fundamental shift? This chapter addresses these questions by tracing the historical development of policing practices and public security policy in Costa Rica, as well as public perceptions of crime and justice institutions.

93

94

Making Police Reform Matter in Latin America

Historical Overview According to idealized versions of history, Costa Rican exceptionalism is tied to its early colonial roots. The Spanish conquered the land making up contemporary Costa Rica relatively late (at the end of the sixteenth century) and found few natural resources or Indigenous labor to exploit. The lack of resources and geographical isolation relegated Costa Rica to a colonial backwater, and the population had more racial and cultural homogeneity than its Central American neighbors. The homogeneity of the population facilitated the emergence of a more egalitarian political culture, as residents tended to see less social distance between themselves and other citizens. Most of the population was concentrated in the Central Valley and engaged in agriculture on small, family-owned farms. This idealized version of history glosses over the socioeconomic inequalities prevalent in Costa Rican society; however, these inequities were on a much smaller scale than in other parts of the region.1 Likewise, Costa Rica was no stranger to military rule in the nineteenth century, but repression was not on the same level as in other countries. Costa Rica’s status as a former colonial backwater insulated it from foreign military intervention, particularly the US occupation that was common in other parts of the region in the late 1800s and early 1900s.2 As Booth points out, “Costa Rica does not escape Central America’s worst social problems, but it experiences them less severely” (1998, 27). These differences in degree have been substantial enough to make civilian governance and the rule of law viable alternatives to the military rule and “rapacious elite culture” of its northern neighbors (Booth 1998, 19). It is easier to rely on the law to govern and settle disputes when the distances between socioeconomic actors is smaller, and the economy does not depend on exploiting a poor underclass. Even though inequality was lower in Costa Rica than other Central American countries, it did feature in economic, social, and political power structures. These idealized versions of Costa Rican history overlook the role that immigration has played in socioeconomic and political development. While the Central Valley tended to be more homogeneous, Costa Rican economic development relied heavily on immigrants to remedy labor shortages, particularly in agriculture and construction. Banana plantations and construction industries depended heavily on Nicaraguan laborers and Black immigrants from the Caribbean, particularly Jamaica.3 Multinational corporations also recruited Chinese immigrants to construct railroads. Police often played a critical role isolating these immigrants and their descendants from the more homogeneous Central Valley. Legislators enacted a series of laws to control the movements of “less desirable” immigrants. In 1862, the Ley de Bases y Colonización (Law of Bases and Settlements) prohibited the “colonization” of African and Chinese races unless they provided indispensable labor. If economic necessities required the

Costa Rica

95

immigration of these groups, the government had the authority to control their movements (Alvarenga 2011, 5). On the heels of this law followed additional measures designed to encourage the “right” immigrants (White Europeans) and discourage “undesirable” (non-White) groups. In 1897, legislation prohibited new Chinese immigrants from entering the country; in 1914, this prohibition extended to Arabs, Turks, Armenians, and Roma (Seligson 1980).4 Police were charged with upholding these laws, placing them in the role of border guards between classes. Seligson (1980) notes that there were informal means for keeping nonWhites out of the Central Valley. Before 1949, many Costa Ricans referred to the country’s central mountain range as a racial frontier, or a natural divide between racial groups. Anecdotally there are reports that non-White inhabitants needed passes to travel over the mountain range from the Atlantic Coast to the Central Valley. Seligson reports that there is no surviving documentation of laws requiring such passes, but some documents suggest that authorities implemented an informal system of segregation. For example, Central Valley residents petitioned officials to patrol the racial frontier more effectively and keep non-Whites from crossing the mountain range. Limited transportation routes made it easier for elites to control the movement of non-Whites and block them from migrating to the Central Valley. In the early 1930s, contracts to open banana cultivation in the Pacific zone included clauses requiring preferential treatment for Costa Rican workers and prohibiting non-Whites from working there.5 In 1930, legislators further limited the movement of immigrants with the Registro de Identificación Inmigratoria (Immigration Identification Registry), which put immigrants under police control. The law required foreigners to register with the local police “to account for their activities, the amount of time they expected to stay in the country, and their means of subsistence” and compelled owners of lodging establishments to report foreigners to the police.6 This police control did not apply to all foreigners equally, as the police were authorized to exempt some foreigners from these proceedings if they deemed them “deserving,” due to their “acknowledged honorability” (Alvarenga 2011, 13). Police featured prominently in regulating the movements of immigrants and non-Whites until 1949. Before the 1949 constitution was put in place, police were part of the Policía Militar de Orden y Seguridad (Military Police of Order and Security). After the brief but extraordinarily violent civil war of 1948, Costa Rica fundamentally broke with the status quo with a series of radical reforms in a new constitution. This six-week war was brief, but with approximately four thousand casualties, it was the bloodiest political event in Costa Rican history and was ultimately a catalyst for radical reforms (Palmer and Molina 2004a, 139). The civil war marked a critical juncture in Costa Rican political development, with important ramifications for the institutional development of the police and policing practices. After

96

Making Police Reform Matter in Latin America

the war, Costa Rica famously abolished the standing army and increased investment in social welfare programs, proclaiming a preference for an army of teachers. The 1949 constitution also eliminated race-based citizenship and provided a clear framework for naturalization.7 Thus, police no longer played a role in monitoring and controlling the movements of nonWhite Costa Ricans and immigrants (at least formally). The 1949 constitution was famous for eliminating the military, abolishing race-based citizenship, and establishing a social welfare state that prioritized nearly universal education and healthcare. Although one should be cautious about romanticizing historical narratives painting a picture of an egalitarian nation of small yeoman farmers, it does appear that comparatively lower levels of inequality in Costa Rica made it easier to abolish the military and create the democratic social welfare state.8 Because the economy did not depend on the repression of labor, it was more feasible to abolish the military. Lower levels of social differentiation facilitated the establishment of universal health and education programs. These lower levels of inequality and investments in human capital have implications for policing. Police repression was not necessary for maintaining the economic status quo, and investments in education and healthcare promoted a minimum standard of living that is typically associated with lower crime levels. Against this backdrop, police do not typically act as border guards between classes, nor does the government rely on police to repress dissent and opposition. Rather, police tend to serve the public good more broadly. As noted in Chapter 1, investments in social welfare are conducive to more inclusive policing practices and more holistic responses to preventing and addressing crime. The 1949 constitution set the stage for a period of peace and prosperity in Costa Rica, and the police have benefited from this political and social stability. While other Central American military and police forces have been tainted by their associations with repressive governments and human rights abuses, in Costa Rica the police have escaped such infamy. Even so, abolishing the military created challenges for police forces. The police became the only state entity responsible for security, including domestic law enforcement and external security. Particularly in more recent times, when organized crime and illicit trade have increased, this expansive role for police became more difficult to fulfill.

Building Professional Police Forces: The Reforms of the 1990s In 1949, the police and military forces were converted into the Guardia Civil (Civil Guard) and entrusted with police functions and the maintenance of domestic order and security. To foster the professionalization of the police

Costa Rica

97

force, the National Academy of the Police was created in 1964. From the late 1960s through the 1980s, policies restructured the police and created new divisions. For example, as conflicts over land tenure escalated during this time, the state created the Guardia de Asistencia Rural (Rural Guard) to contain and confront protests and maintain order in the countryside. In the 1990s, Costa Rica began to implement a series of reforms to modernize policing practices and improve professionalism. Even though the police were highly regarded, particularly for their lower levels of corruption and higher levels of respect for human rights, critics argued that police posts depended too much on political patronage, and there were concerns about the lingering use of military titles and forms of organization (Moya 2012). The 1990s reforms were also part of a regional and international movement, as the end of the Cold War and region-wide transitions to democracy sparked new interest in promoting the rule of law and human rights. Due to past investments in a professional civilian police force, Costa Rican policymakers were able to use this critical juncture to deepen early processes of police reform. While there were international influences, Costa Rican elites led the process of police reform as they aimed to winnow out clientelist practices in the public security sector, create better working conditions for police officers, and improve the administration, organization, transparency, and professionalization of police forces (Eijkman 2007; Solís Moreira 2014). To modernize the police force for contemporary times, the Costa Rican legislature passed the General Law of the Police (no. 7410) in 1994, followed closely by a new Criminal Procedure Code in 1995 and Criminal Code in 1998 (Eijkman 2007). The General Law of the Police aimed to reduce clientelism in the police force and enhance the police’s apolitical nature, through measures such as barring the removal of personnel when new governments took power (RESDAL 2013, 20). This 1994 legislation established the Fuerza Pública (Public Force), the primary national police force in Costa Rica today. The General Law of the Police described the position of police officer as a professional career, codifying the legal requirements for advancement and salaries. This reform emphasized the need to provide enhanced training and opportunities for specialization. In addition to modernizing the police force, the General Law focused on preventing crimes and improving relationships between police and the community. According to Article 4, “the public force of the police will be at the service of the community, in charge of vigilance, maintaining public order, preventing crime, and cooperating to control crime according to legal orders.” In 2001, the Law of the Civil Police reinforced this legislation and aimed to endow the police with a more civilian character, as critics had charged that prior police modes of operation were too militaristic. The Law of the Civil Police also created the department of Police Legal Assistance, which supports the police in their duties by providing legal counsel (RESDAL 2013).

98

Making Police Reform Matter in Latin America

The Costa Rican police system had been criticized for its excessive number of divisions, particularly because coordination among them is sometimes difficult. The police have traditionally been divided into several smaller forces, each assuming duties in a specific sphere. To facilitate coordination among these diverse police divisions, the 1994 police reforms created the National Council of Public Security, led by the president. Coordination continued to be cumbersome, however, leading to a series of additional reforms. For example, in 1995, the Ministry of Interior and Police was combined with the Ministry of Security, to form the Ministry of the Interior, Police, and Public Security. In 2004, the Modernization of the State Act overhauled police divisions again, and the last revision to this structure was enacted in 2013. The Public Force is the most recognizable of the Costa Rican police units and is responsible for maintaining peace and public order. Their other charges include: • • • •

Ensuring citizens are free to exercise their rights Protecting the constitutional order and citizens’ safety Guarding the security, integrity, and property of the people Preventing and repressing legal infractions

This description of duties emphasizes the police’s role in upholding constitutional and citizen rights and balances preventive and repressive measures. The Public Force includes specialized divisions, such as Tourist Security and Community Prevention, Investigation, and Development. The Office of Preventive Police Programs operates under the auspices of the Public Force and promotes community outreach and prevention. Following regional trends, Costa Rica established the Dirección de Programas Policiales Preventivos (Direction of Preventive Police Programs), which, among other duties, manages the Departamento de Prevención Comunitaria (Department of Community Policing). Similar to Chile and Uruguay (see Chapters 3 and 9) the community policing program began with a pilot in 1996 and was fully implemented in 1998. Costa Rica also inaugurated proximity policing programs in 2001. Over time, the Public Force has become more active in public projects. In 2002, it collaborated with civil society groups to promote domestic violence awareness, reduce the consumption of illegal drugs, and protect the rights of minors. In 2008, Costa Rica rolled out additional community-oriented reforms, much like those unfolding in Chile (Chapter 3), Uruguay (Chapter 9), and Colombia (Chapter 4). Indeed, there have been some policing exchanges between Costa Rica and Colombia, highlighting the region-wide adoption of community-oriented strategies. The Public Force is supplemented by a series of specialized police divisions, such as the Coast Guard, the Border Police, and the Anti-Narcotics

Costa Rica

99

Police. This latter agency has become increasingly important given the growth of organized criminal groups using Costa Rica as a transit station for illicit trade and money laundering. Given US interests in curbing the international drug trade, the US government has provided aid to improve the equipment and training for the police to combat drug trafficking (RESDAL 2013). In 2016, for example, the United States pledged $30 million in aid to support efforts to confront organized crime, including cargo planes, armored vehicles, and infrastructure for maritime surveillance. Some aid will also bolster preventive measures, such as scholarships for at-risk youth (Bargent and Leimann 2016). In addition to the national Public Force and specialized agencies, new security actors have emerged on the scene. There are eighty-one municipalities in Costa Rica, and due to sharp increases in crime in the 2010s, several began to deputize municipal officers and finance them through municipal budgets. According to Solís Moreira (2014), gaps in the 1994 Law of Police opened the door to the creation of municipal forces; however, there are concerns over how well municipalities can regulate such forces, and what type of education and training such police would have (65). In 2011, there were approximately 600 municipal policies officers, compared with 13,720 in the national Public Force. Although there have been attempts to regulate and legislate municipal investments in policing, legal concrete rules for decentralizing the police are lacking so far (Solís Moreira 2014, 68). Private security investments and actors have also proliferated, and private security forces are a critical part of the security apparatus. By 2013, there were 25,960 private security officers and 1,129 private security agencies registered in the country, compared with fewer than 2,000 private security officers in 1994 (Solís Moreira 2014, 67). In 2003, legislation and executive decrees sought to increase the amount of regulation and oversight of private security agencies and officers. The Ministry of Public Security oversees the operations of private security actors through a special agency under the auspices of the vice minister of Special Units. Despite enhanced oversight, concerns remain about the extent to which the private security sector is held accountable to the state. Overall police reforms in Costa Rica have unfolded without much controversy, driven by a technocratic approach to policymaking that emphasizes balancing prevention with repression. There has generally been elite consensus on the need for incremental police reforms and sustained political will to design, implement, and reevaluate them. Reforms have developed against a backdrop of investment in social welfare, creating supports for more holistic policing and public security policies. Still, there have been several areas of concern. First, some critics have pointed out that there is no formal code of police, and that many police officers have not completed critical basic training courses, such as the Curso Básico Policial (Basic

100

Making Police Reform Matter in Latin America

Police Course). By some estimates, only 65 percent of police patrolling the streets have successfully completed this course, even though it has been offered by the School of National Police since 1995 (Moya 2012). Similar concerns have emerged over the training of specialized forces, as investments in training for specific skills has occasionally lagged behind the creation of specialized units. Second, domestic and international observers have raised alarms about the capacity of the Costa Rican police to confront the contemporary challenges of organized criminal groups, illicit trade, and money laundering. The General Law of Police took effect when Costa Rica was a very different country, and crime was of a completely different magnitude. Even as laws and executive decrees have sought to ensure that police are equipped to adapt to these changes, and public resources are increasingly devoted to improving domestic security, Costa Rican officials have raised alarms that the police and justice system is sobre cargado. Just as reforms began to modernize and improve police performance in the 1990s, crime trends in Costa Rica began to change dramatically. The next section provides an overview of trends and recent public policy responses to this dramatic shift in the status quo.

Crime Trends and Public Fear of Crime Crime has historically been low in Costa Rica, but this trend started to change in the 1990s. The 1990s and early 2000s were a time of economic, social, and political transformation. Given the rocky nature of these changes, some observers began to caution about possible “trouble in paradise” (Seligson 2002). The economy underwent dramatic change, as neoliberal reforms restructured the economy and liberalized markets. Among other things, these economic changes resulted in higher unemployment and inequality. Between 1990 and 2001, the Gini coefficient increased from 45.3 to 51.1 before decreasing to 48.6 in 2012.9 Unemployment rates rose from 5.6 percent in 1991 to 7.8 percent in 2012.10 Not surprisingly, these changes had political consequences. Analyses of survey data from 1978 to 2012 have documented steady declines in the legitimacy of political institutions in the eyes of the Costa Rican public (Seligson 2002; Alfaro-Redondo and Seligson 2012). Lower levels of political legitimacy coincided with lower levels of voter turnout, particularly for the two major political parties. In the early 2000s, voters increasingly gravitated toward minor parties, a trend that culminated with the new Citizens’ Action Party (founded in 2002) winning the 2014 and 2018 presidential elections. As Figure 5.1 indicates, during this period, crime rates began to rise. Costa Ricans had long prided themselves on the low rates of violence in their country, particularly in contrast to the violence that has long consumed

Costa Rica

101

their neighbors to the north. By the end of the twentieth century, patterns of crime and violence began to change. Homicide rates, which had always been low in Costa Rica, doubled from 4.6 (per 100,000 people) in 1990 to 9.6 in 2014.11 Homicide rates reached historic highs in 2017, climbing to 12.1 per 100,000 people. During this time frame, Costa Rica became an increasingly important country in the global illegal drug trade as both a transit country and a growing domestic market (Parkinson 2013). Experts estimate that over $4 billion in illegal profits are laundered in Costa Rica each year (Riesenfeld 2015). Costa Rican homicide rates were still far lower than Panama or the Northern Triangle countries; however, citizens reacted with alarm to the abrupt change in the status quo. Eight districts consistently reported a homicide rate greater than 20 per 100,000 inhabitants: Limón, San José, Garabito, Orotina, Tibás, Pococí, Matina, and Corredores (Estado de la Nación en Desarrollo Humano Sostenible 2019). Costa Rica registers low rates of violent crime when compared with other Central American countries, but these rates are at historic highs within the country.

Figure 5.1

Homicide Rates in Costa Rica (1995–2019)

Source: World Bank, World Development Indicators. Note: Reliable homicide data is available for Costa Rica in the 1990s, but comparable data from several other Central and South American countries was not valid and reliable until the 2000s, when data collection and reporting improved.

102

Making Police Reform Matter in Latin America

Victimization from other types of crimes has also increased. As LAPOP’s national surveys show, self-reported victimization rates increased gradually from 2002 through 2010. Then, after dipping in 2012 and 2014, the rates climbed to historical highs, reaching 22.2 percent in 2018. Between 2002 (the first year LAPOP posed this question in Costa Rica) and 2018, selfreported victimization rates increased by 66 percent. The most common crimes tend to be robbery and theft, and Costa Rica reports the highest rates of these types of victimization in Central America. However, these higher rates could reflect the underreporting of such crimes to police in the other Central American countries (RESDAL 2013). Rates of reporting crime tend to correlate with public trust in police, as victims need a minimum level of trust to be willing to turn to police to address victimization (Malone 2014). Even so, focusing on the survey data from Costa Rica, between 1990 and 2010 rates of robberies per 100,000 inhabitants increased by 750 percent, and rates of theft per 100,000 inhabitants increased almost 200 percent. This is a sharp and dramatic increase in a twenty-year time frame, one that has caused citizens to react with alarm. Not surprisingly, the public has reacted to such increases in criminality with alarm. As shown in Figure 5.2, fear of crime decreased between 2006 and 2010 but increased significantly from 2010 to 2014 to match the regional average. Given that Costa Rica had long enjoyed a reputation as a peaceful and stable oasis, it is noteworthy that citizens expressed levels of fear of crime on par with those of the region, as the regional average includes countries with dramatically worse records in violence and public safety. While perceptions of insecurity decreased slightly in 2016, by 2018 fear of crime had returned to its highest levels. As discussed earlier, 2017 saw the highest levels of per capita homicides in the country, a total of 603 cases, equivalent to a rate of 12.2 homicides for every 100,000 inhabitants. By 2019, murders totaled 560, equivalent to a rate of 11.2 per 100,000 residents. As crime rates and public fear of crime rose, Costa Rica sharply increased its investments in public security. By 2013, the budget for the Ministry of Public Security increased five times over its funding in 2003 (Solís Moreira 2014). Part of this budgetary growth funded an increase in police personnel, from 237.7 officers per 100,000 inhabitants in 2000 to 287.5 in 2011 (Solís Moreira 2014, 66). As public fear of crime increased, some of the blame for criminality was cast at the feet of immigrants. In addition to economic changes and rising crime rates, in the late 1990s and 2000s, immigration also increased, particularly from Nicaragua. The arrival of these new immigrants has correlated with higher levels of fear of crime (Solano 2017). Since the early 2000s, tensions between Costa Ricans and immigrants have increased as the size of the Nicaraguan migrant community has grown. In 1998, Costa Rica declared an amnesty to legalize undocumented migrants from Central

Costa Rica Figure 5.2

103

Fear of Crime in Costa Rica (2004–2018)

Source: LAPOP’s 2004–2018/19 AmericasBarometer.

American countries. Since this agreement, large numbers of Nicaraguans have migrated for seasonal and long-term employment opportunities. Political repression in Nicaragua after April 2018 has further fueled migration trends (see Chapter 6). Nicaraguan migrants are overwhelmingly urban: 70 percent live in cities, particularly in and around the capital of San José.12 Unfortunately, Nicaraguans living in Costa Rica have become scapegoats for increases in crime. Sandoval García (2004) documents the negative depictions of Nicaraguan immigrants in the Costa Rican media, noting the propensity to portray them as violent criminals taking advantage of Costa Rica’s generous social services, and this scapegoating is often used to justify ill treatment and discrimination. Sandoval García (2004, 2015) notes that xenophobia increasingly peppers public discourse, even among public intellectuals long associated with social democracy and leftist politics.13 In a comprehensive survey of attitudes toward immigrants, Delgado Montaldo (2008) documents how Costa Ricans perceive these arrivals. He reports that a substantial majority of Costa Ricans (over 70 percent) indicate the country should not accept any more immigrants, and over 90 percent say the government should deport undocumented immigrants. About one in four Costa Ricans (26.1 percent) believe that immigrants are very integrated into the communities where they live, 59.2 percent believe they are somewhat

104

Making Police Reform Matter in Latin America

integrated, and 14.7 percent say foreigners are not at all integrated into their communities. Despite these negative responses, 90 percent of Costa Ricans are in favor of immigrants getting access to health services, 95 percent are in favor of immigrants having access to education, and more than 87 percent believe they should have access to the judicial system. Approximately twothirds of Costa Ricans are in favor of immigrants having access to school scholarships, 65 percent would extend access to housing programs, and 68 percent express support for immigrants getting access to micro-enterprise credits. Costa Ricans may favor limiting future migration; however, they tend to support immigrants already living in Costa Rica with adequate social services. When asked about civic rights, 81 percent would extend the right to organize to immigrants, and 77 percent would support extending citizenship rights (Delgado Montaldo 2008). Analysis of survey data also points to a relationship between public fear of crime and attitudes toward migration. In 2014, LAPOP asked respondents, “How much do you agree or disagree that Costa Rican culture is threatened by people that come here from other countries?” Respondents who agreed that people from other countries threatened Costa Rican culture registered significantly higher levels of fear of crime (r = −0.107, p < 0.01). Thus, when we examine trends in fear of crime, it is important to remember that fear of crime does not solely reflect a preoccupation with victimization and violent crime, it also reflects other public attitudes and insecurities (Dammert and Malone 2006).

Public Security Policy Public security debates reflect these broader changes in Costa Rican society. In many ways, the sharp increases in crime in the early twenty-first century marked another critical juncture, as police institutions and public security policies needed to adapt to address changing security conditions. Crime emerged as a highly salient topic in public discourse, and there have been increased calls to employ more punitive policies to curb crime, particularly among those exhibiting xenophobic attitudes (Malone 2019). Sandoval García (2014) reports that xenophobia shapes how police officers interact with suspected criminals. To date, at the national level, public security policy has not become more punitive. Xenophobia has shaped immigration legislation,14 but has not yet fundamentally altered national public security policy. For example, the controversial 2006 General Law of Migration and Alien Affairs framed migration as a matter of state security and expanded police powers to include measures such as indefinite detention (Fouratt 2014, 161). In 2009 the legislature revised this statute, however, and softened some of its more controversial provisions (Sandoval García 2015).15

Costa Rica

105

President Laura Chinchilla (2010–2014), a longtime public security expert, was able to hold off calls for most punitive measures, reserving harsher sentences for those convicted of large-scale operations in drug trafficking. As President Chinchilla stated, “Insecurity affects all of us today; it is a complex problem which we must seriously consider. Insecurity has deep roots and requires action in two areas: social security and public safety.”16 Throughout her term, as well as her earlier tenure as minister of Public Security, Chinchilla emphasized the need for a holistic response to the problem of crime in Costa Rica. In office during a time of historically high crime rates, she repeatedly emphasized the need to avoid a knee-jerk reliance on punitive policies. “Punish with a strong hand, prevent with an intelligent hand,” was a slogan designed to take the middle road of public security policy, assuring citizens that criminals would be held accountable, while recognizing the need to address the root causes of crime. Chinchilla’s voice was a moderate one, in great contrast to other Central American leaders calling for mano dura (and even super mano dura) approaches to fight crime. In 2014, Luis Guillermo Solís Rivera (2014–2018) assumed the presidency as a member of the center-left Citizens’ Action Party (PAC), marking the first time the new party succeeded in winning the presidency. Solís’s presidency focused heavily on reducing corruption, but he also highlighted the dangers posed by organized criminal groups to Costa Rican security and transparency (Solís Moreira 2015). As an academic, Solís had authored books about the challenges of organized crime, particularly for democratic governance, focusing on the problem of youth recruitment in gangs and the role of civil society in the fight against organized criminal groups (Solís and Aravena 2008). Prior to his election, Solís had advocated holistic public security policies to tackle the growing power of organized crime. For example, he highlighted the problem posed by the marginalization of urban youth, arguing that schools should be placed at the center of prevention efforts. He advocated investing in primary education to improve retention rates and educational outcomes, with the goal of insulating more youth from potential recruitment from criminal groups (Solís and Aravena 2008, 140). Solís also stressed the role of employment opportunities to incorporate juvenile offenders back into society and flagged the problem of juvenile access to light arms. In line with his anticorruption efforts, Solís approved purging the police ranks of officers with credible corruption charges, leading to the dismissal of 115 officers under Article 140 of the constitution.17 However, this move was potentially undercut by his administration’s changes to police training. To respond to increasing rates of violent crime, Public Security Minister Gustavo Mata announced that the police academy would shorten the selection process and training by six months and deploy more officers to the streets quickly (Clavel 2017). Concerns abound about the efficacy of this

106

Making Police Reform Matter in Latin America

approach, as less time in the academy could lead to less training on critical areas such as human rights, legal procedures, and transparency. At the end of his term, Solís recognized that public security continued to be a problem, as Costa Rica reported a historical high of 603 homicides in 2017. As he stated, “During the last years there has been a significant increase in cocaine trafficking from South America . . . This situation has produced a small-scale drug dealing phenomenon in Costa Rica, evidenced by the disproportionate increase of violent crime between the local drug cartels” (Tico Times 2018). Despite this backdrop, crime and insecurity took a back seat in the 2018 elections for president, as voters focused more attention on economic matters (Costa Rican debt), corruption (linked to construction kickbacks), and LGBTQ rights. Indeed, when asked to identify the most pressing problem facing the country, most Costa Ricans pinpointed economic concerns, and only 6 percent cited security concerns. This is a stark contrast from 2012, when 42 percent of respondents highlighted security concerns, or even 2016 when 22 percent identified security as the principal challenge. According to LAPOP’s survey data, “security”-related problems (e.g., crime, violence, drug trafficking) increased as the most important problem facing the country between 2004 and 2010 from 29 percent to 48 percent of respondents. From 2012 to 2018, security declined from 42 percent to a low of 6 percent. This decline could be attributed to the fact that the economic situation deteriorated after 2012, so economic matters increasingly shared the spotlight with security ones. Carlos Alvarado Quesada, also from the PAC, won the presidency in 2018. In contrast to his primary opponent, Fabricio Alvarado Muñoz, Alvarado did not focus heavily on crime and public security in his campaign. Crime and citizen security did not feature prominently in his party’s platform, although he did stress the importance of strengthening the judiciary to improve public security. The PAC platform defined security as a public good, and there has been an emphasis on improving women’s security through measures such as the 2020 Law Against Street Sexual Harassment. Still, the Alvarado administration prioritized economic issues and climate change. Like his predecessors, Alvarado acknowledged the problem of organized crime, particularly drug trafficking and weapon proliferation (PAC 2017, 54). Observers cautioned that Alvarado did not have a coherent plan for tackling organized crime, despite the evidence produced by Costa Rican agencies linking rising rates of insecurity to the presence of organized criminal groups and the trafficking of illicit trade (Asmann 2018).18 Alvarado proposed curbing the power of organized criminal groups by enacting stricter gun control measures to restrict access to firearms, although again, these measures have been met with skepticism due to concerns of feasibility and efficacy. Asset forfeiture has also been identified as a way to target organized crime; however, the Costa Rican Institute on Drugs has often

Costa Rica

107

struggled to maintain and sell seized assets (Pechinski 2021). Alvarado stressed the need for enhancing police infrastructure and training, although the latter proposition was met with skepticism given the recent changes to shorten training under his predecessor. Alvarado maintained a focus on preventive measures to strengthen citizen security, although his proposals lacked specific details. In 2019, he stressed the need to improve security in border areas, especially in the Huetar Norte region (where the border area with Nicaragua is located) to regulate the flow of foreign migrants into the country (Arguedas et al. 2020).

Public Perceptions of Police We examine public perceptions of police against this backdrop of reform and rising crime. Due to the longevity of its democratic regime, survey data from Costa Rica date back to the 1970s. LAPOP conducted its first national Costa Rican survey in 1976 and administered additional national and metropolitan surveys through the 1980s and 1990s. In contrast to today, where questions concerning crime and insecurity dominate political discourse and consequently national surveys, in the 1970s and 1980s such questions appeared infrequently in surveys. Indeed, most questions pertaining to insecurity related to concerns that violence in other Central American countries could spill over and affect Costa Rica. Questions pertaining to policing focused narrowly on police interactions with protesters, particularly after the debt crisis of 1983 sparked a series of sustained protests across the country. Overall perceptions of security were quite positive throughout the 1970s and 1980s. When a 1978 survey of Central Valley residents asked how well the government guaranteed the protection and security of citizens, more than 75 percent of respondents responded positively, and 65 percent thought the incumbent government did a good job of fighting crime. In 1983, LAPOP conducted a survey of the metropolitan area of San José, and once again, respondents were quite positive about the state of public security. Seventy percent had positive evaluations of the current government’s track record of fighting crime. Only 14.2 percent indicated that they were dissatisfied with the security and vigilance where they lived. A survey of the Central Valley in 1985 found a slightly less rosy picture on security matters, but even so, just over 50 percent of respondents gave President Luis Alberto Monge’s government positive evaluations for fighting crime, and only 19 percent were unsatisfied with the security and vigilance where they lived. By 1995, crime rates were still quite low, and homicide rates were low for the hemisphere and lower than the global average. In a survey of the Central Valley in that year, however, respondents registered slightly more pessimistic views of public security. Only 34.7 percent evaluated President

108

Making Police Reform Matter in Latin America

Rafael Calderón’s government positively for fighting crime, and 45.1 percent indicated that the incumbent government provided protection and security for citizens. When asked specifically about the police, only 47 percent stated that the police respected the lives of people like themselves. However, people who reported having personal experience with the police (n = 287) indicated that their experiences were positive by a large margin; only 22.6 percent indicated that their personal experiences with police were bad or very bad. LAPOP first asked a question about levels of trust in police in a national survey in 1999, five years after the General Law of Police sought to modernize the police. In 1999, the police ranked below other domestic political institutions on average, and slightly below the midpoint of the scale (X = 3.6). These low evaluations could be linked to perceptions of corruption, as 56.8 percent of respondents indicated that the police were either somewhat or very corrupt and the correlation between this measure of perceptions of corruption and trust in police was quite strong (r = −0.497, p < 0.01). By 2004, average levels of trust in police had rebounded, and police ranked above other domestic political institutions. In 2006, however, trust in police declined significantly in real terms and when compared to other institutions (see Figure 5.3). This drop parallels a sharp rise in rates of violent crime. By 2018, however, trust in police had returned to earlier levels, even as trust in other domestic institutions declined. The drop in trust in police in 2006 also mirrors reports of bribe solicitation. As Figure 5.4 indicates, the percentage of people who reported that the police had asked for a bribe in 2004 was low for the region, at 3.5 percent. This began to change in 2006, as the those reporting personal experiences with police corruption rose substantially to 8.7 percent and remained high in 2008 at 8.4 percent. In 2010 these levels started to decline, and by 2018 self-reported bribe solicitation returned to its 2004 levels. To examine the impact of crime and police misconduct on trust in police, Table 5.1 reports the results of a regression analysis on the 2018 data.19 As these results indicate, personal experiences with victimization and public fear of crime significantly decrease trust in the police. Bribe solicitation also has a significant negative effect on trust in police. The public appears poised to penalize the police for the deteriorating public security environment, as well as police misconduct. In terms of the socioeconomic and demographic variables, older and rural respondents were significantly more trusting of police, but levels of education, wealth, and sex were not significant. In addition to this diffuse trust at the national level, we examine evaluations of police in respondents’ communities. When asked whether police protect people in the community from crime or engage in criminal activities themselves, responses were very stable between 2008 and 2012. Slightly less than half of respondents indicated that police protected communities

Costa Rica Figure 5.3

109

Trust in Police and Other Institutions in Costa Rica (1999–2018)

Source: LAPOP’s 1999–2018 National Surveys.

from crime in both years, in contrast to the roughly one third who thought police were involved in crime. Although one third may seem like a large percentage to view the police as enmeshed in crime, it is one of the lower rates in the region, on par with Uruguay and Colombia (Yagoub 2017a). Only Chile (15 percent) and Panama (18 percent) reported much lower percentages. When asked about their satisfaction with police performance in their neighborhoods in 2014, a majority of Costa Ricans reported some level of satisfaction: 55.5 percent of respondents were satisfied or very satisfied. When asked about police response time, responses were generally positive from 2014 to 2018; on average, Costa Ricans thought that police would respond to a burglary call in thirty to sixty minutes. Overall Costa Ricans report one of the better police response times in the region, ranking fourth among Latin American countries and on par with Chile, Panama, Ecuador, Argentina, and Uruguay. Even so, this response time is significantly lower than the United States and Canada, where people estimate that the police will respond in less than thirty minutes, on average (Zechmeister 2016).

110

Making Police Reform Matter in Latin America

Figure 5.4

Police Bribery in Costa Rica (2004–2018)

Source: LAPOP’s 1999–2018.

Finally, turning to assess respondents’ faith that the justice system will hold criminals accountable, levels of trust have not changed significantly since 2004, despite a dip in 2006. Overall, there is only tepid belief that criminals will be sanctioned by the justice system, an indication that amid the current crime wave, Costa Ricans are skeptical about the efficacy of their justice system on this dimension.

Conclusion To understand Costa Rican exceptionalism in terms of policing and public security, a few things stand out. First, the 1949 constitution was a critical juncture for the Costa Rican security apparatus. It ushered in a stable democratic government, which has provided institutional stability for developing policing and public security institutions. The constitution created an expansive role for police, but also supported policing and public security reforms with simultaneous investments in social welfare policies. When levels of inequality are lower, police are not typically put in the position to act as bor-

Costa Rica Table 5.1

Trust in the National Police in Costa Rica (2018)

Independent Variables Level of education Age Victim of crime Police officer asked for a bribe Perception of neighborhood insecurity Sex Wealth quintiles Urban/rural Observations R2

111

Trust in National Police 0.051 (1.161) 0.086** (0.0457) −0.082** (0.0186) −0.115*** (0.0405) −0.158*** (0.0232) −0.024 (1.509) 0.011 (0.563) 0.053* (1.605) 1,440 0.069

Notes: Standardized beta coefficients. Standard errors in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001.

der guards between socioeconomic classes. More egalitarian policing practices tend to emerge when levels of social differentiation are low. Second, these investments in social welfare, particularly education and healthcare, help address many of the root causes of criminality. Thus, for most of the second half of the twentieth century, criminality in Costa Rica tended to be low, and police had the resources to maintain public order effectively. A technocratic approach tended to mark the design and implementation of public security policies, and there was elite consensus on the need for incremental police reforms. Overall policing and public security was not politicized. This chapter also points to the limitations of Costa Rican exceptionalism, which became more prominent in the early twenty-first century. The country succeeded in developing a modern, professional, civilian police force by the end of the 1990s, but changing domestic and international factors have dramatically altered the status quo in which the police operate. Domestically, Costa Rica has witnessed major changes because of economic disruption and sustained migration. Internationally, Costa Rica has increasingly become ensnared in thriving illicit markets, consequently raising the prevalence of criminality beyond the levels police have traditionally maintained. Costa Rica has increased its investments in public security and thus far maintained its balance between preventive and

112

Making Police Reform Matter in Latin America

repressive public security policies. This balance will be difficult to maintain, however, if illicit activity continues to thrive. Indeed, citizens appear to penalize the police for increases in crime, perceiving them as ineffective in controlling criminality and maintaining order. At the micro level, those victimized by crime, as well as those who fear future victimization, register significantly lower levels of trust in the police. At the macro level, average levels of trust in police declined just as homicide rates increased. If the police are perceived as ineffective in curbing crime, it seems the legitimacy of the institution could decline.

Notes 1. Even when inequalities increased because of the concentration of land ownership during the coffee boom of the late 1800s, Costa Ricans continued to imagine themselves as an egalitarian nation of small yeoman farmers (Seligson 1980). 2. The United States did intervene militarily in Costa Rica in 1921 when a border dispute erupted between Panama and Costa Rica; however, this was a brief naval intervention and did not involve the occupation of Costa Rican land. 3. According to the 1892 census, Europeans were the largest migrant group, followed by Nicaraguans and Jamaicans. By the twentieth century, census records document that immigrants hailed primarily from Central America and the Caribbean, and Nicaraguans made up the largest group. This trend intensified throughout the twentieth century; Nicaraguans represented 2 percent of the Costa Rican population in 1927, compared with 6 percent in 2000 (Castro 2011, 25). 4. Migration laws did not focus solely on nationality and race, however, as statutes also targeted people who were indigent, physically disabled, and associated with anarchist groups. 5. In addition to furthering aspirations for a “whiter” society, this prohibition aimed to isolate the unionization and labor activism of Black workers in the Atlantic zone and keep organized labor associations and activists from spreading their work to the Pacific zone (Seligson 1980, 68–69). 6. Because many Afro–Costa Ricans were not considered citizens, they were classified as foreigners (even though they were born in Costa Rica) and subject to this law. 7. Requirements for naturalization include the ability to read, write, and speak Spanish; documentation of good conduct; and the successful completion of an exam on Costa Rican history and values. For the complete text of the 1949 Costa Rican constitution, see Georgetown University’s Political Database of the Americas, http://pdba.georgetown.edu/. 8. For a historical overview of Costa Rican political development during this time, see Booth (1998). Booth, Wade, and Walker (2020) provide an overview of contemporary Costa Rican political development in the regional context of Central America. 9. The Gini coefficient is frequently used to measure inequality; it is a summary measure of the dispersion of income across an income distribution. A value of 0.0 reflects perfect equality within a group, while a value of 1.0 indicates maximum inequality. 10. Although unemployment rates did rise overall during this time, this was not a steady increase. In 2000 unemployment rates dropped to 5.1 percent, and in 2007 they declined to 4.6 percent before increasing again in 2009 (World Bank World Development Indicators n.d.).

Costa Rica

113

11. In the 1990s, homicide rates in Costa Rica were well below those of the United States: approximately 5 per 100,000 for the former and 9 per 100,000 for the latter. By the end of the 1990s, violent crime such as homicides began a steady drop in the United States at the same time that rates began to rise in Costa Rica. 12. According to the 2011 census, Nicaraguans make up 6 percent of the Costa Rican population, and these numbers have steadily risen since the 1990s (INEC 2011). Gatica (2011) approximates that there are roughly 130,000 undocumented migrants; when added to the numbers documented in the 2011 census, this would bring the percentage of Nicaraguans living in Costa Rica to 10 percent of the total population. 13. To be sure, there is not unilateral acceptance of xenophobic discourse. One play challenging xenophobia and highlighting discrimination in Costa Rica, “El Nica” received widespread acclaim and a 2003 Ministry of Culture award. Palmer and Molina (2004b) note that the Ministry of Culture has begun to emphasize the multiculturalism of Costa Rican society (230). 14. In 2006, growing opposition to Nicaraguan immigrants culminated in a punitive law that increased deportations, increased border security, and made legal residency requirements stricter (Orcés 2015, 1894). The Costa Rican legislature revised this law in 2009 and replaced the security-laden discourse of the earlier laws with language focused on human rights (Fouratt 2014). Still, the reformed legislation contained provisions that many found disconcerting, as it established a fee structure for participation in the public social security system and extensions and changes to migration status (Sandoval García 2015). 15. This new law still contained measures rendering migration proceedings more difficult and expensive, consequently impeding migrants’ ability to regularize their status in Costa Rica and access social services. 16. These comments were made during the 2010 Costa Rican presidential election campaign. They synthesize Chinchilla’s empirical investigations on public security promotion, such as her 2002 book, Citizen Security in Latin America: Towards a Holistic Policy (Rico and Chinchilla 2002). 17. Article 140 gives the power, in light of loss of confidence, to remove members of the public security force without requiring the official dismissal process. 18. In 2017, the Interagency Commission on Citizen Security and Living Conditions published a report on the proliferation of organized criminal activities in Costa Rica and the implications of this for democratic governance and citizen security. 19. Once again, the independent variables included in this regression are different from those of prior chapters, as different variables emerge as theoretically important in the Costa Rican context.

6 Nicaragua: A Return to Political Policing

SINCE THE EARLY 1990S, MANY OBSERVERS HAVE APPLAUDED police reform in Nicaragua and considered policing there to be an important exception to regional trends. While most of its neighbors became engulfed in high levels of violent crime, Nicaragua has consistently maintained low homicide rates. Furthermore, the country has largely refrained from politicizing crime. In contrast, elected officials in other parts of Central America have tended to campaign with an iron fist, favoring militarized approaches to policing that unleash violence and repression against suspected criminals. These zero-tolerance measures have rarely lowered homicide rates but have exacerbated cycles of repression and criminal violence (Wolf 2017). Nicaragua has largely eschewed these tactics, and until April 2018, state and societal forces tended to shun militarized and repressive policing practices, favoring preventive and community-based initiatives (Espinoza and Herrera Rodríguez 2009).1 The community orientation of Nicaragua’s national police force, which had long emphasized the limited use of violence and use of preventive programs, was a critical component to this country’s different trajectory. The community orientation of police and investment in prevention programs would be remarkable given trends in the region, but the Nicaraguan case was all the more unusual in that these reforms occurred in a country marked by some of the highest levels of poverty and institutional weakness in the region (Cruz 2015). These preventive and community-based strategies are often credited (at least partially) with maintaining comparatively low levels of crime (Sibaja et al. 2006).2 While Nicaragua has all of the preconditions for high rates of public insecurity and crime, in 2017 its homicide rate was the lowest in Central America. Like Guatemala, El Salvador, and Honduras, 115

116

Making Police Reform Matter in Latin America

inequality and political violence have marked Nicaragua for much of its history. The country reports a GDP of less than a quarter of that of Costa Rica, yet since the early 2010s, its homicide rate was roughly the same. Nicaragua’s violent crime rate has been 80 percent lower than Guatemala and El Salvador, despite registering lower levels of GDP per capita. While Nicaragua’s history and poverty are similar to its neighbors to the north, its crime rates are comparable to its neighbors to the south. Nicaragua had maintained comparatively lower crime levels while extolling the benefits of policing strategies that respect human rights, maintain strong ties to the community, and are holistic and preventive. For these reasons, observers had pointed to the Nicaraguan case as a potential model for other developing and democratizing countries (e.g., Stone 2012; Cruz 2014). Nevertheless, police attacks on demonstrators in April 2018 shattered perceptions that the Nicaraguan police could serve as such a model. Tragically, this was not an isolated incident. Overt, systematic police repression against dissidents and demonstrators has continued from 2018 to the present. According to the Comisión Interamericana de Derechos Humanos (CIDH) and the Asociación Nicaragüense Pro-Derechos Humanos (ANPDH), approximately three hundred people were killed in violent clashes with the police between April and August 2018, more than two thousand people were injured, and more than two thousand were arrested (many in arbitrary fashion or kidnapped by security forces) (CIDH 2018; Baldizon 2018).3 In April 2020, CIDH rapporteur Antonia Urrejola reported that the killing and jailing of dissidents had tapered off, however, this was because fewer Nicaraguans were willing to protest or criticize the regime, as they feared police repression (Wallace 2020). In other words, from the standpoint of the regime, systematic police repression yielded the desired results: stifling dissent and instilling fear in the population. The violent police crackdowns against protesters after April 2018 raise several questions. Were these attacks a dramatic departure from the police’s community orientation? Or were they the culmination of a police force gradually becoming more repressive and politicized? In retrospect, there is evidence that the widely publicized attacks revealed the increasing politicization of the institution since Daniel Ortega’s return to power in 2007. Indeed, before April 2018, there were less publicized reports of smaller scale crackdowns on protesters and dissidents. As early as 2008, the police did not act to curb violence perpetrated by groups linked to the government during the municipal elections (Cuadra 2018). The police also cracked down on social protest in rural areas, particularly those opposing canal construction in 2013 (Orozco 2015).4 Observers noted several cases where the police began to use disproportionate force in routine stops and excessive force against mine workers protesting in 2015 and protesters from Unidad Nacional del Adulto Mayor in 2013 (Baldizon 2018). According to these accounts, the

Nicaragua

117

2018 protests were not so much an abrupt departure from the status quo but a steady evolution of more repressive policing that began in 2007. These abuses of power did not receive extensive attention, perhaps because they were not as deadly as the abuses on full display in the other countries of Central America. In retrospect, these earlier episodes were perhaps harbingers of more extreme and widespread police attacks on dissidents in the future. In this chapter, we examine how Nicaragua created a community-oriented policing model and identify the weaknesses of this model that ultimately jeopardized its sustainability over time. We trace the evolution of policing practices in Nicaragua and rely on longitudinal survey data from 1996–2019 from the Latin American Public Opinion Project (LAPOP) to analyze public attitudes toward police as these reforms unfolded. We seek to understand how a police system once applauded for its ability to control crime without violating citizen rights was so quick to use lethal force against protesters and dissidents. We also analyze data from a survey of police officers in Nicaragua collected in 1997, a time when police reforms were consolidated under a democratic government and well before Ortega’s reelection in 2007. We complement our analysis of public opinion data with field research conducted in Nicaragua in 2011 and 2013. This chapter not only offers insights for Nicaragua but also has implications for the sustainability of police reforms in other democratizing countries.

The Evolution of Policing in Nicaragua Like most Latin American countries, the police in Nicaragua historically had served as an instrument of the government to support its power and control citizens. In 1936, Anastasio Somoza used the newly created National Guard to assume control of Nicaragua, establishing a family dynasty that ruled the country until 1979. The Somoza regime epitomized the use of police for repressive and political aims, as it relied heavily on the National Guard to repress political opponents and instill fear in citizens. The Sandinista-led revolution toppled the Somoza regime in 1979 and destroyed the authoritarian and personalist security forces. The revolution marked a significant critical juncture that led to deep changes in the country’s security apparatus. Military and police functions were separated constitutionally, legally, and operationally. The police sought to build close community ties and emphasized prevention and rehabilitation rather than punishment and repression. Postrevolutionary Reforms, 1979–1990 Following the 1979 revolution, Nicaragua made a decisive break from the repressive policing of the National Guard and created a new civilian police

118

Making Police Reform Matter in Latin America

force oriented toward the community with a mandate to prioritize prevention. The new government separated national defense from public security and created both an army and a police force to uphold each (Orozco 2015). As longtime former police director Aminta Granera explained, “We didn’t know how to be police. We only knew we didn’t want to be like the Somozan Guard” (The Economist 2012).5 According to one official who participated in the reforms immediately after the revolution, “What was clear was that we needed to create a police force that would not be repressive. That the police could never do what they had done in the past.”6 Although critiques of the early Sandinista government abound, there is consensus that the new police system was a vast improvement over its notorious predecessor, particularly in the arena of human rights (Weegels 2018a, 2018b). Reformers emphasized the need to elaborate rules (and clearly explain them) to limit the use of violence and develop strategies to deter the excessive use of force and strengthen preventive mechanisms. Nicaragua was unique in that it enshrined limitations on police use of force in its new constitution. Article 97 of the 1987 constitution explicitly stated that the police’s role is preventive, not repressive; this established an early constitutional precedent for creating an apolitical, professional, civilian police force that would respect citizen rights and the rule of law. The new National Police included a network of institutions at the local level that would not only prevent crime but also exercise government presence. The Committees for the Defense of the Sandinista Revolution, as well as the so-called heads of sector, served as security specialists that helped social cohesion and consolidated a political structure that bolstered the Sandinista government at the local level. For the first decade after the revolution, the Sandinista police were charged with guaranteeing the status quo, confronting crime, and protecting the state (Rocha Gómez 2005). Still, these policing reforms and new practices unfolded against the backdrop of the Contra war, which ultimately claimed the lives of nearly 34,000 Nicaraguans, with casualties heavily concentrated in rural areas (Seligson and McElhinny 1996). Growing insecurity and violence from the Contra war corresponded to increased repression in the Sandinista-led government, so the new police force operated in a context of high levels of violence and insecurity, alongside government coercion, until the end of violence in 1990. Reform Under Democracy, 1990–2006 The first cycle of police transformation closed in 1990. In a historic election in 1990, Nicaraguans witnessed a peaceful transfer of power when Violeta Chamorro (1990–1997) replaced Daniel Ortega (1985–1990) as president. This change in power reflected a postrevolutionary critical juncture that ushered in a second stage of police reform, which prioritized depoliticization,

Nicaragua

119

innovation, and increased community orientation. The new government approved the Law of the Ministry of Interior, and the protection of citizens was defined as a police function. According to Beltran (2009), subsequent political transformations in the 1990s prioritized police reform as an essential component for constructing a “firm and lasting peace” (1), as the 1990 loss of the Sandinista party led to new relationships among the police, the victorious coalition led by President Chamorro, and the population. While the process of institutional depoliticization had begun, Rocha Gómez (2005, 2007) argues that the Sandinista structure in the police persisted. In this changing electoral landscape, Cruz (2014) contends that “Law enforcement leaders understood that in order to survive, they had to work with the community and be efficient” (2). Police training and orientation tended to stress the nonrepressive nature of the police and emphasized their role as one who responds to the needs of the community.7 Such an emphasis is an integral part of community-oriented policing strategies, which prioritize the community perspective in police formation and mechanisms for a public accounting of police action (González 2016; Muggah 2017; Ungar 2019). Here, the Nicaraguan National Police have stood out in the region, as police reformers went beyond mere rhetoric and implemented meaningful reforms that were communitarian, proactive, and preventive (Espinoza and Herrera Rodríguez 2009; McNeish, Prado, and Ehrlich 2019). Still, overall investments in police personnel have been comparatively low. The number of police personnel per 1,000 inhabitants ranks among the lowest in the region, with lower numbers only in Guatemala.8 Nicaragua pioneered several innovations in its policing strategies. For example, it recruited a far greater percentage of women into police ranks, and by 2009, approximately one third of the force was female (Beltran 2009; Drysdale-Walsh 2019). It was one of the first countries in the region to open specialized women’s police stations in 1993, and these stations were staffed exclusively by women to provide comprehensive support for victims of violence. There are some indicators that these specialized stations and recruiting women into police forces have generated more public trust in the police. In a comparative analysis of violence against women in Nicaragua and Guatemala, Drysdale-Walsh (2019) found that as of 2011, reported cases of intrafamilial or domestic violence were three times higher in Nicaragua than Guatemala, even though actual rates of violence against women in these countries were comparable. This indicates that women were more willing to come forward to police in Nicaragua. Additional specialized policing units have addressed the security of families and the needs of youth (Cruz 2014). Indeed, Sibaja et al. (2006) point to the ways the Nicaraguan policing model had been supported by parallel investments in social welfare and criminal justice innovations. As we noted in Chapter 1, these investments are critical because they can

120

Making Police Reform Matter in Latin America

address the root causes of criminal careers. When investments in justice reform are matched with those of social protection, these reforms can buttress one another and provide a more holistic response to addressing crime. These initiatives are often credited with reducing the numbers of disaffected youth who join organized criminal groups in Nicaragua (Rocha Gómez 2006, 114; Espinoza and Herrera Rodríguez 2009). Nicaragua’s emphasis on crime prevention makes it harder for gangs to recruit disaffected youth and use prisons as training grounds for new members. Nicaragua has focused “most of its efforts on prevention and intervention, which have had important results in reducing criminality and youth violence” (Sibaja et al. 2006, 8). Even incarcerated populations have access to some rehabilitative measures, such as activities and job training. Some youth join smaller-scale neighborhood gangs (pandillas), but large, sophisticated gangs have not established a stronghold in the country.9 Policing strategies are not the only reason gangs and organized crime have not fully penetrated Nicaraguan society, but they are an important part of the story (Sibaja et al. 2006). While violent crime poses problems for governance, organized crime is an even more formidable opponent, as it tends to be a well-financed, wellorganized counterpart with ready access to weapons and ammunition. In many cases, organized crime can systematically challenge the state’s monopolization of force and legitimacy. This has certainly been the case in many Central American countries, as the Mexican antidrug offensive has pushed organized crime into Central America, particularly the Northern Triangle of Guatemala, Honduras, and El Salvador. In Nicaragua, throughout the 1990s and 2000s such groups were not powerful enough to challenge the state’s monopoly of force, leading to homicide rates that are often the lowest in the region. Nicaragua’s low rates of violent crime, and ability to insulate itself from the infiltration of gang members, are often cited as evidence of successful policing strategies. While the Nicaraguan policing model highlights the importance of a community orientation, some critics have argued that control measures were not completely canceled and violent incursions were routine (Rodgers 2006). To be sure, the emphasis on prevention and community orientation does not mean that harsher, more punitive measures have been completely absent. In 1999, the police cracked down on small neighborhood gangs, but changed course the next year to bring policing practices into conformity with the constitutional mandate. In 2005, a punitive antigang law was debated in the National Assembly; it was ultimately defeated after legal experts argued it would violate the constitution (Sibaja et al. 2006). Some argue that community policing in Nicaragua included policies aimed at young people at risk but often stigmatized and kept them in the system (Weegels 2017). Still, while some harsher measures have surfaced over the years, before 2018 they tended to be few and have a short life span; they

Nicaragua

121

did not define the nature of police work as they did in other parts of Central America. Finally, as we noted earlier, some critics argued that the process of institutional depoliticization was incomplete and the Sandinista structure in the police persisted despite efforts to transform the police into a truly apolitical organization (Rocha Gómez 2005, 2007). A survey of police officers conducted in 1997 allows us to examine the merit of some of these critiques. LAPOP conducted a national probability sample of 2,500 Nicaraguans, as well as 1,100 specialized interviews with police officers, the military, teachers, and individuals affiliated with various nongovernmental organizations.10 We focused on the 150 police officers who were interviewed face to face while at work. These interviews took place seven years after the end of the Contra war and the election of a democratic government, well before Ortega’s 2006 election. In 1997, President José Arnoldo Alemán had just begun his presidential term, replacing the broad-based coalition of President Chamorro. Alemán’s party originally ran on a strong anti-Sandinista platform, before Alemán made an infamous power-sharing pact with Ortega in 1999.11 Thus, this 1997 survey provides a unique opportunity to assess police attitudes toward the public and the government and compare police attitudes to those of the public at large, at a critically important historical juncture. Of the 150 police officers interviewed, 78 percent were men and 22 percent were women. On average, the police in this sample had completed or nearly completed their secondary education and were thirty-one years old. When asked about their ideological orientation, 36 percent identified with the far left, and 27 percent squarely in the middle of the ideological spectrum.12 Not many police identified with the center right or the far right—only 21 percent placed themselves to the right of the political center.13 The ideological orientation of the police in 1997 was geared far more to the left, compared with the general public. On this ideological scale, which ranged from 1, left, to 10, right, police scored 4.0 on average. In contrast, the general public scores on this same measure were almost two points higher, indicating that the ideological spectrum was much broader among the general public. This pattern holds when we examine partisan affiliation and voting intentions. As Table 6.1 reports, when asked who they voted for in the 1996 presidential elections, 76.1 percent of the police indicated that they would vote for Ortega’s party, compared with only 34.7 percent in the general public. These partisan differences carried over into evaluations of President Alemán’s government. Police registered significantly lower levels of approval, and when asked whether Alemán’s government helped or hurt in reducing crime, 22.8 percent of the public replied that his government had helped, compared with only 13.8 percent of the police. On a variety of indicators, the police skewed to the political left and were much more supportive of Ortega even after seven years of reforms under a democratic government.

122

Making Police Reform Matter in Latin America

Table 6.1 illustrates that there were other differences between the police and the public. The police were more likely to state that respect for human rights had improved since 1990, as 67.8 percent of police thought human rights had improved, compared with 57.0 percent of the public. The police also supported more spending on the military (78.5 percent); here, the majority of the public (58.5 percent) was also in favor of increased military spending, but this support was still well below that of the police. Still, when asked about the role of the military, police respondents registered far less support for military coups than the general public on average. Only 20.7 percent of police respondents indicated they would support a military coup, compared with 28.2 percent of the general public. Police respondents also registered more support for political institutions. Given police attacks on protesters in 2018, it is worthwhile to examine police attitudes toward protest at this earlier stage of the police’s institutional development. In 1997, LAPOP posed a series of questions to the public and the police to gauge their level of support for different types of protest activities. Among these items were the following three questions: • “I’m going to read a list of actions or things that people do to achieve their political goals and objectives. Could you tell me how much you (10) approve or (1) disapprove that people engage in these actions . . . participate in lawful protests?” • “There are people who only talk bad about the government. How much do you (10) approve or (1) disapprove that these people carry out peaceful protests to in order to express their views?”

Table 6.1

Comparison of Police and Public Political Attitudes in Nicaragua (1997)

Percentage Reporting That . . . They voted for Ortega in 1996 They voted for Alemán in 1996 President Alemán’s government   helped reduce crime Human rights situation has 1990   improved since Military budget should increase Would justify a military coup that   interrupts a democratic process They have high levels of respect for   political institutions

Public

Police

34.7 56.4 22.8

76.1 18.2 13.8

57.0

67.8

58.5 28.2

78.8 20.7

61.4

69.6

Source: Nicaragua 1997: Cultura política y democratización en Nicaragua, LAPOP, https://www.vanderbilt.edu/lapop/nicaragua.php.

Nicaragua

123

• “If a law were passed prohibiting public protests, how much would you (10) approve or (1) disapprove?” As Figure 6.1 illustrates, police respondents registered high levels of support for citizens’ right to protest peacefully. When asked about citizens protesting peacefully to achieve their goals, police support was extremely high on average—8.6 on this 1–10 scale. When the question primed police respondents to think more concretely about people who talk badly about the government, approval of protest activities declined significantly, but was still high at 7.2 points. In a similar vein, police respondents tended to disapprove of laws that would limit citizens’ rights to protest, with average approval hovering at only 3.8 points. When we compare police responses to these three questions to those from the general public, we see that police respondents were more supportive of citizens’ right to protest on average.14 Overall, our survey data from 1997 indicate that after seven years of police reforms under democracy and the peaceful transition of presidential power from one party to another, police respondents tended to have higher levels of respect for political institutions and low levels of support for military coups. Police registered very high levels of support for people’s right to protest peacefully, even when they critiqued the government. These political attitudes bode well for democracy, as they indicate that police Figure 6.1

Police Attitudes Toward Protests in Nicaragua (1997)

Police Law passed outlawing protests Critics of the government protesting peacefully People participating in peaceful, lawful protests

Public

Error bars: 95% CI

Strongly disapprove

Strongly approve

Source: Nicaragua 1997: Cultura política y democratización en Nicaragua, LAPOP, https://www.vanderbilt.edu/lapop/nicaragua.php.

124

Making Police Reform Matter in Latin America

strongly supported key democratic norms and rights. However, our survey data indicate that the police had stronger partisan ties to the Sandinistas than the general public did and tended to skew more to the left. These partisan leanings provide some evidence that the goal of depoliticizing the police was perhaps only partially achieved. Reform Versus Retrenchment: 2006–Present Ortega’s 2006 presidential victory was another critical juncture for Nicaragua’s policing structures. While some of his early efforts to co-opt the police were thwarted by opposition in the legislature, Ortega’s 2007 inauguration ultimately created an opening for police reforms that had long-term disastrous consequences, particularly by undermining institutional structure and rendering the police more squarely under the control of the executive. Under Ortega’s presidency, Orozco (2015) argues that the National Police began to take on two roles. On one hand, the police continued to uphold their standard public security responsibilities (e.g., investigating crime, ordering traffic, apprehending and detaining suspected criminals). On the other hand, they became more vested in guaranteeing the security of Ortega’s regime and began to define security threats as not just those that jeopardized public safety but also those that challenged the regime. Protests against canal expansion and electoral reform were reframed as security threats, and the police began to crack down on protesters and potential regime opponents, especially in the countryside. Police stood by passively and did not intervene when groups linked to the government engaged in violence during the municipal elections in 2008 (Cuadra 2018). Reports of police using disproportionate force in routine stops began to circulate with greater frequency. Baldizon (2018) chronicles police abuse of mine workers protesting in 2015, which injured twenty-five people and led to the arrest of twenty-three protesters. Police also accosted protesters from Unidad Nacional del Adulto Mayor in 2013. Legal changes to the police were indicators of growing politicization. After a series of thwarted attempts, Ortega succeeded in passing a major change to police practices in 2014 (Law 872) and was able to eliminate the Ministry of Governance, which had supervised police and served as an intermediary between the president and chief of police. Ortega ensured that the chief of police would serve more clearly at the behest of the president, leading several observers to caution that these new laws ran the risk of politicizing policing (Mejia Giraldo 2014). At the local level, committees that had ostensibly existed to promote citizen participation and public security began to serve the regime more overtly. As Orozco argues, these Gabinetes de Familia (Family Cabinets)

Nicaragua

125

know who’s in the opposition in each block on each street; they know who everyone’s going to vote for, who supports the government. . . . They’ve got it all mapped out at the base, performing functions similar to those of the Sandinista Defense Committees (CDS) in the eighties, exercising social control at the neighborhood level, gathering information, identifying threats and informing. (2015, 22)

Nicaragua became more vulnerable over time to the influence of organized crime and corruption. For example, Interior Minister Ana Isabel Morales sharply rebuked judges and magistrates who consistently reduced prison sentences (or simply released) drug traffickers, which she cited as evidence of organized crime’s influence over the justice system. In a highprofile trial, officials from the Nicaraguan National Police and Supreme Electoral Council were charged with complicity in money laundering and drug trafficking operations (Meléndez and Orozco 2013). Nicaragua’s investment in specialized police forces also began to languish. While the women’s police stations were widely applauded in the late twentieth and early twenty-first centuries, more recent analyses of these stations have identified problems of sustainability, as resources have been siphoned off for other purposes, leadership has weakened, and conflicts between the broader women’s movement and the government have eroded the political will to address violence against women (Drysdale-Walsh 2019, 186–187). Indeed, in 2019 La Prensa reported that the government had falsified data concerning women’s police stations and steadily closed several stations and reduced women’s access to support and justice (Silva 2019). Indeed, several scholars point to Ortega’s 2006 election as a watershed year for police institutions. Cuadra (2018) contends that the 2018 repression is rooted in changes to the police that began when Ortega was elected to his first term. She argues that when President Ortega assumed office, he began to co-opt the police, allowing grupos de choque (repressive forces) to form inside the institution that could be used for political purposes. In 2007, these groups originally formed ostensibly to provide protection to Ortega and donned blue shirts to distinguish themselves. Some journalist accounts noted that this group consisted of approximately 150 men who were former members of the military, security personnel, or other Ortega sympathizers (Salinas and Solano 2009). Rocha Gómez (2008) argues that the government recruited youth from former pandillas to join these grupos de choque (often from the youth programs run by the National Police) and that they operated with the tacit approval of police. Over the course of ten years, Cuadra (2018) argues that these groups evolved into full-fledged para-police, which were unleashed with violent consequences in April 2018, and that the National Police coordinated and directed the activities of these groups. Over time, these groups expanded to cover larger territories,

126

Making Police Reform Matter in Latin America

with the complicity of the mayors’ offices in several major cities, such as Managua, Matagalpa, León, and Jinotega.

Policing and Crime Trends Even as these developments were taking place, empirical indicators on the ground, particularly in urban areas, painted a very different picture of public security policy in Nicaragua compared with its counterparts in the Northern Triangle. For many sectors of the population, the police with whom they had daily contact performed much as they had in the past. Nicaragua continued to invest in a more holistic and preventive approach to fighting crime and maintaining public security. As noted earlier, a holistic approach combining justice reforms with social protections creates a matrix of cross-cutting reforms that can buttress one another. This different policing style is partially credited with containing rates of violent crime. After the Cold War, violent crime rates in Nicaragua were starkly different from the Northern Triangle countries. Postconflict countries are particularly vulnerable to infiltration by organized crime, as political and economic transitions take place against a backdrop of demobilized soldiers and incomplete disarmament. Indeed, in two postconflict countries where organized crime has been prevalent, El Salvador and Guatemala, homicide rates have been very high (Malone 2014).15 Nicaragua follows a different trajectory (see Figure 6.2). Despite plenty of demobilized combatants and a disarmament process that generated criticism, rates of violent crime have risen only modestly. According to World Bank estimates, following the 1990 elections in Nicaragua, homicide rates decreased substantially, declining from a high of 15.2 per 100,000 inhabitants in 1995 to 11.2 in 1999. This is not the typical scenario of postconflict countries, as El Salvador illustrates most tragically. While homicide rates in Nicaragua rose from 2000 to 2006, they declined and in 2013 fell below those of Costa Rica, the country widely regarded as the most peaceful in the region. Nicaragua was able to defy the trends of most postconflict countries, and despite its endemic problems of poverty, its crime rates remained comparatively quite low, even as the police became more politicized under Ortega.16 These promising trends were interrupted in 2018, when police became responsible for an increase in the homicide rate. While statistics on the killings of protesters are incomplete, civil society and nongovernmental organizations (such as the CIDH and the ANPDH) approximate that 300 people were killed between April and August 2018, and 595 disappeared as of July 2018. This breaks down to three or four murders per day, a situation that marks an important departure from 2017, when Nicaragua recorded a total of 347 murders. Due to government threats, the major civil society groups have not made additional data available (to our knowledge). According to the National

Nicaragua Figure 6.2

127

Homicide Rates per 100,000 People in Nicaragua (2005–2020)

Source: World Bank, World Development Indicators and InSight Crime, Homicide Roundup.

Police of Nicaragua, the homicide rate in 2018 was 11 per 100,000—a clear reversal of earlier promising trends (Policía Nacional de Nicaragua 2018). Despite this sharp reversal in 2018, it is remarkable that Nicaragua was able to contain increases in violent crime for an extended period of time given broader trends in the region. Still, while homicide rates are a valuable and widely used indicator, they do not capture the prevalence of other types of crime. Official statistics on other types of crime (e.g., burglary, assault, and theft) are notoriously unreliable, particularly because they tend to gauge citizen trust and willingness to turn to the police rather than the actual occurrence of crime. For that reason, scholars and practitioners typically turn to national surveys to gauge trends in other types of crime. We rely on LAPOP’s national surveys to measure rates of self-reported victimization in Nicaragua and then compare these rates to broader trends in Latin America. These surveys show that self-reported victimization rates in Nicaragua rose slightly from 2004 to 2019, despite a drop in 2012. Independent sample t-tests confirm that rates of self-reported victimization

128

Making Police Reform Matter in Latin America

increased by 2.5 percent between 2004 and 2019. Although this increase was not large, it was statistically significant. With the exception of 2014, self-reported victimization rates in Nicaragua were on par or lower than those of Costa Rica and lower than the regional average. Particularly when compared with other postconflict countries, Nicaragua has managed to contain victimization rates of these other types of crime. Public perceptions of crime have also not been exaggerated or politicized. It might seem intuitive that if crime rates are low, public fear of crime will also be low. However, as we saw for Chile (Chapter 3) and Costa Rica (Chapter 5), this is not always the case. Although there is overlap between the occurrence of crime and public fear of crime, other factors play a role in shaping public perceptions of crime, such as elite discourse, media coverage, and anxiety about other socioeconomic societal changes. In Figure 6.3, we see that average levels of fear of crime in Nicaragua are consistently low, which reflects a lower incidence of crime and a lack of politicization of crime. Political elites and the media have largely refrained from sensationalizing and/or politicizing crime, and in public opinion surveys, the public has largely not prioritized crime as a major societal problem. This overview of policing and crime trends in Nicaragua illustrates two things. First, despite some well-warranted criticism, as well as politicization under Ortega, before 2018 Nicaraguan police emphasized communityoriented approaches and refrained from the wide-scale adoption of militarized policing tactics that were common in other parts of Central America. Second, this community-oriented strategy has partially contributed to very different trends in violent and nonviolent crime. Some of these community-oriented initiatives opened the door to the politicization of the police and public security. Indeed, some of the structures that contained crime and impeded the growth of large transnational gangs lent themselves to politicization. Sibaja et al. (2006) note that “lingering socialist structures such as the neighborhood watch” as well as the effect of the Committees for the Defense of the Sandinista Revolution have helped civil society self-police their neighborhoods and keep more dangerous foreign gangs out (5–6). Yet recent analyses have highlighted that this community-oriented approach was also politically oriented, as it eventually included a series of institutional structures such as the Gabinetes de Familia or jefes de sector (neighborhood officers) to consolidate social capital. This approach increased police linkages with the community but also allowed the police to gather political information on specific people and organizations (Orozco 2015; McNeish, Prado, and Ehrlich 2019, 16). Such civil society groups run the obvious risk of politicization, and vigilance of organized crime can swiftly be converted into vigilance of political opponents. Indeed, Ortega moved quickly to create the Consejos del Poder Ciudadano (CPC, Councils of Citizen Power) by presidential decree in 2007, ostensibly to strengthen community organization and promote the citizen partici-

Nicaragua Figure 6.3

129

Fear of Crime in Nicaragua (1999–2019)

Source: LAPOP’s 1999–2019 National Surveys, https://www.vanderbilt.edu /lapop/nicaragua.php.

pation in government. Among other functions, the CPCs included positions to liaise with the National Police to engage local communities in issues pertaining to public safety.17 In 2014, the Gabinetes de Familia, Comunidad y Vida (Cabinets of Family, Community, and Life) replaced the CPC, and these organizations were widely criticized as increasing government control over civil society. To track the growing politicization of the police, we examine how citizens have viewed their police during these processes of reform and retrenchment. We rely on LAPOP’s public opinion data from 1996–2019, which asked citizens a series of questions about their interactions and views of the police at several critical junctures.

Public Perceptions of the Police LAPOP first included survey questions on the police in its 1996 national survey, asking respondents: “To what extent do you trust the National Police? (1) not at all – (7) very much.” This item was part of a battery of

130

Making Police Reform Matter in Latin America

questions measuring trust in a wide range of institutions (e.g., legislature, courts, elections). We chart the average level of trust Nicaraguans registered in the police from 1996 to the present in Figure 6.4. As a point of comparison, we include average levels of trust in other political institutions over the same time.18 Survey data indicate that in terms of garnering public trust, the Nicaraguan approach to policing appears to have yielded some dividends. As Figure 6.4 illustrates, average trust in police was on par with that of other institutions in 1996 and dipped below them in 1998. However, by 1999 trust in police was higher than the average of other institutions and remained higher through 2016, when the gap started to narrow. Trust in police declined precipitously in 2019, however, implying that the public reacted to the 2018 crackdowns with significantly lower levels of trust. In addition to these longitudinal indicators, LAPOP included questions pertaining to different dimensions of police performance in different waves of the AmericasBarometer survey. In 2004, LAPOP focused on people who had direct experience with the police alongside a variety of government

Figure 6.4

Trust in Police in Nicaragua (1996–2019)

Source: LAPOP’s 1996–2019 National Surveys, https://www.vanderbilt.edu /lapop/nicaragua.php.

Nicaragua

131

entities. By that year, democratic governments had consolidated their overhaul of policing institutions and consistently promoted community-oriented and holistic public security policies. Thus, these survey questions help us gauge perceptions of police after more than fourteen years of democratization, yet before Ortega’s presidential term, which several observers argued ushered in an age of politicization (e.g., Cuadra 2018). LAPOP asked respondents, “After using the services of the following institutions, are you (4) very satisfied, (3) somewhat satisfied, (2) somewhat unsatisfied or (1) very unsatisfied?” The entities included were the National Police, courts or justice tribunals, public prosecution offices, and respondents’ municipalities. Respondents who did not have interactions with these entities were excluded.19 Nicaraguans who had personal interaction with the police were not particularly satisfied with these interactions. Whereas average levels of satisfaction are not abysmally low (2.5 on a 4point scale), they were quite tepid and statistically identical to levels of satisfaction with other government agencies. In 2004, LAPOP also gauged perceptions of police corruption vis-à-vis other public and private actors: “I am going to name various public and private institutions. I am interested to know the degree to which you think that the representatives of these institutions are honest or corrupt. I will ask you to grade each one from 1 (very corrupt) to 10 (very honest).” On this indicator, the police scored in the middle of the range at 4.9 points. Police fared the same as other political institutions, such as the courts, president, and mayors; only religious officials, professors, and the military score significantly higher than the police, whereas ministers, deputies, and leaders of political parties rank significantly lower. In 2008, LAPOP expanded on its battery of questions assessing daily interactions with police and asked respondents how much they would trust the police to apprehend criminal suspects if they were victimized by a crime. That year marked a milestone, as the survey was administered after Ortega’s first year of office. While Ortega began to politicize the police at the beginning of his term, it is unlikely that average citizens would have noticed a marked change in policing practices in their daily routines. Average responses in Nicaragua fell only in the midpoint of the range, indicating that respondents had limited faith that the police would apprehend suspected criminals if they were victims of a crime. Still, Nicaraguans were more optimistic compared to their counterparts in Mexico and Central America. In 2004–2019, LAPOP asked respondents to evaluate the ability of the justice system as a whole to punish the guilty party. Average responses hovered just above the midpoint of the scale, indicating moderate amounts of trust that the justice system could hold criminal accountable. In 2008, LAPOP included questions to gauge another key measure of public confidence in the police: willingness to report crime. Indeed, for

132

Making Police Reform Matter in Latin America

many crimes, rates of reporting reflect public confidence that police will do their jobs well, rather than rates of actual crime. Crime reporting is a valuable behavioral measure of trust in police. LAPOP asked respondents who had been victimized in the past year whether they reported the crime to the police. Of the 471 respondents who self-reported some type of victimization, 44.4 percent stated that they had reported the crime to the police. The most frequent reason for not reporting a crime was that “it wouldn’t do any good (49 percent).” Fourteen percent responded that it was dangerous, or they feared reprisals, and another 14 percent stated that they didn’t have any evidence. Only 10 percent stated that they didn’t report the crime because it was not that important.20 In 2008 and again in 2012, LAPOP measured perceptions about the legality of police activities: “Some people say that the police in this community (town, village) protect people from criminals, while others say that the police that are involved in the criminal activity. What do you think?” Responses were recoded so that (1) police protect; (2) police do not protect, but are not involved in crime or protect and involved in crime; and (3) police involved in crime. Here, the majority of respondents in each survey wave indicated that police primarily protected respondents, but a sizable group thought the police were primarily involved in crime (approximately one quarter of respondents). There was also a slight deterioration of positive evaluations of the police between 2008 and 2012, coinciding with Ortega’s return to power and an increase in organized criminal activity, particularly as drug trafficking routes shifted to Central America. Between 2008 and 2012, the percentage of respondents who reported that the police served primarily to protect communities declined slightly from 59.5 percent to 54.7 percent. In 2014, LAPOP measured satisfaction with police performance specifically in respondents’ neighborhoods. Rather than generalized trust, this item aims to gauge satisfaction with police specifically in the respondent’s immediate environment. When compared with the other Central American countries and Mexico, the Nicaraguan police ranked third on this measure, just behind Panama and Costa Rica, with an average score of 2.5 (on a 4-point scale). In 2014 and 2016, years that witnessed a series of police reforms that troubled several observers, LAPOP included a question to measure perceptions of police responsiveness, asking respondents to estimate how long it would take for police to arrive if they were victimized by a burglary (1 = less than 10 minutes, to 6 = there are no police/they would never arrive). While this is not a direct measure of citizen–police interactions, it does gauge citizens’ anticipation of responsiveness to their calls. On average Nicaraguans did not have high hopes that the police would respond efficiently if they called for help, estimating average police response time to fall between one and three hours in both years. In a regional context,

Nicaragua

133

Nicaraguan evaluations of police time were quite pessimistic, with some of the longest estimated response times in Latin America. As indicated earlier (Figure 6.4) public trust in police is higher vis-àvis other political institutions, yet the police have consistently earned a reputation for corruption. As Figure 6.5 demonstrates, the prevalence of bribery rose steadily from 2004 through 2008, and even though it declined in 2010, by 2014 it had rebounded again and continued to increase through 2016. Except for 2014 and 2016, the percentage of Nicaraguans who reported that police solicited a bribe was slightly lower than the Latin American regional average. Police bribery still exacted a toll, however. According to bivariate analyses, police bribe solicitation has a significant and negative impact on trust in police (r = −0.146, p < 0.01). In addition to charting these longitudinal trends in public trust in police, we wanted to determine whether attitudes toward the police were polarized alongside partisan lines. This is one way to measure the extent to which the police were politicized. Earlier we discussed the institutional changes during Ortega’s presidency that transformed the community-oriented police into a

Figure 6.5

Police Bribery in Nicaragua (2004–2019)

Source: LAPOP’s 2004–2019 AmericasBarometer, https://www.vanderbilt.edu /lapop/nicaragua.php.

134

Making Police Reform Matter in Latin America

political force. Given this evidence at the institutional level, we wanted to determine whether this politicization was reflected in police–community relationships more broadly. One way to determine whether the public at large perceived the police in more politicized terms is to empirically measure whether political identification with Ortega is linked to trust in the police. To this end, we conducted a series of regressions to examine the relationship between trust in police and political affiliation (alongside the traditional socioeconomic and demographic controls typical in survey research). Here, we rely on the LAPOP national survey data from 2008–2019, which contained identical survey questions that allow us to gauge the relationship between political affiliation and trust in police in each survey wave after Ortega’s inauguration. We measured support for Ortega with the following survey question: “If the next presidential elections were held today, what would you do? (1) wouldn’t vote; (2) would vote for the candidate or party of the incumbent; (3) would vote for some candidate or party different from the incumbent; (4) would vote but cast a null or blank vote.” We recoded these responses so that a value of one indicated the respondent’s intention to vote for President Ortega, the incumbent, and zero that the respondent would vote for the opposition, not vote, or cast a null vote. In our models, we include this measure of support for Ortega alongside other variables found in the literature to be significantly linked to trust in police: police corruption, fear of crime, and crime victimization (Malone and Dammert 2020).21 We also included the socioeconomic and demographic variables standard in survey research to control for individual attributes. To take these characteristics into account, this analysis follows the conventions of survey research and incorporates variables measuring sex, age (measured in cohorts),22 education (measured as the number of years of formal schooling respondents completed), the size of respondents’ hometowns,23 and income according to the number of household possessions owned by respondents.24 Table 6.2 reports the results for these ordinary least squares regression models.25 While the intention to vote for the incumbent (President Ortega) in the next presidential election is significant in every survey year, over time the size of this coefficient increased steadily. In 2008, willingness to vote for Ortega was significant but had a negligible effect on trust in police; those who supported Ortega were 0.560 more trusting in police than those who did not (on a 7-point scale). The model predicted only 4.0 percent of the variance in trust in police.26 By 2016, the magnitude of the support for Ortega variable had almost doubled, and the amount of variance the model predicted in the dependent variable of trust in police was 12.1 percent. In 2019, we see an even sharper contrast. The size of the coefficient for the support for Ortega variable is almost four times larger than it was in 2008, indicating that respondents who intended to vote for Ortega in the next

Nicaragua

135

election had 2.185 higher levels of trust in police than those who did not. On a seven-point scale, the impact of this variable on trust in police is quite large. The overall model predicted 28.2 percent of the variance in trust in police. Together, these results indicate that partisanship played a small role in predicting trust in police in 2008, but in 2019 public trust in police was heavily contingent on people’s views of the incumbent government. Trust in police became heavily politicized among the mass public after the police crackdown on protesters in 2018. Indeed, the politicization of trust in police is on par with that of more overtly partisan organizations that have also played a role in public security: the CPC and the Gabinetes de Familia, Comunidad y Vida. In 2008, 2010, and 2012, LAPOP included questions measuring trust in the CPC as part of the battery of questions measuring trust in other institutions. In 2014 and 2016, this question was replaced with one measuring trust in the Gabinetes de la Familia.27 We ran the same regression model in Table 6.2 to predict trust in these two more overtly political organizations (Table 6.3 reports

Table 6.2

Trust in Police and Willingness to Vote for Governing Party in Nicaragua (by Year) 2008

2010

2012

2014

2016

Constant

2019

4.259*** 4.775*** 4.572*** 5.661*** 5.031*** 5.159*** (0.198) (0.191) (0.189) (0.195) (0.200) (0.216) Vote for 0.560*** 0.721*** 0.851*** 0.797*** 1.015*** 2.185***   incumbent (0.127) (0.104) (0.095) (0.095) (0.100) (0.119) Police solicited −0.504** −0.753*** −1.077*** −1.057***−0.834*** −0.867***   bribe (0.188) (0.189) 0.217 (0.176) (0.155) (0.168) Fear of crime −0.208*** −0.074 −0.112* −0.424*** −0.129** −0.372*** (0.057) (0.051) 0.052 (0.048) (0.054) (0.050) Crime −0.478** −0.319* −0.474** −0.243 −0.360** −0.074   victimization (0.144) (0.125) 0.144 (0.131) (0.133) (0.138) Gender 0.178 0.074 0.068 −0.028 −.015 −0.452*** (0.105) (0.097) (0.096) (0.096) (0.098) (0.103) Income 0.009 0.217 0.064 −0.058 −0.757** −0.420 (0.293) (0.297) (0.297) (0.280) (0.278) (0.274) Education 0.015 −0.017 −0.023 −0.035** −0.026 −0.033* (0.014) (0.014) (0.014) (0.013) (0.013) (0.014) Municipality 0.015 −0.105** 0.018 −0.003 −0.012 −0.143*** (0.038) (0.034) (0.038) (0.036) (0.036) (0.036) Age 0.037 −0.031 0.009 −0.035* −0.012 −0.005 (0.020) (0.018) (0.017) (0.017) (0.017) (0.018) Adjusted R2 0.040 0.062 0.073 0.136 0.121 0.282 N 1,454 1,507 1,676 1,535 1,493 1,484 Notes: Coefficients are unstandardized, with standard errors in parentheses. *p < 0.05; **p < 0.01; ***p < 0.001.

136 Table 6.3

Making Police Reform Matter in Latin America Trust in Consejos del Poder Ciudadano and Gabinetes de la Familia, Comunidad y Vida (by Year) 2008 (CPC)

Vote for incumbent N

2010 (CPC)

2012 (CPC)

2014 2016 (Gabinetes de (Gabinetes de la Familia) la Familia)

1.867*** 1.797*** 1.868*** 1.170*** (0.134) (0.107) (0.100) (0.107) 1117 1437 1653 1362

1.103*** (0.103) 1418

Notes: This table presents only the coefficients for the primary variable of theoretical interest. Coefficients are unstandardized, with standard errors in parentheses. *p < 0.05; **p < 0.01; ***p < 0.001.

results). Although one must use caution when comparing results across different models, we see that the coefficient of vote for the incumbent is larger in the 2019 model predicting trust in police than it is in any of the models predicting trust in the CPC or the Gabinetes de la Familia. These survey findings demonstrate that trust in police increased from 1996 through 2016 and for much of this time was significantly higher than average trust in other political institutions. However, trust in police declined following the 2018 repression of protesters and dissidents. As Figure 6.6 illustrates, trust in police becomes highly politicized after 2018. There are sharp partisan divides in how the public views the police in 2019, as Ortega’s supporters registered significantly higher levels of trust in police, just as this trust plummeted for those who do not support Ortega. While support for Ortega did shape attitudes toward police from 2008 to 2016, the 2018 police attacks on protesters sharpened this divide. It appears that politicization may not have registered with the public until after the 2014 legal changes and the deeply polarizing events of 2018.

Conclusion Nicaragua’s police forces underwent significant changes after the 1979 Sandinista revolution. The revolution represented a critical juncture that broke the authoritarian and personalist “path” in which the structure of security forces was established in the 1930s when the Somoza family took control of the National Guard and the state. The 1990 elections ushered in another period of reform that further demilitarized the police and deepened its commitment to community ties and preventive policies. Investments in social welfare supported these reforms by targeting some of the root causes of crime. Unfortunately, the return of Daniel Ortega to power in 2007 and

Nicaragua Figure 6.6

137

Average Trust in Police and Support for President Ortega (2008–2019)

Source: LAPOP’S 2008–2019 AmericasBarometer, https://www.vanderbilt.edu /lapop/nicaragua.php.

subsequent reforms undermined much of the earlier work and repoliticized the police, making them more reliant on the presidency and the ruling party. Repression and politicization reached a peak after the April 2018 popular protests. By the end of 2020, President Ortega had consolidated power, repressed, and jailed most of the political opposition, and destroyed much of the police reforms that came to characterize Nicaragua’s National Police since the revolution. The pattern of repression continues today with renewed attacks on the Catholic Church, the private sector, and NGOs. In early 2023, the Ortega government released from prison and exiled more than 200 political prisoners, arbitrarily canceling their Nicaraguan nationality in clear violation of international law. Our analysis of survey data shows that Ortega’s consolidation of power and politicization of the security apparatus has undermined levels of trust in the police. Data indicate that in the late 1990s and early 2000s, public trust in police was higher in Nicaragua than in other political institutions and higher than the average in the Latin American region. Although there were problems with police corruption and response time, overall the public tended

138

Making Police Reform Matter in Latin America

to register support for police. This changed in 2019, as average levels of trust in police dipped nationally and became more contingent on partisanship. These patterns of public trust in police follow similar trajectories to how policing began to change under Ortega. As we have documented, Ortega implemented a series of legal changes to the police and reoriented how they interacted with the public, particularly during protests. While Ortega originally relied on smaller grupos de choque to control critics of the government, in April 2018 the size and momentum of the protests superseded their ability to control them, and the government decided to change course and send greater numbers of ex-military members and former combatants to confront them, and eventually uniformed police followed (Cuadra 2018). Para-police groups acted on the street with full knowledge of police personnel (CIDH 2018, 22). Police repression against political opponents quickly became the norm in 2018, raising many questions about the sustainability of the best practices implemented in the 1990s and until 2018. Indeed, the Nicaraguan case raises critical questions—even when countries succeed in reforming their police forces, how can they insulate their initial successes from setbacks? How can reformers keep authoritarian impulses from reemerging in the form of repressive police practices? The political nature of these community-oriented reforms is essential for understanding how Nicaraguan policing changed as President Ortega consolidated political power (Romero, Alvarez, and Moncada 2015). In recent years, Nicaraguan policing changed to respond to political developments, rather than adhering to the larger trajectory of institutional police reform after the collapse of the Somoza dynasty in 1979 and the democratization of police reforms in the 1990s (Cuadra 2018, 245). Furthermore, the CPCs, created in 2013, supplanted the crime prevention committees, which in turn joined the Family Cabinets, twisting prevention into direct and constant surveillance. This context allows us to understand how the community-oriented reforms were undermined, setting the stage for the 2018 crackdowns. On the surface, the evolution of policing in Nicaragua appears to share some similarities with those of Chile (Chapter 3) and Colombia (Chapter 4). In all three cases, countries successfully reformed their police forces and embraced at least some community-oriented practices, then quickly abandoned this trajectory to repress government critics. However, there are important distinctions among these three countries. In Nicaragua, police violence against protesters was not rooted solely in police failure to respect human rights but was the culmination of an ongoing authoritarian takeover of policing. In contrast, in Chile and Colombia, police violence against protesters was less reflective of an intentional authoritarian agenda or the decades-long progress on reform. While the creation of a new constitution could lead to reforms that could address police violence against protesters

Nicaragua

139

in Chile, in Nicaragua this is less likely to occur, given the broader authoritarian context of the Ortega regime. Police reforms that emphasize a nonrepressive role for police have no real place in an authoritarian regime. The tragic shift in policing practices in Nicaragua underscores the political and cyclical nature of reform. Our path-dependency framework helps us understand that historically established institutional structures can change after moments of significant political or social tumult, such as a violent revolution or key electoral outcomes. Nicaragua’s National Police has undergone periods of significant positive reform since 1979, however, since Ortega’s return to power, and particularly since the repression of April 2018, the regime’s democratic backsliding has deeply affected the institutional legitimacy of the police. When the community challenged the leadership of President Ortega, the National Police chose sides and now defends the regime through repression. In this sense, the revolution has tragically come full circle, with the police fulfilling a similar role as they did under dictatorship. What was once a promising case of police reform is now a warning sign. When the regime reverts to authoritarian rule, it is highly unlikely that police will stay above the fray. This is particularly troublesome given the retreat of democracy and the rise of populist leaders such as President Ortega, both of which are occurring on a global scale.

Notes 1. Some of the findings in this chapter were reported earlier in Malone and Dammert (2020). 2. To be sure, policing practices alone cannot receive the credit for low rates of violent crime. Orozco (2015) notes that emigration and remittances reduce social tensions and economic deprivation in Nicaragua, which in turn led to lower levels of violence in the country. For an interesting, in-depth discussion of the dynamics of violence in urban Nicaragua, see Rodgers (2016). 3. For a broader overview of the political, economic, and social ramifications of this repression, see Aguilar Antunes, De Gori, and Villacorta (2018). 4. The controversial project, financed mostly by Chinese investment, proposes to build an interoceanic canal using the San Juan River as an access route to Lake Nicaragua. The project generated substantial opposition from environmental groups (Passary 2015). 5. For full interview, see: https://www.economist.com/the-americas/2012/01/28 /a-surprising-safe-haven. 6. Author interview, Managua, Nicaragua, July 2011. 7. Participant observation of police orientation sessions in Managua, July 2013. 8. According to the Red de Seguridad y Defensa de América Latina (RESDAL), the number of police per 1,000 inhabitants is 4.4 in Panama; 4.1 in Costa Rica; 3.3 in El Salvador; 3.3 in Honduras; 3.1 in Nicaragua; and 2.8 in Guatemala (2012–2013 estimates) (RESDAL 2014). 9. Survey results corroborate this point. When LAPOP asked respondents to indicate the greatest threat to security, most respondents indicated common criminals

140

Making Police Reform Matter in Latin America

(42 percent). A sizable portion, 27 percent, identified pandillas as the most important threat, but only 10 percent indicated organized crime. 10. For more information on this survey, see LAPOP’s technical description at https://www.vanderbilt.edu/lapop/nicaragua/NIC1997-Technical-Information-v1.0 -FINAL-eng-200604.pdf. 11. For a concise overview of this pact, see Rogers (2009). 12. To measure ideological orientation, respondents were asked to place themselves on a scale of 1, left, through 10, right. 13. Only 21 percent of the sample reported values of 6–10 on this ideological scale. 14. Although police respondents were more supportive of citizens’ right to protest on average, it is important to note that this difference was only statistically significant in the case of the first item: lawful protests to achieve political goals. On the other two items, police responses were higher than those of the public, but these differences were not statistically significant, most likely due to the smaller sample size of our police respondents (n = 150). 15. Although Honduras did not experience civil war, its geographical location led it to inherit many of the problems of its postconflict neighbors (Malone 2012). 16. Extraordinarily high rates of violent crime in El Salvador and Honduras, and to a lesser extent Guatemala, have been widely attributed to organized criminal groups. Still, the link between high levels of violent crime and the presence of organized criminal groups is complex. Several studies highlight that it may not be merely the presence of criminal groups that drives violence, but whether these groups contest resources among themselves or face reprisals from the state. For example, Bailey and Taylor (2009) point to several scenarios in which organized criminal groups are present but do not resort to violence. The absence of violent crime in Nicaragua does not necessarily mean that organized criminal groups are absent; it could also indicate that organized criminal groups have hegemonic control over resources and/or territory and find it in their best interests to avoid extensive violence. 17. The Canadian Immigration and Refugee Board provided an overview and assessment of the CPC in English, available at https://www.refworld.org/docid/534c e41a4.html (Canada: Immigration and Refugee Board. 2013). 18. Due to data availability, we computed an index of average trust in political institutions including courts, legislature, municipal government, and political parties. 19. Thirty percent of respondents indicated that they had interactions with the police, compared with 14 percent with the courts or justice tribunals, 7 percent with the public prosecution office, and 33 percent with the municipality. 20. The exact text of the survey items read: “Did you report the crime to any institution?” and “Why did you not report the crime? (1) Does not work (2) It is dangerous and afraid of retaliation (3) Did not have any proof (4) It was not that serious (5) Did not know where to report (6) Other reason (8) DK/DR.” The first question asks respondents if they reported the crime to “any institution”; when asked later in the survey about locations for crime reporting, more than 92 percent of all respondents indicated that if they had to report a crime, they would go to the police (vic11 in 2008 survey). 21. Once again, we have included the independent variables that are most important in the Nicaraguan context. Crime victims were coded as (0) not a victim; (1) victim of crime in past twelve months. Figure 6.4 reports the question wording and coding for our measure of fear of crime, and Figure 6.7 reports the question wording for police solicitation of bribes. 22. Measured in age cohorts: (1) 18–25; (2) 26–30; (3) 31–35; (4) 36–40; (5) 41–45; (6) 46–50; (7) 51–55; (8) 56–60; (9) 61–65; (10) 65 and over.

Nicaragua

141

23. The variable measuring the size of respondents’ town or city was coded as (1) rural area, (2) small city, (3) medium city, (4) large city, (5) capital city. 24. The income scale was calculated based on answers to the following survey items: Do you or any member of your household have any of the following possessions? TV; car; refrigerator; telephone; cell phone; computer; microwave oven; washing machine; drinking water; sewage system. Responses were coded as (1) yes and (0) no. We created an index of personal income using a means formula that included a case if there were valid responses to at least eight of the ten items. 25. We report our ordinary least squares regression results for ease of interpretation. We replicated these results using ordinal logistic regression, and the variables in these analyses had the same signs and levels of significance. Results of ordinal logistic regression are available on request. 26. The control variables included the traditional survey research socioeconomic and demographic variables: age, gender, education, income, color of skin, and size of municipality. 27. These questions followed the same format as those measuring trust in other institutions: “How much do you trust the CPC or Councils of Citizen Power?”; “How much do you trust the Gabinetes de la Familia?” Answers to these questions ranged from (1) not at all to (7) a lot.

7 Panama: Rebuilding Security After Invasion

PANAMA IS A FASCINATING YET UNDERSTUDIED CASE OF POLICE reform. Like most of its Central American neighbors, Panama has a long history of military rule and inequality. The police forces were developed shortly after independence as part of a weak constabulary force under the tutelage of the United States, mainly to protect US interests and maintain elite rule. The nature of Panama’s emergence as an independent country, closely linked to the US ambitions for an interoceanic canal, meant that for most of the early twentieth century the police force was relatively small, weak, and dependent on the United States. From 1903 through 1931, Panama relied almost entirely on US forces stationed in the Canal Zone for security. Political instability in the early 1930s led the United States to push for the establishment of a national police. By the 1950s, as in most Central American countries, the National Police had transformed into a militarized National Guard whose main function was to maintain elite power and repress domestic opposition. The National Guard was transformed into the Panamanian Defense Forces (PDF) in the early 1980s as part of strengthening the security apparatus in preparation for transferring the canal to Panama as a result of the Torrijos-Carter Treaties.1 Under General Manuel Noriega, the PDF became more militarized and consolidated political power. The structure of integrated police and military structures with significant political influence under a uniformed officer was not dismantled until the crisis that led to the US military invasion of December 20, 1989.2 The US invasion was the critical juncture that set off significant reform of Panama’s security forces and the building of today’s demilitarized, civilianled Public Force (Fúerza Publica). In the aftermath of the 1989 US invasion, Panama followed Costa Rica’s lead and abolished its standing military (see 143

144

Making Police Reform Matter in Latin America

Chapter 5). The US invasion led to the dismantling of the Panamanian security apparatus, allowing for a clean break from its authoritarian past and the establishment of a new, professional civilian police force. Unfortunately, also like Costa Rica, rates of violent crime surged in the early 2000s, particularly as organized criminal groups and illicit trade proliferated. As in Colombia (see Chapter 4), these sharp increases in violent crime acted as another critical juncture that led Panamanian officials to rethink public security policy, and they consciously choose a very different path. With the help of extensive international assistance and expertise, Panamanian officials crafted and implemented community policing models, aiming to integrate police officers more cohesively into the communities they served. For example, President Martín Torrijos (2004–2009) steered crime control policy away from mano dura tendencies, extending the Mano Amiga (Friendly Hand), which revolved around crime prevention and offering alternatives to dissuade youth from joining maras (gangs) (Ribando Seelke 2005). When some preventive measures did not yield the desired results in short order, officials were willing to turn to more repressive tactics, particularly against gangs. Panamanian reformers have also sought to increase professional standards among police by raising salaries and improving training, and civil society groups have linked these reforms to improved professionalism on the streets and a reduction in bribe solicitation from average citizens.3 Panama still faces many challenges. Like Colombia (Chapter 4), police reforms in Panama unfold very differently in urban and rural settings. The vast majority of reforms targeted urban areas and did not really reach rural parts of the country. In particular, the Servicio Nacional de Fronteras (National Border Service) (SENAFRONT) has come under harsh criticism amid charges of corruption and abuse of human rights in the rural areas of drug trafficking corridors. Also like Colombia, police reforms have advanced against a backdrop of high levels of inequality. While per capita income almost doubled between 2010 and 2020, and poverty rates declined over 40 percent between 2006 and 2016, inequality remained stubbornly persistent. According to the most recent World Bank estimates on inequality, Panama ranks seventeenth in the world, with the third highest levels of inequality in Latin America (behind Brazil and Colombia, respectively).4 Such high inequality can pose challenges for the process of police reform by shaping police responses to different socioeconomic classes. In this chapter, we trace the evolution of policing practices and public security policies in Panama. We identify the critical junctures in which policing practices underwent extensive overhaul and the incentives that officials faced to design and implement community-oriented and holistic public security policies. We also note the challenges that remain, particularly Panama’s role in illicit trade, high levels of inequality, and urban/rural divides.

Panama

145

Building a New Police Force Panama’s current police and security agencies emerged in the wake of a military intervention from an extraterritorial power. On December 20, 1989, the United States invaded Panama to depose and apprehend General Manuel Noriega. The aftermath of the invasion led to the demilitarization and depoliticization of the security forces. The PDF, at the time the country’s armed institution, was dissolved, its command destroyed, and its personnel the subject of a massive restructuring. From the ashes of the PDF emerged a Fúerza Publica of 12,000 men and women.5 Three factors were decisive in transforming the public security structures. First, the invasion destroyed the operational capabilities of the PDF. Second, the looting and ensuing anarchy dramatized the urgent need for an organization whose task could ensure public security. Third, the country required a new security apparatus independent of the party in power. After the invasion, the Minister of Government and Justice, Ricardo Arias Calderón, decided to replace the PDF with the Public Force composed of three services: National Police, National Air Service, and National Maritime Service (RESDAL 2013). The decree that created these entities assigned them responsibility for public order and national defense and envisioned the possibility that special units with external defense functions might be created in the future. These services were almost completely demilitarized in terms of their structure, philosophy, weapons, and training. This was not the case in terms of their personnel, who were largely recruited from the former PDF. The new security forces were subordinated to civilian authorities through direct control by the Ministry of Government and Justice and budgetary oversight by the Office of the Comptroller General and the Legislative Assembly.6 Decree 38 created a Public Security and National Defense Council to coordinate public policy in matters of national security, chaired by the president and composed of key ministers and senior government officials. In February 1990, the government created the Institutional Protection Service to provide security for the president. This unit was placed under the control of the minister of the Presidency, entirely separate from other security forces.7 The old intelligence service, the Departamento Nacional de Investigaciones (Department of National Investigations), was reorganized as the Technical Judicial Police to carry out criminal investigations under the control of the attorney general’s office. Transit, penitentiary, and immigration security services were all brought under the Ministry of Government and Justice. Of the 16,000 members of the former PDF, 12,000 positions were assigned to the National Police, 400 each to the Air and Maritime National Services, 300 to the Institutional Protection Service, and 750 to the Technical Judicial Police. The system of living in barracks ended, replaced with

146

Making Police Reform Matter in Latin America

eight-hour service shifts. As a result, at any given moment only about 3,500 police officers were in service. Many of the exclusive perks enjoyed by the PDF, such as separate health, banking, supermarket, and social club facilities, were abolished. The purpose of these measures was to dismantle the separate style of military life and integrate the police into civilian life. In addition, the Public Force adopted a new uniform. Military equipment and weapons were replaced with those corresponding to police functions, such as 9mm pistols, .38 calibre revolvers, and 12-caliber shotguns, leaving in reserve a restricted number of T65, M16 rifles, and Uzi submachine guns. Official titles and military ranks were modified from “commanderin-chief” to “director general,” from “General Staff” (Estado mayor) to “Directorate,” from “base” to “station.” In 1994, all military titles for ranks above major were changed to “subcommissioner” and “commissioner.” Of the 1,121 officers, ranked from sublieutenant to general in the PDF in 1989, only 608 remained in 1990: 525 in the National Police, 62 in the National Air Service, and 61 in the National Maritime Service. No general, colonel, or lieutenant colonel remained. Finally, numerous properties accumulated by the PDF were transferred through an inter-ministerial committee. From 1990 to 1991, seventeen bases and eight residences of the chiefs of military zones and of the Association of Wives of the Defence Forces were handed over to civilian agencies. All equipment of the Military Engineering Battalion was transferred to the Ministry of Public Works. Twelve military planes, including a Super Puma with a market value of $10 million, were placed at the disposition of the new government. These changes reflected significant advances over the previous national security model. One notable achievement was the clear separation of state security agencies with distinct mandates for law enforcement, air rescue, coastal surveillance, criminal investigation, and presidential security. The second major accomplishment was the subordination of state security agencies to civilian authority through legislative, judicial, and executive oversight and control. The changes to the security forces were in equal part a demilitarization process, a police reform, and an (imposed) transition to democracy in which the political domination of the security forces ended abruptly. In July 1994, President Guillermo Endara proposed a series of constitutional reforms to the National Assembly, including abolishing the army. The reforms placed responsibility for public order and security on the police. In cases of foreign aggression, Panama would organize special police forces to defend the national territory. To reinforce the apolitical nature of the police, the reforms prohibited the Public Force as an institution and its individual members from participating in political protests, making political declarations, or intervening in political party rivalries. Members retained the right to vote in elections. The constitutional reforms were approved on August 23, 1994, and Article 305 abolished the standing military.8

Panama

147

The institutional success of demilitarization is bound to the internal changes in one of the main political parties in Panama, the Partido Revolucionario Democrático (PRD). The military government created the PRD in the late 1970s, designing it to be the political vehicle for leading the transition process after the 1968 coup. Given the party’s long-standing ties to the powerful military establishment, its acceptance of civilian authority was critical. Without at least minimal PRD support of civilian authority, and the accompanying constitutional and institutional changes undertaken since December 1989, it is doubtful demilitarization would have been consolidated. The embrace of demilitarization was the result of structural changes in the party that purged the influence of former members of the PDF and their most ardent supporters, and more important, a change in attitudes toward the role of the military in the state. The latter was perhaps more due to political convenience than real conviction, but nonetheless it was important (Pérez 2018). The PRD stopped short of totally repudiating its military past, picking out the elements that shored up popular appeal while remaining in the norms of the new political order. It accepted the new rules of the game and made no call or attempt to orchestrate disturbances or destabilize the reforms. There was consensus among a critical number of elites to proceed with police reforms.9 Once the constitution was amended, it was possible to institutionalize the other aspects of public security reform, notably the organic law for the National Police. Overall, the National Police’s organic law (Law 18 of June 3, 1997) consolidated the changes that had already taken place since 1990. Key aspects emphasized the civilian nature of the institution—for instance by replacing ranks (rangos) with levels (niveles). The law clearly defined how and when force could be used and eliminated the concept of due obedience. It reiterated the subordination of the police to civilian authority, clearly stating the prohibition on political activities, intervention, and partisanship—all of which were punishable by dismissal. Law 18 enabled the professionalization of the police forces by clearly establishing their mission, functions, and regulation, including a strict code of conduct and a clear career path. The strengthening and professionalization of the security forces, especially the National Police, continued throughout the mid-1990s, with budgets increasing from $77,286,258 in 1995 to $103,929,183 in 1999; and the number of National Police staff increased from 12,000 in 1994 to 16,000 by 2002.

The Growth of the Illicit Sector The transformation of security agencies postinvasion promoted democratic civilian control but did not eliminate deep concerns over the country’s security situation, which had been severely compromised by the growth of the

148

Making Police Reform Matter in Latin America

illicit trade. Panama’s proximity to strategically important trade routes has long made it an ideal route for legal and illegal goods. Even under colonial times, elites built their power base around the control of illicit activities; for example, the slave trade offered numerous opportunities for enrichment and stimulated investments in local real estate and business ventures. Illicit trade continued to thrive throughout Panama’s history, and this backdrop facilitated the country’s emergence as a key actor during the cocaine boom of the 1980s, particularly when law enforcement crackdowns in the Caribbean led traffickers to seek alternate routes to ship drugs north. Panama’s border with Colombia and the isolated Darién Gap region further made it an asset for the cocaine trade. Under Noriega’s rule in the 1980s, observers began referring to Panama as a narco-state, given Noriega’s work with Colombian cartels to smuggle drugs, protect traffickers, guard cocaine-processing plants in Panama, and launder money. Even after the US invasion destroyed the Panamanian army and captured Noriega, problems of corruption and illicit trade proved intractable. Panama continued to play a critical role in money laundering, given its strong international banking center, use of the US dollar, and lax foreign exchange regulations.10 After the invasion, drug trafficking actually increased in Panama, particularly as organized criminal groups took advantage of the power vacuum in the wake of the dismantling of the PDF (Sullivan 1997). Two years after the invasion, the US government noted that drug trafficking in Panama had increased (General Accounting Office 1991). In 1999, the US State Department described Panama as “a major transhipment point for illicit drugs smuggled from Colombia” (US Department of State 1999).11 It is estimated that between 2005 and 2010, the number of gangs operating in Panama tripled (Ministerio de Seguridad Pública 2011). By 2013, signs indicated that several Mexican criminal organizations operated in Panama: the Sinaloa Cartel, the Zetas, the Beltrán Leyva Organization, and the Juarez Cartel (InSight Crime 2018). These organizations were responsible for drug trafficking, kidnappings, extortion, and money laundering. Banking secrecy laws, a dollarized economy, and weak judicial oversight made Panama an ideal base for organizing illicit activities. In the early 2000s, concerns abounded as to the effectiveness of the Panamanian security forces to meet the new challenges posed by international criminal networks, terrorism, and drug trafficking. The United States, concerned over this situation, encouraged Panama to restructure its security apparatus. President Martín Torrijos came to power in 2004 and obliged the United States by pushing the first significant reorganization of security agencies since the early 1990s, including creating the National Intelligence and Security Service and modifying Article 41 of the National Police Organic Law to allow former military officers to be named as chief of the National Police. Torrijos was also able to pass additional reforms to consolidate the Air and Maritime Services into the Aeronaval National Service, create the National

Panama

149

Border Patrol, and reorganize the Council on Public Security and National Defense as a separate entity. Torrijos justified the controversial reforms largely because of the spillover effect of the Colombian civil war. At the time, it was not unusual for FARC guerrillas to operate relatively freely in and out of Panamanian territory in the sparsely populated border area. This institutional restructuring was accompanied by investments in preventive public security policies. To respond to an increase in violent crime in the early 2000s, Torrijos advocated holistic public policies to address the root causes of crime, distancing Panama from the mano dura approaches popular in other parts of Central America. Under the Torrijos administration, Panama launched the Mano Amiga program, which revolved around crime prevention and providing at-risk youth with alternative educational and economic opportunities, with the goal of dissuading them from joining gangs. With the help of domestic and international nongovernmental organizations, Mano Amiga provided approximately 10,000 Panamanian young people with access to extracurricular activities, as well as rehabilitation for former gang members (Ribando Seelke 2005). In 2006, the government developed a Programa de Seguridad Integral (PROSI), its largest prevention program, which targeted four municipalities with the highest crime rates. The Inter-American Development Bank invested $20 million to jumpstart this program for an initial timeline of five years (RESDAL 2013). The program approached prevention from several different angles. For example, PROSI worked with community organizations and leaders to form committees to engage youth in high-risk communities. It collaborated with the National Institute of Women to address cases of domestic violence and with other government ministries to work with youth gang members and minors in prison (RESDAL 2013). PROSI also prioritized strengthening the police and created the Observatory for Violence in 2012.12 Additional measures aimed to improve the collection, analysis, and dissemination of data related to public security, such as the Directorate of the National Integrated System of Criminal Statistics (created in 2007). This organization’s primary duties are to design, regulate, collect, process, and analyze studies of government data related to criminality and public security (RESDAL 2013). When the Ministry of Public Security was created in 2010, the Directorate of National Integrated System of Criminal Statistics fell under its auspices and created its own system to receive, compile, and analyze data from other government agencies.13 In 2009, President Ricardo Martinelli (2009–2014) was elected, and his campaign and time in office were marked by a very different rhetoric. Martinelli touted his mano dura approach of ratcheting up security measures and spending. He drew sharp distinctions between his security policies and that of his political opponent, noting that his opponent “believed in the friendly hand, and I believe in the iron fist” (La Estrella 2008).14 Martinelli’s campaign promises included stricter penalties for adult and juvenile offenders.

150

Making Police Reform Matter in Latin America

Once elected, he deployed the National Police to crackdown on suspected criminals (particularly suspected gang members), as well as protesters (Murillo 2012). Under his administration, Panama’s spending on security increased from slightly under 2 percent of the country’s GDP in 2009 to approximately 2.75 percent in 2014 (RESDAL 2014). In August 2009, Martinelli criticized Torrijos’s reforms and sought to repeal the decree that created the National Intelligence and Security Service. He made additional significant changes to Panama’s security apparatus to more effectively deal with rising crime rates and other security threats, such as drug trafficking. On March 31, 2010, the National Assembly approved a law establishing the Ministry of Public Security. This change came after the Ministry of Government and Justice was basically split in half and turned into two smaller ministries. The Ministry of Security is composed of the National Police, the Servicio Nacional de Fronteras (SENAFRONT), the Servicio Nacional Aereo Naval (SENAN), and the National Immigration Service. The other part of the split, the new Ministry of the Interior, is responsible for the Civil Protection System, the Office for Attention to Refugees, the post office system, the Office of Indian Policy, the office for coordination with local governments, the passport office, the Ground Traffic and Transportation Administration, the Civil Aviation Authority, and prisons. In addition to structural changes, the Martinelli administration established a series of interagency “cabinets” to coordinate activities aimed at tackling crime and its underlying causes. Starting in 2011, these cabinets dispersed across the country, holding sessions in local communities to provide a space for the views of community leaders, local authorities, and ministers (RESDAL 2013). Preventive crime control strategies and community policing initiatives coexisted with Martinelli’s mano dura advocacy. In 2012, the first community police unit, the Unidad de Policia Comunitaria (UPC), was deployed in Curundú, with 180 agents (including 35 women) who received training from the Rio de Janeiro police (Solis 2016). In September 2011, Brazilian firm Odebrecht was working on a private residential complex in Curundú and contacted the Ministry of Public Security over some security concerns. Under the auspices of its corporate responsibility program, Odebrecht offered to install a PCU, which received training from the Rio de Janeiro police (RESDAL 2013).15 The unit distinguished its agents with different uniforms and worked to develop a strong relationship with the local community to build trust. In June 2013, another UPC unit of 388 agents deployed in El Chorrillo, which had an astonishingly high homicide rate of 91.1. According to the police, crime levels in the area declined by 70 percent after the introduction of the UPC (Solis 2016). After his time in office, Martinelli fled the country to avoid criminal charges related to money laundering and the Odebrecht scandal. Upon his

Panama

151

extradition back to Panama, he was placed under house arrest, but in 2021 a Panamanian court declined to convict him. His sons pled guilty in a US court to corruption and money-laundering charges. The conviction of Martinelli’s sons bodes poorly for the former president, as US prosecutors included language in the indictment that the brothers laundered millions of dollars in bribes “on behalf of a close relative who was a high-ranking government official in Panama” from 2009 until 2014 (Ives 2021). After Martinelli’s pivot toward mano dura, reformers aimed to return to earlier strategies based on prevention and community orientation. This return to more holistic public security policies was supported by the fact that Martinelli’s mano dura approach did not curb violent crime. Under his watch, homicide rates reached a historic high of 17.3 per 100,000 in 2013. In interviews with high-ranking police officials and public security experts conducted a month after Martinelli left office, reformers lamented the inability of mano dura policies to curb crime successfully, not just in Panama but in the northern Central American countries of Guatemala, Honduras, and El Salvador. Officials and consultants noted that more punitive, repressive approaches to fighting crime had not deterred homicide rates and organized criminal activity. Those interviewed found the evidence supporting more community-oriented and prevention-based models of public security, such as those pioneered in Chile, to be more persuasive. As one police official noted: We see what is happening in Honduras and El Salvador, and we feel that if we don’t act now, that will be us in ten years. The gangs, the murder rates, the drug traffickers—all of it will be worse. We have the money from the canal to invest in education and youth programs, and if we don’t do it now, it is going to be too late. If we let the gangs grow as big as they are in El Salvador, we’ll never be able to get crime and violence under control.16

Another public security official highlighted the promising collaboration between experts in public security and those in education, housing, and health. He pointed to the need to improve educational retention and recreational opportunities for youth, and discussed his collaboration with university teacher training programs: The problem sometimes is that our teachers tend to come from nice, wellmannered, middle-class families, and they don’t have much experience with kids who are raised in poverty, sometimes living without adequate food and in informal housing settlements. They get these kids in the classroom, and the kids don’t behave, so they say “I can’t teach them,” they want to have the kids suspended or expelled from school because they don’t behave, they disrupt the classroom. They throw up their hands. But then the kids are out of school, and into trouble. There’s a whole bunch of bad characters waiting to get these kids involved in one illegal scheme or another. So we have to socialize our teachers and train them to teach all different kinds of kids, not just the ones who know how to behave.

152

Making Police Reform Matter in Latin America

The official explained how he had reached out to professors of education at a university to talk about how they could work with teachers to train them to be effective in classrooms with children from poorer backgrounds to improve educational retention rates. He repeated the phrase, “Afuera, no. Adentro” (Outside, no. Inside). That is, youth needed to remain inside the classroom, not suspended or expelled for poor behavior or lackluster educational achievements. He worked with education professors to take their students to informal housing communities and provide tutoring and educational support. Panamanian officials noted that international donors expressed a decided preference for community-oriented policing practices in urban areas.17 Indeed, one high-ranking police chief pointed to a large stack of papers and said they were all invitations from community policing experts and consultants to learn more about police reforms and how community policing programs worked in other countries in the region. The official said, “I’ve gotten the message. They all want community policing.” Despite the slight exasperation with the amount of correspondence, the official did note that community-oriented programs appeared promising, particularly since mano dura did not appear to work in Panama or other parts of Central America. The official also noted that investments in prevention appeared to yield dividends, particularly when they offered alternatives for at risk youth to dissuade them from joining maras. As previously mentioned, Panama provided some Panamanian youth with access to extracurricular activities such as theater and sports, and rehabilitation for former gang members. But when those community-oriented or preventive measures did not yield results quickly, officials pivoted to more repressive policies. For example, President Juan Carlos Varela (2014–2019) launched the Safe Neighborhoods with more Opportunities and a Firm Hand program in 2014, which included a month-long amnesty period for gang members to voluntarily surrender their weapons (Córdova 2014). In the first six months after implementation, approximately two thousand former gang members joined the program, and two hundred weapons were turned in. The government concentrated on regions with the highest levels of gang activity and worked with civil groups to implement the program, setting up five commissions to collect the weapons and work with the gang members to provide social services, education, and economic opportunities to reincorporate them into society (Cook 2014). Despite early successes, the amnesty and job training program had mixed results. While homicide rates declined, other types of crime did not, and gang membership continued to plague certain high-violence and economically depressed areas. As a result, Varela moved toward a mano dura approach by introducing a threehundred-strong special forces unit aimed at taking down gangs in Panama City and Colón. The special forces unit were drawn primarily from the Institutional Protection Service, the presidential security force (Gobierno de la República de Panamá, March 24, 2017).

Panama

153

Panama continued to be a major transit state for the illicit drug trade, and as Colombian cocaine production boomed in the late 2010s, Panama interdicted record amounts of drugs. Official figures released by the ministry in October showed that in 2016, fifty tons of drugs were confiscated and over 1,500 people arrested in antidrug operations by the National Police, SENAFRONT, and SENAN (Pérez 2017b ). According to the Ministry of Security, the National Police confiscated 26.5 tons of illegal drugs, with SENAN and SENAFRONT confiscating 21.5 and 2.3 tons, respectively (El Panamá América 2016). In July 2016, Varela announced the creation of a new special antinarcotics force (Fuerza Especial Antinarcóticos) (La Prensa 2016). This followed the release of a report by the United Nations Office on Drugs and Crime (UNODC) on Colombia, which confirmed that there was a significant increase in coca plantation areas and potential cocaine production from 2014 to 2015 (UNODC 2016c). The new force followed a series of high-level operations in 2015 by specialized joint task forces of the National Police, SENAFRONT, SENAN, and the Institutional Protection Service. While the government focused on extending security measures, corruption continued to plague the police, justice, and prison systems. Panama’s Interior Minister, Milton Henríquez, announced the dismantling of a corruption ring in the country’s prisons, resulting in thirteen arrests, including four officials and three former officials from the prison service (Pérez 2017b). An Interior Ministry press release stated that the operation was carried out by the Public Ministry (attorney general), the National Police, the general inspectorate of prisons, and the directorate of judicial investigations. According to the press release, the corruption ring, which had been operating for years, had altered judicial sentences and manipulated the transfer of inmates from one prison to another, among other things (Mendoza 2016). Efforts to reform the justice system received a boost in late 2016, after years of delays and numerous setbacks, when Panama announced a new legal system based on the adversarial approach (where two advocates represent their parties’ positions before a jury or judge), ending a process that began in 2011. The new system replaced the inquisitorial system, where the courts are actively involved in investigating the facts of the case. The UNODC, which helped the transition process, highlighted the expectation that the new system “will allow for a speedier and more transparent resolution of conflicts between parties” (UNODC 2016a).18 Unfortunately, corruption and inefficiency continue to plague Panama’s judicial system. In 2017, the World Economic Forum ranked Panama 120th out of 137 countries for judicial independence (Schwab and Sala-i-Martin 2017; InSight Crime 2018). In a move that baffled international observers, and many Panamanians concerned with the rise in violent crime, newly elected President Laurentino Cortizo (2019–) and his security minister announced in January 2020 that they were allowing a law that limited the number of firearms per import shipment to expire.19 Panama’s geographic position and open economy has

154

Making Police Reform Matter in Latin America

long made it a hub for arms, human, and drug trafficking, and lifting restrictions on importing firearms could affect the illicit transit of weapons across Central America and the United States. Firearms could also become more widely available for domestic gangs and criminal networks. In sum, despite persistent problems with corruption in the justice system more broadly, and through higher levels of government, many observers credit the police reforms since the early 2010s with improving professionalism and reducing the prevalence of petty corruption among the police on the streets. Civil society groups noted that police treatment of citizens had improved steadily, despite isolated incidents of excessive use of force (such as that deployed by the Martinelli administration). Still, these civil society leaders cautioned that while the community-oriented public security practices had taken root in urban areas, in rural zones (particularly in areas with high rates of illicit trade), policing practices were increasingly militarized, and human rights abuses continue to be prevalent.20

Crime Victimization and Perception of Insecurity These changes to institutional configurations and public security policy aimed to address changes in rates of violent crime. Violent crimes such as homicides did rise in the mid-1990s, following the US invasion and subsequent restructuring of the security sector. Despite some minor fluctuations, these rates stabilized between 1996 and 2004. As Figure 7.1 illustrates, homicides began to increase significantly in 2006, more than doubling by 2009. Homicide rates dipped in 2010 but then peaked to an historical high of 17.3 per 100,000 inhabitants in 2013. Following this peak, rates declined back to their late 1990s levels, but then started to creep upward again in 2019 to 11.2 per 100,000 people. Rates of victimization from other types of crime rose steadily between 2012 and 2018. As reported in LAPOP’s 2004–2019 AmericasBarometer surveys, the percentage of Panamanians who said they were victims of a crime during the past twelve months dropped by half between 2004 and 2006: from 14.8 percent to 7.1 percent. This mirrors the drop in homicide rates during this period. However, this reprieve was short-lived. Victimization rates rose in 2008 and again in 2010; despite a drop in 2012, rates climbed steadily between 2012 and 2018. Between 2014 and 2016, self-reported victimization rates doubled to 16 percent, and increased again to 21.5 percent in 2018. This 2018 rate was the highest percentage of self-reported crime victims since LAPOP began conducting its surveys. Notably, we found that education and sex are the two statistically significant factors influencing crime victimization in Panama, as men with higher levels of formal schooling are more likely to say they have been victims of a

Panama Figure 7.1

155

Homicide Rates per 100,000 People in Panama (1995–2020)

Source: World Bank, World Development Indicators.

crime.21 According to our analysis, men were 7.8 percent more likely to say they are victims of crime than women, and crime victimization increased as the level of formal schooling increased. These results mirror studies by the World Bank that found that men between the ages of fifteen and thirty-four make up about 60 percent of all homicide victims (World Bank 2011). These increases in self-reported victimization rates between 2012 and 2018 are matched by increases in public perceptions of insecurity. Figure 7.2 shows that levels of insecurity improved steadily from 2004 to 2012. Since then, fear of crime has increased significantly. It seems that after comparing trends in homicide rates and self-reported victimization, perceptions of public insecurity in Panama follow actual experiences of victimization more closely, as opposed to the more unusual crime of homicide. In each survey wave, LAPOP asked respondents to name the most important problem facing their country. In 2004, only 10.8 percent of respondents considered crime to be one of the most important problems at the national level; public fear of crime was moderately high, but public preoccupation with other national issues, such as economic matters, overshadowed crime. This changed dramatically, as the percentage of respondents who

156

Making Police Reform Matter in Latin America

Figure 7.2

Fear of Crime in Panama (2004–2019)

Source: LAPOP’s 2004–2018/19 AmericasBarometer.

thought crime was the most important problem doubled between 2004 and 2006, and increased again in 2008 and 2010. In 2010 public identification of crime as the most important national problem reached its peak, with more than half of respondents ranking this issue first. This percentage decreased somewhat in subsequent years, but from 2012 to 2018, at least one quarter of respondents identified crime as the most important national problem. The general increase in perception of crime as important might reflect the increase in homicide rates observed in Figure 7.1.

Trust in the National Police We turn to examine public trust in the police amid these trends of reform and public insecurity. As Figure 7.3 illustrates, from 2004 to 2019, the public reported more trust in police than in other political institutions. Even when this trust dipped in 2006 and 2008, average trust in police was still higher than it was for other key political institutions. In a similar vein, when trust in other political institutions declined sharply between 2004 and

Panama Figure 7.3

157

Trust in National Police in Panama (2004–2019)

Source: LAPOP’s 2004–2018/19 AmericasBarometer.

2019, trust in the police remained stable and significantly higher than the regional average. The results reflect the success of the new police agencies to establish relatively high levels of institutional legitimacy despite the concerns over crime. Police corruption is one of the most serious problems facing police institutions. Among the most insidious is the practice of asking those with whom the police interact for an extralegal payment (a bribe). Figure 7.4 shows the percentage of Panamanians who say a police officer asked them to pay a bribe. The graph shows that after a slight decline in 2008–2012, the percentage increased significantly in 2014 to more than 11 percent of respondents. The percentage declined in 2017 but remains at a higher level than the earliest LAPOP surveys. Although several reforms aimed to improve the professionalism of police through higher pay and well-defined career paths, the problem of bribe solicitation has clearly been persistent. In the most recent 2019 survey, Panama’s level of bribe solicitation was significantly below the regional average but had increased by 28 percent. This increase may appear small, but given reform efforts to improve police professionalism, it is troubling that self-reported bribe solicitation increased rather than decreased.

158

Making Police Reform Matter in Latin America

Figure 7.4

Police Bribery in Panama (2004–2018)

Source: LAPOP’s 2004–2018/19 AmericasBarometer.

Table 7.1 reports the results of a regression analysis of the determinants of trust in the National Police using the 2018 survey results. The regression uses the standard sociodemographic variables and measures of crime victimization, fear of crime, and whether the respondent says a police officer asked them for a bribe. These results indicate that corruption, fear of crime, and crime victimization all have a significant and negative effect on trust in police. Bribe solicitation had the largest effect, as respondents who reported that police had solicited a bribe scored more than one full point lower on trust in police than those who had not, holding all other independent variables constant. Corruption by individual police officers clearly undermines levels of institutional legitimacy. Of the socioeconomic and demographic variables, wealth and urbanization shaped levels of trust in police. Wealthier respondents registered less trust in police, as did those living in urban areas. These latter two results are perhaps counterintuitive, as one might think that wealthier respondents might be accustomed to better police treatment, and police reforms tended to focus on cities rather than urban areas. Our analysis indicates that this is not the case, however. Our findings on urbanization could be attributed to the fact that cities usually have more

Panama Table 7.1

159

Determinants of Trust in the National Police in Panama (2018)

Independent Variables

Trust in Police

Police bribe solicitation

−1.058*** (0.171) −0.296* (0.117) −0.180** (0.052) −0.021 (0.014) 0.006 (0.016) −0.152 (0.096) −0.586* (0.250) −0.733*** (0.111) −0.039 (0.028) 1,501 0.104

Crime victim Fear of crime Education Age cohort Sex Wealth Urban Skin color N Adjusted R2

Source: AmericasBarometer, 2018/19, LAPOP. Notes: Unstandardized coefficients; standard errors in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001.

criminal and gang activity than rural areas, and police are often more overburdened in urban areas. In Panama, results show that 24.1 percent of those living in the capital say they were victims of crime in 2018, compared with a rate of 20 percent in rural areas. In contrast, the remaining socioeconomic and demographic factors were not significant, indicating that trust in police did not vary according to educational levels, age, sex, or skin color. In the 2006, 2010, and 2012 surveys, LAPOP posed a question to measure how safe citizens felt when they have contact with the police. Table 7.2 shows the percentages of citizens who felt safe around the police in each year. Citizens who felt secure near a police officer increased significantly from 2006 to 2010 and 2012. In 2006, 56.7 percent of Panamanians felt secure when around a member of the police. That number increased to 81.5 percent in 2010 and 82.2 percent in 2012. These results reflect a growing level of trust and support for the police in Panama. These results are bolstered when we analyze the level of familiarity with police. LAPOP asked respondents whether they knew a police officer by name, face, or both and the extent to which they perceived the police they knew as reliable.

160

Making Police Reform Matter in Latin America

Table 7.2 Perception of Security When Near a Police Officer in Panama Response Very secure Somewhat secure Somewhat insecure Very insecure

2006 (%)

2010 (%)

2012 (%)

11.3 45.4 25.7 16.6

17.4 64.1 14.7 3.8

25.9 56.3 11.2 6.6

Note: The exact text of the survey question read, “When you are near a member of the National Police, how safe to you feel? (1) very safe; (2) a little safe; (3) a little unsafe; (4) very unsafe; (5) neither safe nor unsafe.”

Table 7.3 shows that approximately half of the respondents in the LAPOP surveys said they knew a police officer with 48.5 percent, 55 percent, and 42 percent in 2006, 2010, and 2012, respectively. Most respondents perceive the police officer they know as reliable. In 2006, 68 percent of Panamanians who said they knew a police officer perceived them to be reliable (28.2 percent very reliable and 39.8 percent somewhat reliable). This number was 82 percent in 2010 and 86.9 percent in 2012. Perceptions of police criminality have steadily improved. In 2008, 2010, and 2012, LAPOP asked respondents to indicate if they thought that police were involved in crime, protected local communities, or did a combination of both. In 2008, 35.5 percent of respondents thought that police were involved in crime, but this percentage was cut in half by 2012, when only 17.6 percent had this pessimistic perspective. Next to Chile, this was the lowest percentage in Latin America.22 These results indicate that citizen perceptions of police steadily improved from 2008 to 2012. Other indicators provide additional evidence that citizen evaluations of police tended to be positive. In 2014, when asked about police performance specifically in their neighborhoods, 63.8 percent of respondents indicated that they were either satisfied or very satisfied. When asked about the ability of the justice system as a whole to punish criminals, average responses were above the midpoint of the scale. However, in the two years when LAPOP asked respondents about the ability of police to apprehend criminals, average responses were slightly lower than those of the justice system as a whole. These tepid responses improved between 2004 and 2014, and in 2014 respondents were relatively optimistic that the justice system could sanction criminals. This optimism was short-lived, as evaluations on this indicator declined significantly between 2014 and 2019. Perceptions that police would respond in a timely fashion also declined over time. In 2014, respondents indicated that it would take police between thirty and sixty minutes to respond to a call regarding a burglary. This was significantly longer than estimated police response times in the United States,

Panama Table 7.3

161

Level of Familiarity with and Trust of Police Known by Respondents in Panama

Do you know a police officer? Yes, by face Yes, by name Yes, both No How trustworthy is the police   officer you know?a Very reliable Somewhat reliable Little reliable Not reliable

2006 (%)

2010 (%)

2012 (%)

15.8 7.7 25.0 50.4

13.9 10.6 30.5 44.9

12.3 8.1 21.6 58.1

28.2 39.8 17.3 14.7

31.2 50.8 14.6 3.4

39.1 47.8 7.5 5.7

Note: The exact text of the survey item read, “Do you know a police officer by face or name in your neighborhood or community? (1) yes, by face; (2) yes, by name; (3) yes, by face and name” (PANAOJ8). “How trustworthy do you think the police you know personally are? (1) very trustworthy; (2) a little trustworthy; (3) very little trustworthy; (4) not at all trustworthy” (PANAOJ9). a. Asked only of those who answered “Yes” on PANAOJ9.

Canada, and Uruguay, but it was short for the region, in a statistical tie with estimates in Chile and Costa Rica. By 2017, this estimated response time increased significantly to over an hour and remained longer in 2019. All together, these survey results provide clear indications that the reforms instituted after the invasion have helped to establish a National Police force that was generally trusted by and reliable to most Panamanians. Although crime victimization remained a clear concern and corruption remains a problem, the Panamanian police have come a long way from when it was subordinate to the armed forces, politicized, and highly corrupt. Trust in police, evaluations of police performance in local communities, and perceptions of police criminality all indicate that the Panamanian police have earned some legitimacy in the eyes of citizens. While these indicators are positive, respondents still register doubts about the police’s ability to respond to citizen calls and sanction criminals. Concerns about the ability to respond to crime persist.

Conclusion This chapter has explored the transformation of the police in Panama. Unlike the other countries examined in this book, Panama’s police force emerged from the consequences of political turmoil and an invasion that

162

Making Police Reform Matter in Latin America

destroyed the nation’s armed forces. The current security system is a byproduct of dual processes of demilitarization and democratization resulting from the US military invasion that toppled Noriega’s regime in December 1989. The invasion was a critical juncture that provided the incentives for deep, significant reforms of the country’s security forces. Before the invasion, the police were subservient to the armed forces and ultimately responded to Noriega. The immediate result of the invasion was to destroy the capabilities and command structure of Panama’s armed forces, thus destroying the operational capabilities of the local police. The vacuum of authority after the invasion led to weeks of looting and increased criminal activities, forcing the newly installed Panamanian government, with the direct assistance of the US military, to restore order by relying on former military personnel. Ultimately, the decision was made by the Panamanian authorities that Costa Rica would serve as a model for the new Public Force. It became an institution subordinate to the Ministry of the Interior, led by civilians without military ranks and divided into three branches: the National Police, National Air Service and National Maritime Service. Subsequent reforms, precipitated by changes in crime levels and legitimacy of the early security structures, unified the air and maritime services and created a unit specific to border security. Decoupling military and police functions is a key element in demilitarization and democratization throughout Central America. Costa Rica and Nicaragua illustrate how these complementary processes helped in developing local police forces dedicated exclusively to domestic security and subordinate to civilian leadership. However, Nicaragua also illustrates how these initial gains can be subject to reversal. In Panama, the success of the police force demilitarization was as much political as operational. The ability of political parties, even those that supported the military regime, to accept the dismantling of the PDF and play by democratic rules led to a consensus on the fundamental institutional structures of the security agencies. This chapter used survey data to explore popular attitudes toward crime victimization and the national police. The evidence suggests that the percentage of Panamanians who say they have been a victim of crime in the past twelve months and those who fear being victims of a crime in their neighborhood has increased significantly since 2012. Our analysis shows that crime victimization is determined by sex and education, with men and those with greater level of formal schooling more likely to indicate being a victim of crime. Despite the increase in crime and insecurity, our findings show that trust in the National Police has remained relatively high compared with other state institutions, although citizens do penalize the police when they perceive them as ineffective at controlling crime, as our regression results suggest (Table 7.2). Still, trust has remained relatively high even though corruption and victimization by the police has increased. The per-

Panama

163

centage of Panamanians who say they were asked for a bribe by a police officer more than doubled from 3.5 percent in 2012 to 8.8 percent in 2018. Similar to Colombia (Chapter 4), the sharp increase in homicides in the mid-2000s created another critical juncture. At first, the growth of crime and gang activity led the government to seek additional powers and restructure police units. For some in Panama these changes threatened the demilitarization process. President Martinelli’s increases in budget and personnel and changes to institutional structure increased concerns that Panama’s police agencies were being remilitarized. Since 2014, the government has sought to balance effectiveness and operational capacity with less repressive policies aimed at reincorporating gang members into society and implementing community policing in some high-crime areas, similar to Nicaragua (Chapter 6). Although some gang members have been reincorporated and homicide rates have declined, crime victimization and perception of insecurity have increased. Indeed, when less repressive policies have not yielded the desired results quickly, policymakers have been willing to rely on more punitive approaches. All signs indicate that Panama will continue to struggle to curb the power of illicit actors and the negative consequences of its role in the transit of illicit goods. Democratic reforms transformed Panamanian institutions throughout the 1990s, but the illicit sector has proven resilient. Its power is largely shaped by international forces, particularly the demand for illegal goods in northern markets, and the supply of such goods from organized criminal groups in Colombia and Mexico. Still, there are domestic reforms that Panama can pursue that would support police reforms and render them more sustainable. Continued efforts to improve the transparency and efficacy of the justice system, stricter regulations to identify and sanction money-laundering activities, and investments in social welfare policies all have the potential to enhance the sustainability of police reforms.

Notes 1. The Torrijos-Crater Treaties signed in 1977 established a process by which the Panama Canal would be transferred to Panamanian control by December 31, 1999. 2. For a full analysis of Panama’s early military and police structures, see Pérez (2011a). 3. Author interviews with leaders of civil society organizations, Panama City, 2014. 4. In 2020, per capita income in Panama reached a high of US$25,940. In 2006, 38.3 percent of the population lived beneath the national poverty line; by 2016 (the most recent year data available) this percentage had declined to 22.1 percent. In 2019, the Gini index for Panama was 49.8, following Brazil (53.4) and Colombia (51.3) (World Bank, World Development Indicators, n.d.).

164

Making Police Reform Matter in Latin America

5. An expanded discussion of this overview of the Panamanian democratic transition is available in Pérez (2011a). 6. “Decreto de Gabinete No. 38 de 10 de febrero de 1990,” Gaceta Oficial, no. 21 (February 20, 1990): 479. 7. “Decreto de Gabinete No. 42 de 17 de febrero de 1990,” Gaceta Oficial, no. 21 (March 1, 1990): 485. 8. Article 305 prohibits a standing army but allows for temporary special defense units in times of urgency. The exact text of the constitution is available through Georgetown University’s Political Database of the Americas at http://pdba .georgetown.edu/Constitutions/Panama/panama1994.html. 9. For a fuller analysis of the role of the PRD in the demilitarization and democratization of Panama after the US invasion, see Pérez (2011a) and Loxton and Mainwaring (2018). 10. The 2016 Panama Papers scandal showcased Panama’s central role in international money-laundering operations. 11. See Panama section, https://2009-2017.state.gov/j/inl/rls/nrcrpt/1999/920.htm. 12. See Oficina de Seguridad Integral, https://www.seguridadciudadana.gob.pa /sobre-la-entidad/. 13. Ministry of Public Security Law (no. 5, April 14, 2010). See Ministry of Public Security, Memoria 2011, https://www.minseg.gob.pa/. 14. For more detail, see: https://www.laestrella.com.pa/nacional/politica/081102 /cree-mano-dura-ricardo-martinelli. 15. Odebrecht later became infamous as the epicenter for the largest corruption scandal in the region, as evidence emerged in 2014 that the firm had paid hundreds of millions of dollars to secure construction contracts across Latin America, implicating a wide range of high-level officials. In 2021, the sons of former Panamanian President Ricardo Martinelli pled guilty to criminal charges related to money laundering as part of the Odebrecht investigations (Asmann 2021). 16. Author interview with police officials, Panama City, August 2014. 17. Author interviews with education officials, Panama City, August 2014. 18. See, https://www.unodc.org/unodc/en/frontpage/2016/June/unodc-supporting -criminal-procedure-reform-in-panama.html?ref=fs2. 19. EFE, “El Gobierno de Panamá reabre la importación de armas,” El Diario, January 22, 2020, https://www.eldiario.es/politica/gobierno-panama-reabre-importacion -armas_1_1069690.html. 20. Author interviews with leaders of civil society organizations, Panama City, August 2014. 21. We relied on binary logistic regression to examine the factors that identify crime victims using the 2018 survey. The analysis uses answers to whether the respondent was a victim of a crime in the past twelve months as dependent variable and the usual sociodemographic factors as independent variables. 22. In 2012, 15.4 percent of Chileans stated that the Carabineros were involved in criminal activity.

8 Peru: The Challenges of Institutional Instability

PERU’S PROCESS OF POLICE REFORM HAS REGISTERED FEW SUCcesses, but we can learn several things from these reforms, many of which were abandoned before they could have a real impact. The Peruvian case illustrates how challenging it can be for the rhetoric of reform—even when such rhetoric is sincere—to ultimately materialize into sustainable practices on the ground. Typically, when people think of public security in Peru, they imagine the violence associated with insurgency, most notably that of the Sendero Luminoso (Shining Path), which launched a series of terrorist attacks in the 1980s and 1990s, countered by government crackdowns on suspected terrorists and civilian populations. These government crackdowns of the 1980s and 1990s, particularly under President Alberto Fujimori (1990–2000), destroyed much of the capacity of this insurgency. While Peru’s long legacy of social exclusion and economic inequality has facilitated the emergence of political violence, by the end of the twentieth century, violence perpetrated by insurgents and state actors declined. Peru has registered impressive economic growth in recent years but has struggled with historically high levels of inequality and social exclusion. There are large gaps between the rich and poor, and the capital city of Lima is marked by stark socio-territorial segregation, leading observers to refer to it as “two cities”: one for the rich and one for the poor. Given these longstanding trends, it is intriguing that Peru registers such comparatively low levels of violent crime. By the beginning of the twenty-first century, violence in Peru was substantially lower than that of its South American neighbors and less than a quarter that of the Latin American regional average. Also intriguing (but understudied) are state and societal reactions to crime. When fighting the terrorist insurgency of the Sendero Luminoso, the Peruvian 165

166

Making Police Reform Matter in Latin America

state unleashed excessive force; according to the 2003 Truth and Reconciliation Commission, state forces committed one third of the human rights violations during the conflict. Although much attention has been given to the state’s use of force against the insurgency, state responses to current security challenges posed by common crime have not received as much scrutiny. Even so, preliminary evidence suggests that Peru deviates in important ways from regional trends and warrants closer study. To study the Peruvian experience in greater depth, this chapter examines institutional development, state policy, and crime trends using field research, interviews with police captains, LAPOP’s survey data from 2006–2019, and Peruvian surveys of urban environments. We examine attempts to reform the police in Peru and identify the obstacles that kept these reforms from sustained implementation. In contrast to Chile (Chapter 3) and Uruguay (Chapter 9), the institutional weakness and instability of the Peruvian political system disrupted reforms before they could take root, even when these reforms were sincere and well intentioned. Attempts at reform occurred after government counterinsurgency campaigns, which critics argued were prone to violate the rights of average citizens caught in the crossfire between insurgents and government forces.

Police Reforms amid Institutional Instability Instability has been a hallmark of Peruvian politics ever since former President Fujimori fled to Japan in 2000 and attempted to resign via fax to evade trial for corruption and human rights abuses. Of Fujimori’s five successors, four have faced corruption charges related to the Odebrecht scandal.1 President Alejandro Toledo (2001–2006) was arrested in the United States in 2007 on an extradition order, and President Alan García (2006–2011) committed suicide in 2019 to avoid a corruption trial. President Ollanta Humala (2011–2016), who led a failed coup against Fujimori in 2000, was arrested on corruption charges in 2017 and was awaiting trial in 2021. President Pedro Pablo Kuczynski served only two years in office (2016–2018) before being ousted by an opposition led by the former president’s daughter, Keiko Fujimori, herself a perennial presidential candidate, who was arrested on charges of corruption related to Odebrecht in 2018. Kuczynski resigned while facing an impeachment vote in Congress, and his vice president, Martín Vizcarra, assumed office (2018–2020). When faced with his own opposition, Vizcarra dissolved the legislature, which retaliated by attempting to suspend his presidency and elevate his vice president, Mercedes Aráoz, as interim president. Aráoz’s resignation the next day intensified the leadership crisis, and Vizcarra capitalized on the moment by calling for legislative snap elections in January 2020. While critics charged that Vizcarra

Peru

167

had plunged the country into constitutional crisis, new elections did take place, and voters registered high levels of support for Vizcarra and very low ones for Congress until his impeachment and removal by Congress in November 2020. Vizcarra’s replacement lasted in office for only five days, as widespread protests led to his resignation and replacement by President Francisco Sagasti Hochhausler. During these demonstrations, police unleashed excessive force against protesters, resulting in the deaths of two people, forty disappearances, hundreds of injuries, and reports of arbitrary detention (Amnesty International 2020). Sagasti completed Kuczynski’s original 2016–2021 presidential term and tried to address the police’s excessive use of force by removing the commander general of the National Police, along with eighteen generals. Not surprisingly, this political instability has filtered through the branches of government, including the police, and hindered public policy responses to growing insecurity. Since the early 2010s, criminality has been one of the principal problems preoccupying citizens in Peru (Dammert 2022). Indeed, Peru illustrates a large disjuncture between the incidence of violent crime and victimization, on one hand, and perceptions of public insecurity on the other. Perhaps due to high levels of social and economic exclusion, public fear of crime remains quite high even as violent crime rates are comparatively low. Citizens have pressed their government officials for more and better police officers on patrol, more police engagement in outreach services, and better mechanisms to curb corrupt practices. Politicians have responded to this growing concern by increasing public expenditures in this area, with annual allocations more than doubling between 2011 and 2019. Public security ranked high among the priorities of the budget proposed for 2018, as Congress appropriated 6 percent of that year’s budget for the public order and security sector (Congreso de la República 2017). This allocation remained high in 2019, when it was 5.9 percent of the budget (Congreso de la República 2018) and rose slightly in 2020 to 6.2 percent of the national budget (Alva 2020). This increase in investment has not translated into concrete public security improvements because of chronic institutional instability. The Policía Nacional del Perú (PNP) resides in the institutional space of the Ministry of the Interior, which has undergone extensive overhaul since the early 2000s. However, these changes have not yielded successful long-term results, much less modernization of police processes and consolidation of civilian leadership. In fact, public security policy suffers from two permanent tensions: a constant crisis with changes in leadership and a struggle between civilian and police leadership. As a result, the leadership of the Ministry of the Interior has become one of the most volatile and controversial positions in the country. The Ministry of the Interior hosted eight ministers under Fujimori, one under Valentín Paniagua’s period of transition (2000–2001), and seven

168

Making Police Reform Matter in Latin America

each under of the terms of Toledo, García, and Humala. Two more ministers marked the brief government of Kuczynski, and Vizcarra went through three more. In other words, in a twenty-eight-year period, Peru has sworn in thirty-five Interior Ministers, which undoubtedly undermines reforms and poses permanent setbacks. In addition, half of these thirty-five ministers were civilians, and the other half were retired military or police officers, which has further deepened tensions in the attempts to promote civilian leadership, an area of reform dating back to the 1990s. Perhaps the two most clear examples of this tension are the ministries of Gino Costa (2002–2003) under Toledo and Carlos Basombrio (2016–2017) under Kuczynski, who presented a civilian proposal to reform the Ministry of the Interior and the police. Most of their efforts centered around anticorruption, institutional strengthening of civilian capabilities, and limiting the discretionary powers of the police. Limited political will and visible confrontations with police personnel in the media undermined these efforts. Almost all of their specific initiatives did not survive the change of ministers, although presidential rhetoric on public security and crime control remained consistent. Not surprisingly, in this context, many of the public policies and reform processes developed in recent years have remained only on paper and not converted into concrete changes in practice. As with most reform processes, proposals for police reform first undergo the slow process of discussion, design, and deliberation. Given the short tenure of each minister, it was typical for proposed reforms to never proceed from deliberation to actual implementation, as by the time the reforms reached the implementation stage, technical teams and priorities had already changed. The example of Barrio Seguro (Safe Neighborhood) clearly illustrates this problem. First designed in 2016, Barrio Seguro aimed to increase police presence and crime prevention in high-violence neighborhoods, with implementation scheduled to begin in 2017. However, less than eighteen months after its original design, the program lost priority due to institutional instability and personnel change. By the end of 2018, its financing and development was dissipated, and implementation faded from possibility. The government still touts Barrio Seguro as one of its priorities, but this narrative does not match the reality of the program, which floundered before it could be properly implemented. Despite this gap between rhetoric and reality, the Peruvian government secured a loan from the Inter-American Development Bank (IDB), with the main objective to consolidate the Barrio Seguro initiative (IDB 2019). The first barrier to implementing sustaining police reforms has been institutional instability. The second obstacle has been the bureaucratic structure of the Ministry of the Interior. There has been a long-standing tension regarding the role of police and civilian leadership in the institution. This police/civilian dichotomy has generated an inefficient bureaucratic structure that tends to duplicate the institutional structure of both civilian

Peru

169

and police sectors. The two towers that make up the Peruvian Ministry of the Interior building could be a metaphor for this permanent duality, which unfortunately translates into less collaboration and trust among the officials. There is a general understanding that the minister will work with the civilian sector and take the lead in designing policy and that the police sector should oversee implementation. In practice this line is blurred, and in many cases redundant institutional structures lead to a disconnect between civilian and police sectors, with each pursuing policy goals independently. For example, the General Directorate of Intelligence of the Ministry and the Directorate of Intelligence of the National Police often share objectives, but there is limited coordination and collaboration capabilities across the two entities. In a similar vein, the General Directorate of Citizen Security at the Ministry and the Directorate of Citizen Security of the national police have virtually identical missions, but communication between them is limited, which diminishes their abilities to engage in a key area for crime control and prevention. As of March 2020, there were approximately 134,000 police staff members. Roughly 10,000 of these operate at the level of administration and management; the remainder are below the officer level and oversee street patrol, traffic control, and most community relations. The PNP carries out various tasks; among them, the fight against drug trafficking and the control of citizen protests stand out. The PNP has recently recognized the importance of crime prevention and that prevention requires measures that transcend traditional police patrol. While there is recognition of the need for policy initiatives that cross sectors and departments, this recognition has not advanced beyond the stage of pilot programs, such as those seeking to consolidate public efforts to provide protection for women victims of violence (Dammert and Castañeda 2019). The police’s relationship with citizens, despite high levels of mistrust, was not recognized as an institutional challenge until the late 1990s. Even then, the institutional response to low levels of trust in the police has been only sporadic. Political debates and media coverage often focus on the police institution (Tello 2012). To gauge how well the public evaluates police work in action, perceptions of the performance of police stations and police on patrol are often better indicators, as they tend to be more proximate to citizens’ daily lives (Frühling 2009; Cruz 2015). When we focus on the work of local police stations, many red flags pop up. Although budget allocations have increased in recent years, and the number of police officers working in stations has risen to 47,000 (INEI 2018), structural shortcomings undermine the capacity of these officers. In 2018, the Instituto Nacional de Estadística e Informática (INEI; National Institute of Statistics and Information) published its sixth census of police stations and found that each officer works with 673 inhabitants on average, without considering shift

170

Making Police Reform Matter in Latin America

work (INEI 2018). Accounting for the different types of schedules and reassignments in each shift, the national strength of the PNP does not exceed 14,000 police working in local stations; that is, more than 2,250 people for each police officer in police stations (INEI 2018). The crisis of local police stations is multidimensional, ranging from infrastructure problems to lack of leadership. To responding to these ongoing problems, the Ministry of the Interior named 2017 the “year of the revolution of the police stations” (Gestión 2017). At the end of 2017, approximately one fourth of police stations did not even have their own premises and lacked legal ownership of the properties, which of course has a major effect on their renovation, construction, and maintenance. Only 64.7 percent of stations have services such as drinking water, drainage, and electrical energy permanently (INEI 2018). In addition to these infrastructure problems, there are leadership challenges. The local police commissioner (comisario) has minimal freedom to define strategies or initiatives, which poses obstacles for addressing unique local needs. Most activities conducted at the level of the local police precinct (comisaria) are planned and designed at national headquarters with a nationwide perspective. This makes it extremely difficult to tailor patrolling and community relations to the needs of individual neighborhoods. In addition, the constant rotation of the comisario (every two or three years) impedes the consolidation of key relationships with local community leaders and leaves little room for local innovation. Despite these profound limitations on police work in local districts, in 1997 the PNP created the Oficina de Participación Comunitaria (OPC), with the aim of strengthening citizen security in its jurisdictions by establishing a solid relationship between citizens and police. By the end of 2017, OPCs operated in 95 percent of the country’s police stations (INEI 2018). Among their primary functions is formulating working guidelines to bring the police closer to the community, with the goal of helping police to prevent and fight crime. Like other regional community policing initiatives, the OPCs aim to work directly with neighborhood councils to develop activities in citizen security, as well as promote social protection programs with the cooperation of the community and in coordination with citizen governing bodies and educational centers.2 The ministry’s legal and regulatory frameworks, as well as the country’s official citizen security forces, recognize and value community policing; however, the situation is very different in practice. In 2017, only 4.7 percent of the police station personnel were dedicated to the OPC. On average, only 13.0 percent of police station activities were devoted to foot patrol, and 28.9 percent of police stations did not engage in any foot patrols at all (INEI 2018). This is highly problematic, as much of the literature on public security finds that vehicular and/or random patrols are very inefficient and ineffective in controlling and preventing crime; in Peru almost

Peru

171

one third of all local police stations are engaged in vehicular patrolling or not patrolling at all—a worrisome number. Despite these challenges, since the early 2000s, the OPCs have become part of the institutional framework to achieve an active and organized cooperation with local communities, and thus make the communities allies in the fight against crime.3 In 2017, we conducted field research and visited various OPCs. We found that they primarily carry out two types of work: civic and emblematic actions. Civic actions include social assistance programs that seek to bring the PNP closer to the population, such as events offering afternoon tea or hot chocolate, haircuts, and parties. Emblematic actions aim to prevent crime through programs that involve the community in an organized manner. These occur in neighborhood councils, programs in schools (e.g., School Police, Self-Protection Brigades, Children’s Club), as well as organizations with important local actors (e.g., motorcycle taxi drivers) (Dammert and Castañeda 2019). Nevertheless, the distance between the rhetoric extolling community engagement and these sporadic activities is evident. Such intermittent programs do not render the community a “co-producer of security,” as envisioned in community policing paradigms. OPC offices must carry out the same programs in all jurisdictions regardless of the different realities of local communities, and these programs do not receive the same reception among communities. Rather than tailoring police response and engagement to meet the needs and context of local communities, OPC initiatives tend to rely on a one-size-fits-all approach. One of the main tasks of the community-oriented police is to establish links with community groups. In Peru, this typically translates into activities that aim to identify areas with a concentration of criminal activity, in conjunction with neighborhood councils and so-called cooperative networks. These networks are associations of citizens that support the police logistically and even financially to carry out their work of prevention and patrolling. In fact, one of the main activities of the OPC is organizing mixed patrols (or mixed rounds) with preventive and dissuasive purposes at critical points. The police personnel in charge of the OPC are noncommissioned officers, and they were brought into these roles due to service needs. They do not receive any special training, nor do they have tools that allow them to advance with a work strategy or strategic plan that effectively consolidates the relationship with the community. Possibly one of the main problems facing the OPC is the disconnect between their work and that of the commissioner and the PNP. Those assigned to the OPC are supposed to coordinate directly with the commissioner, but it is not clear how the information they collect from their activities is used. During our fieldwork, several interviewees expressed discomfort with their roles because they perceived that their work was undervalued by other

172

Making Police Reform Matter in Latin America

members of the police station (Dammert and Castañeda 2019). They felt that many officers did not believe in prevention or in working with the community, when what people demand is good treatment and responsiveness from all the police. Some OPC officers lamented that when they reported citizens’ concerns, they were ignored or boycotted. This lack of coordination ultimately undermines the OPCs’ ability to promote citizen trust in the police and those assigned to their police station. If citizens respond to OPC overtures and report something to them, but these reports are subsequently disregarded, OPC officers feel that they have failed the very people who trusted them. These perceptions are magnified in the area of community concerns regarding drug consumption, as the most effective response to reduce consumption is through public health programs to treat addicts, but OPC officers know that the only response their reports will elicit is a police raid. Ultimately, this cycle ends up empirically demonstrating that the information collected by the OPC offices is not systematized and used to tailor effective police responses to community problems. Rather than improving police–society relationships, the OPC runs the risk of exacerbating the frustration that citizens feel with the police, further widening the disconnect between them. Beyond the experience of the OPCs, there have been multiple initiatives developed by the PNP to try to improve police relationships with citizens. For example, based on widespread recognition that corrupt practices are closely linked to collecting bribes for traffic offenses, in 1998 the Peruvian government commissioned the World Bank and other agencies to carry out studies to address this problem. The main proposal suggested recruiting more female police personnel, who were perceived to be “more honest, disciplined, hard-working and reliable” (Karim 2011). The government accepted the proposal and increased the number of women serving in the police force, beginning with the traffic police, which by 2009 consisted primarily of women. Although this move aimed to reduce corruption, it unfortunately created equity problems in the police force. By 2011, estimates indicated that most women police officers spent more than 81 percent of their careers on traffic tasks, creating gendered divisions of labor in police ranks (Karim 2011). This trend has continued and intensified. At the end of 2017, most women who were assigned to transit finished their training as noncommissioned officers, with little consideration given to their preferences or specialization capabilities. In terms of corruption, we cannot definitively say that the increase in women transit police caused bribery levels to decrease; however, we find evidence that police bribery decreased after this reform. When we examine LAPOP’s survey question on police bribery, there is a sizable decline after 2008. As Figure 8.1 reports, in 2008, 22 percent of Peruvian respondents indicated that the police had solicited a bribe from them in the past year.

Peru Figure 8.1

173

Police Bribery in Peru (2006–2019)

Source: LAPOP’s 2006–2018/19 AmericasBarometer.

This number declined steadily right after the reforms to transit police, and by 2012 levels of reported bribery had fallen significantly to 16.4 percent. Despite an increase in 2016, bribery levels dropped to 15.7 percent in 2019—the lowest level recorded since LAPOP began asking this question in Peru. Although this percentage is significantly higher than the regional average, it still indicates improvement, as respondents reported significantly lower levels of bribe solicitation in 2019 compared with the high of 2008. During this time frame, bribe solicitation declined almost 30 percent, according to survey responses.

Crime Trends and Public Fear of Crime Despite institutional instability and lack of sustained engagement with the community, rates of violent crime have remained low in Peru, comparatively speaking. As Figure 8.2 illustrates, homicide rates register far below the Latin American regional average but have increased from historic lows of 5.4 in 2011 (CEIC 2018). Peru is one of the safest countries in the region when it comes to violent crime; however, the total number of murders has

174

Making Police Reform Matter in Latin America

Figure 8.2

Homicide Rates per 100,000 People in Peru (2011–2020)

Source: World Bank, World Development Indicators and InSight Crime, Homicide Roundup. Notes: We begin reporting homicide data in 2011, as the data prior to 2011 had sizeable discrepancies. Even within the same UNODC, data for some years was reported in starkly different terms in different publications. For example, in 2011 the UNODC reported the homicide rate of Peru as both 5.5 and 9.6 per 100,000 in different publications (GSH 2013 and UNODC) https://dataunodc.un.org/content /data/homicide/homicide-rate). We thus begin reporting with the year 2011, when the World Bank and UNODC online data began to converge.

risen steadily since 2011 and reached a new high in 2018. Until 2014, homicide rates were on par with the global average; however, they began to steadily rise the following year, and the sharp spike in 2018 could signal a deviation from these promising trends. It is also important to remember that the national average can mask more complex realities, such as that of Madre Dios, an area with a strong presence of illegal mining and drug trafficking, whose homicide rate in 2017 far exceeded that of other areas with a figure of 48.0 per 100,000 (Ministerio del Interior 2018). According to some reports, higher rates of homicide are linked to ongoing political unrest, which intensified in 2018 as allegations of involvement in the Odebrecht corruption scandal led to the resignation of President Kuczynski, the suicide of former president García, and the apprehension

Peru

175

and extradition of former president Toledo (Asmann and O’Reilly 2020). Following Kuczynski’s resignation, Vice President Vizcarra assumed the presidency in 2018 and announced a fight against corruption and impunity. By September 2019, Vizcarra dissolved Congress and was subsequently suspended by Peruvian lawmakers, but as discussed earlier, he succeeded in holding snap legislative elections in 2020 with high public approval ratings before Congress successfully removed him again in November 2020. As we noted earlier, the political instability of the Peruvian executive and legislative branches has reverberated throughout other institutions, making it difficult for reforms to gain traction. Furthermore, some observers have argued that this instability, or what former Interim President Aráoz called “broken constitutional order,” created opportunities for organized criminal groups to increase coca cultivation in a security vacuum along the Bolivian border; illegal mining to expand to new regions; and Shining Path guerrillas to increase drug trafficking operations and attacks against the state (Asmann and O’Reilly 2020). These illicit activities are not new. In 2015, three Peruvian provinces announced a state of emergency due to drug trafficking and insecurity. Still, the most recent rounds of constitutional crisis could have created further opportunities for organized criminal groups, which thrive in the absence of a strong state presence. Homicide rates are very low in Peru compared with the Latin American regional average, but rates of self-reported victimization are very high. According to LAPOP’s national surveys, the percentage of Peruvians selfreporting victimization increased steadily from 2008 through 2019; in 2019, self-reported victimization in Peru was significantly much higher than the regional average. Victimization rates reached a high in 2019, when more than a third of Peruvian respondents reported they had been victimized in the past year. In some survey years (2006, 2010, 2012, 2014), LAPOP included follow-up questions to probe what type of crime victimization had occurred. In 2006, approximately two thirds of self-reported victims indicated that the crime had not been violent (e.g., robbery without violence, robbery from one’s home, vandalism). By 2014, only 46.8 percent of selfreported victims indicated that the crime was not violent, marking a rise in the number of violent crimes such as physical assault or aggravated robbery (excluding homicides). National crime data, based on crimes reported to the police, also indicate that the use of violence in the commission of crime has increased (INEI 2018). Taken together, these data on homicide and selfreported victimization indicate that although homicide rates are low for the region, many Peruvians are still victimized by high levels of nonviolent crime such as robbery, and violence in the commission of crimes appear to be on the rise. National averages often hide situations that may be more complex or have a greater effect on the political agenda. Lima is home to almost half of

176

Making Police Reform Matter in Latin America

the national population and has a key role in framing the discourse around public security and policing. Like many Latin American capitals, Lima dominates the political agenda, yet it deviates in important ways from national trends. For example, victimization rates in Lima are significantly higher than the national average and have increased at higher rates over time. In 2019, this difference was most pronounced: 43.8 percent of Lima’s residents reported experiencing some type of victimization in the past year. In other urban areas, rates of victimization have been high but stable from 2010 to 2019. Not surprisingly, self-reported victimization rates in rural areas are significantly lower than urban ones; however, even in rural areas victimization has increased sharply over time, increasing 37.5 percent between 2014 and 2016 in rural areas (from 17.8 percent to 24.5 percent). Residents of rural areas report less victimization than the national average, but they are still witnessing steep increases in recent years. Public perceptions of crime reflect these recent increases in victimization. According to data collected by the observatory Lima Cómo Vamos, insecurity persists as the Lima’s biggest problem. When asked to identify pressing problems, 82.2 percent of respondents in 2019 signaled insecurity, followed by public transportation with 46.2 percent (Lima Cómo Vamos 2019).4 In contrast, at the national level, INEI data showed that in the period of September—December 2018, most respondents (64.9 percent) reported that the main problem of the country was corruption (64.9 percent); crime took second place with 38.9 percent (INEI 2019). As Figure 8.3 illustrates, fear of crime is consistently highest in Lima. After declining for several years, fear of crime in Lima increased between 2012 and 2019. Fear of crime in other cities and rural areas was much lower than in Lima, but still on an upward trajectory from 2012 through 2019. Despite some smaller fluctuations, when asked about potential victimization in their own neighborhoods, Peruvians registered consistently high levels of insecurity. National averages of fear of crime in one’s own neighborhood are consistently higher in Peru than the Latin American regional average. In Lima, there are important socioeconomic differences in perceptions of insecurity. For example, in 2019, respondents from the lowest income sector reported much higher feelings of insecurity in their neighborhoods than those from the highest income sectors: 54.6 percent of the most vulnerable respondents indicated that they felt very insecure in their neighborhoods, compared with 43.1 percent of those in the highest socioeconomic group (ENAPRES 2019). This disparity reflects the marked socio-territorial segregation of Lima. Furthermore, 25.5 percent of respondents in the highest income category reported that they felt very secure (a rate exceeding global averages), compared with only 18.3 percent of the most impoverished group. These differences highlight the very separate realities the poor and the wealthy experience in the same city and underscore the pres-

Peru Figure 8.3

177

Fear of Crime by Area in Peru (2006–2019)

Source: LAPOP’s 2006–2018/19 AmericasBarometer.

ence of “two cities” (Merrifield 2007). This segregation is also reflected in the 2018 data on the presence of gangs in Lima. When asked about gang activity, only 6.3 percent of those in the highest socioeconomic level reported the presence of gangs in their communities, compared with 13.2 percent of the most vulnerable respondents (Lima Cómo Vamos 2018). When we analyze LAPOP’s 2019 national data, we find that fear of crime in one’s neighborhood is significantly linked to income levels and the size of respondents’ municipalities. As Table 8.1 reports, poorer respondents are significantly more fearful of crime in their neighborhoods, and fear of crime increases significantly as the size of the municipalities increases. Men and older respondents reported significantly higher levels of fear of crime. Most important, crime victims are systematically and significantly more fearful of future victimization in their neighborhoods. Given that rates of self-reported victimization are quite high in Peru, it is likely that fear of crime will continue to reflect this reality and remain high as well. Taken together, these data on crime indicate that although Peruvian homicide rates are low for the region, they are increasing. Furthermore, since the mid-2010s, rates of victimization have been quite high and

178

Making Police Reform Matter in Latin America

Table 8.1

Fear of Crime in Peru (2019)

Independent Variables

Fear of Crime

Education

−0.009 (0.016) 0.073*** (0.018) −0.543*** (0.103) −0.051*** (0.013) 0.200*** (0.035) 0.630*** (0.108) 0.104 1,350

Age Gender (men) Income Size of municipality Self-reported victimization   in past year Adjusted R2 N

Source: AmericasBarometer 2018/19, LAPOP. Note: Table reports B coefficients from ordinal logistic regression analysis. ***p < 0.001.

increasingly more violent. High rates of fear of crime appear to respond to victimization, and these trends have combined to render public fear of crime in Peru quite high. Peru might not face the same challenge of homicide and organized crime as some other Latin American countries, but its public security policies unfold against a backdrop of high levels of victimization, public fear of crime, and political instability.5

Perceptions of Police The lack of clear and sustained public policies over time, as well as high levels of impunity, have contributed to decidedly negative perceptions of the police. Multiple studies have documented that most Peruvians view their police as corrupt, slow, inefficient, and in some cases violent (Dammert 2022). Here, we rely on LAPOP’s 2006–2019 survey data to examine key indicators of public perceptions of police performance over time. Our findings confirm those of other studies that highlight low levels of trust and poor public evaluations of the Peruvian police while pinpointing areas that are particularly problematic. We begin with a survey item that allows us to gauge public perceptions of a fundamental component of police work. When asked whether police protect people from crime or are themselves involved in crime, respondents revealed a very low regard for the police. In 2008, only about one third of

Peru

179

respondents thought police protected local communities. In 2012, this percentage had declined further: only 28.6 percent of respondents thought that police fulfilled their primary function in protecting communities. In these years, a little over half of respondents said that police are primarily involved in crime, a very high percentage that ranks in the top third in the region, on par with Mexico. These survey responses indicate that Peruvians perceive their police more as criminal actors than as community protectors. When we examine public perceptions of police response time, we find that many Peruvians register quite pessimistic evaluations. On average, respondents tended to think that it would take police one to three hours to respond to a routine call for a burglary, and this average has remained very stable across multiple survey waves. This tends to be a longer estimate than most countries in the region, situating Peruvian perceptions of police responsiveness in the bottom half of Latin American countries. Perceptions of overall police performance in respondents’ neighborhoods was equally pessimistic. In 2014, LAPOP asked respondents to assess police performance in their immediate communities, and only 1.4 percent stated that they were very satisfied. An additional 29.2 percent were satisfied, but an overwhelming 69 percent indicated that they were either dissatisfied or very dissatisfied with police presence in their neighborhoods. In more surveys of urban areas, the data suggests where both fear of crime and victimization are most acute, respondents perceive that police presence is weak, and penalize the police for this with poor evaluations (INEI 2018, 2020a, 2020b). Indeed, the lack of police presence and increasing pressure from citizens seems to have provided the impetus for the creation of a special municipal unit, the Serenazgos. This local municipal unit oversees street patrols, mediates local problems, and calls the police when needed. Because municipal workers do not have the authority to carry weapons, in most cases the Serenazgos’ presence is perceived as a deterrence factor as well as an opportunity to resolve some community needs. According to INEI (2018), at least 30 percent of Peruvians report that the police do not maintain a presence in their neighborhoods. Only 26.2 percent reported that the PNP actively engaged in surveillance of their neighborhoods, while an additional 34.3 percent indicated that surveillance was conducted by municipalities via the Serenazgos. These trends continued into 2020, as surveys conducted in January 2020 found that 26.7 percent of respondents reported PNP surveillance in their neighborhoods, and 36.6 percent reported Serenazgo surveillance (INEI 2020a). These factors coalesce to create a scenario where citizen trust in police is limited. When we examine trust in police longitudinally, we find that from 2006 through 2019, average levels of trust in police in Peru are below the already low Latin American regional average (see Figure 8.4). Trust in police tends to fall just above trust in other political institutions, but as our

180

Making Police Reform Matter in Latin America

earlier discussion of political instability and the 2019–2020 constitutional crisis revealed, there is a widespread crisis of public trust in virtually all Peruvian institutions. According to an INEI survey conducted in urban areas in Peru, between July and December 2019 trust in police has started to slip below that of other institutions. INEI (2018) found that police earned some of the lowest levels of institutional trust, as only 16.4 percent of respondents indicated they had some trust in police, compared with 81.3 percent who registered distrust.6 These low levels of public trust in police fuel the perception that the justice system overall is not capable of curbing criminality and serves as a mere revolving door for criminals. When asked whether the justice system will be able to punish guilty parties, on average Peruvians have quite pessimistic views.7

Public Perceptions of Public Security Policies We turn to determining if these low levels of trust in the police lead to support for more punitive public security policies. Support for penal populism

Figure 8.4

Trust in Police in Peru (2006–2019)

Source: LAPOP’s 2006–2019 AmericasBarometer.

Peru

181

has proven quite strong in Peru. Citizen frustration with police performance, particularly in electoral periods, has been a key trigger for increased political concern. Low levels of trust in the police have been linked to punitive populism, which has featured prominently in public policy debates over the past two decades. Proposed measures have included the death penalty for rapists and reopening a prison located on an island near Lima to house dangerous criminals. These initiatives generated debate and citizen approval but to date have not been implemented. When we examine LAPOP survey responses over time, we find consistently high levels of support for proposals that would allow authorities to act on the margin of the law, police torture of suspected criminals, and domestic military operations to combat crime. Support for preventive measures tended to be lower than for punitive ones, as we demonstrate in this section. From 2006 through 2012, LAPOP asked respondents, “In order to catch criminals, do you think that the authorities should always abide by the law or that occasionally they can cross the line?” Throughout this time frame, a sizeable portion of the public supported police acting on the margin of the law, ranging from 42.3 percent in 2012 to 50.5 percent in 2010. In 2012, LAPOP included a measure that was more extreme: support for police torture of suspected criminals: “If the police torture a criminal to get information about a very dangerous organized crime group, would you approve of the police torturing the criminal, or would you not approve but understand, or would you neither approve nor understand?” Here, 39.6 percent of Peruvian respondents supported police torture in this scenario, and an additional 33.4 percent would not support but would understand such extralegal action. Less than a third of respondents categorically rejected police torture, a low percentage given the severity of the topic. We identify high levels of support for military involvement in domestic security matters. In LAPOP’s 2012 and 2014 surveys, on average, there was a great deal of support for domestic military operations in the name of fighting crime.8 Average responses remained stable from year to year, at 5.2 on a 7-point scale, indicating high levels of agreement that military operations would be welcome. Given the comparatively low rates of homicide, as well as the dubious track record of domestic military operations during the counterinsurgency campaign of the 1980s and 1990s, these high levels of support are notable. In 2012, LAPOP asked respondents to indicate their level of support for three kinds of public security policies. Respondents did not have to choose among them, they only needed to state their level of support for each item separately; they could agree strongly that all three were viable approaches.9 As Figure 8.5 illustrates, support for harsher laws and preventive programs enjoyed high levels of support (3.6 on a 4-point scale). Interestingly, investments in private security garnered significantly less support.

182

Making Police Reform Matter in Latin America

Figure 8.5

Best Way to Fight Crime in Peru (2012)

Harsher laws

Prevention programs

Private security Error bars: 95% CI 1.0 Strongly disagree

2.0

3.0

Average support

4.0 Strongly agree

Source: LAPOP’s 2012 AmericasBarometer.

When respondents were asked to select between punitive and preventive measures, punitive measures clearly garnered more support. LAPOP posed the following question in 2012, 2014, and 2019: “In your opinion, what should be done to reduce crime in a country like ours: (1) Implement preventive measures; (2) Increase punishment of criminals; (3) [Don’t read] Both.” When asked to choose between the first two options, punitive measures proved far more popular. In 2012 and 2014, 56–57 percent of respondents voiced support for punitive measures over preventive ones. In 2019, this percentage increased to 63 percent—almost two thirds of Peruvian respondents favored more punishment over investments in prevention programs. In 2019, the percentage of respondents who favored a balanced approach between punishment and prevention, even though this option was not read aloud in the survey, shrank to its smallest yet, 4.2 percent. When asked to choose among public security policies, there is a consistently high level of support for harsher punishments, compared with more tepid levels of support for investing in preventive measures to tackle the root causes of crime. In 2012 and 2019, LAPOP asked respondents if they would buy a gun for protection if they had the means. In both years, roughly half of respondents expressed an interest in acquiring a gun for protection. For some crimes, many Peruvians indicated that harsher punishments could include the death penalty. In 2019, during a time the death penalty

Peru

183

was debated publicly, LAPOP included several questions to gauge public support for such an extreme punitive measure in cases of homicide, rape, and corruption. Perhaps due to the intensity of the public debate, support for the death penalty was extraordinarily high for crimes involving rape of a minor and the rape of a woman: 88.6 percent and 80.2 percent, respectively. Support for the death penalty for homicides was also high at 67.9 percent; fewer respondents supported the death penalty for corruption crimes, but it is still notable that more than half—53.8 percent—stated that they would support the death penalty under this scenario. For a corruption crime, this is an extraordinarily high level of support, but this statistical finding could be because corruption had intensely dominated public discourse for the two years before the survey, which could perhaps lead respondents to be particularly harsh in their responses to this item (Malone 2023). In most cases, support for these types of punitive policies is significantly linked to respondents’ trust in the police. Table 8.2 reports the bivariate correlations between these measures of punitive populism and trust in the police, highlighting significant relationships.10 Lower levels of trust in police translate to more support for obtaining a personal gun, police torturing criminals, the death penalty, and punishment over prevention. These results suggest that people’s support for these measures at least partially reflects their frustration with police. When people do not trust the police, it makes them more willing to endorse blunter, harsher measures to reduce crime. There are a few exceptions to this trend. There is no significant relationship at the bivariate level between trust in police and support for the military to engage in crime-fighting measures. When respondents did not have to choose among different crime fighting measures (e.g., harsher laws, prevention, and private security), we do not see a significant relationship between support for these policies and trust in police. Even taking these exceptions into account, the results in Table 8.2 suggest that trust in police is linked to attitudes on several different punitive security measures. In our concluding chapter, we probe this relationship in greater depth in the region.

Police Perceptions of Institution and Public Security Policies Thus far, we have examined the frustrated process of police reform in Peru, contextualized against a backdrop of low rates of homicide but high rates of victimization and public fear of crime. We have noted that Peruvians regard the police quite negatively and register high levels of support for punitive public security policies. Missing from this analysis so far is the voice of the police themselves. How do they view their work within the larger institution? How do they regard their work and their relationships with the

184

Making Police Reform Matter in Latin America

Table 8.2

Punitive Populism and Trust in the Police in Peru

Measure of Punitive Populism

Correlation with Trust in Police (1) Not at All – (7) A Lot

Obtain a gun for protection   (0) no, (1) yes Police could torture criminal to get information   (1) would not approve or understand;   (2) would not approve but would understand;   (3) would approve The armed forces should participate in   combatting crime and violence   (1) strongly disapprove – (7) strongly approve Support for death penalty (index)a   (0) never – (6) in every circumstance listed Harsher laws best way to fight crime   (1) strongly disagree – (4) strongly agree Prevention best way to fight crime   (1) strongly disagree – (4) strongly agree Private security best way to fight crime   (1) strongly disagree – (4) strongly agree Prefer more punishment over prevention to   reduce crime   (1) prevention; (2) both; (3) more punishment

−0.077** −0.058*

0.024 −0.064* −0.011 0.018 0.042 −0.055*

Notes: The table reports Pearson’s correlation coefficients and significance levels. In Peru, LAPOP included questions measuring trust in police in every survey wave, and questions gauging support for punitive populism in several different survey years. Correlations between trust in police and obtaining a gun for protection and support for capital punishment are from the 2019 survey. The correlation between trust in police and support for the military to fight crime are from pooled data from 2012 and 2014. The correlation between trust in police and support for punishment over prevention is from pooled data from 2012, 2014, 2017, and 2019. Finally, the correlation between trust in police and support for police torture of suspects is from the 2012 survey. a. This index was created by using an additive index to combine the four separate survey items reported in Figure 8.9. The Cronbach’s alpha for this index is 0.749. *p < 0.05; **p < 0.01; ***p < 0.001.

communities they serve? What types of public security policies do they see as most viable? In spring 2018, we conducted a survey of thirty-five high-ranking police officers in Peru who were enrolled in a program at the School of Government of the Pontificia Universidad Católica del Perú, an educational prerequisite for promotion to general. We now analyze these data and compare public attitudes on policing and public security to those of the police themselves. When asked about their training and qualifications, police officers in our survey were quite positive in their responses, perhaps reflecting the fact

Peru

185

that public investment in the police has risen over the past three years. Forty percent of respondents indicated that they had earned university degrees, and 71.4 percent said that their training was either comprehensive or very comprehensive. Indeed, 77.8 percent said they had received specialized training less than a year ago, indicating that there were sustained efforts to strengthen and support higher level police officers. When asked about salary, 85.7 percent of respondents reported satisfaction with their compensation, indicating that their salaries were high enough. The police respondents did identify areas for improvement, however. When asked about the areas of training and education that were lacking, training on the new penal system and oral trials ranked first, followed by training on laws and procedures.11 Methods of conflict resolution and mechanisms for transparency and accountability were frequently ranked second.12 In contrast, items related to weapons training and police tactics were not cited often. When asked about their daily work, once again the police were overall quite positive. All respondents either agreed or agreed strongly that they liked the type of work they did and they got along well with their colleagues.13 One issue that emerged as problematic was their workload: 42.9 percent either agreed or strongly agreed that the quantity of work made it difficult for them to do their jobs well, and an additional 17.1 percent neither agreed nor disagreed. When asked about how the police worked today compared with the past, 61.1 percent said it was better and 25.0 percent said it had stayed the same; only 13.9 percent said it had worsened. When asked more specifically about work conditions, officers reported high levels of overall satisfaction with the organization and their autonomy within it. Most indicated that they were open to new ways of doing things and new ideas in their work. However, levels of satisfaction with equipment were much lower. For community policing models to work well, it is imperative to discard police performance metrics that prioritize punitive elements such as number of detentions and focus more squarely on indicators that can address different dimensions of police performance that are community oriented. The police officers in our survey tended to agree that these latter measures of police performance better tapped into the work that they performed. When asked “Which of the following do you think are the best results of your work?,” only 25 percent indicated that the number of arrests and proceedings would be the best result of their work, compared with 42 percent who answered “by the number of requests for help attended.”14 When asked why they became police, 88.9 percent said to help to serve (first); combat insecurity was the second response most police gave (55.6 percent), but a smaller number indicated other second options like family tradition (13.9 percent). Sixty-six percent said they thought there were

186

Making Police Reform Matter in Latin America

opportunities for advancement within the police, while 30.6 percent said such opportunities were few. Eighty percent said that they expected to advance in their careers either in their current branch or another part of the institution. On the other hand, 67.6 percent said that it would be difficult to leave the job (agreed or strongly agreed), even if they wanted to make the change. This could indicate a lack of alternative career paths for those unhappy with or tired of police work. We asked the police officers to identify two things that the police as an institution should prioritize. “Application of strategic operations” ranked first for 30.6 percent of our respondents, and another 22.9 percent ranked it second. Twenty-five percent listed “collaboration with citizens” as first, followed by 22.2 percent who prioritized work equipment (34.3 percent ranked work equipment second, but only 5.7 percent listed collaboration with citizens second). We included questions to assess what the officers thought the best way was to reduce crime and address public insecurity: “Now, we are going to discuss things that one hears in the street and on the news, when people talk about options to fight crime. Please, could you tell us if you (1) strongly disagree – (5) strongly agree with the options below.” As Figure 8.6 illustrates, respondents strongly favored public security policies that focused on prevention and addressed the root causes of insecurity (e.g., unemployment). There was less support for harsher penalties for criminals, with average responses in the “neutral” zone of the range. More repressive measures, such as military patrols and allowing authorities to act on the margins of the law, received very low levels of support, which is in clear contrast with the popularity of these measures in the broader public. The police and the public concurred that private security measures were undesirable. We also included a question regarding police torture of suspected criminals, posed earlier to LAPOP respondents. In contrast to the public support for police torture, police in our survey were very hesitant to endorse this measure: 94.4 percent said that they would neither support nor understand police who tortured suspects. Of course, it is important to note that social desirability bias could inflate this percentage; as our respondents were pursuing this educational program for later promotion to the rank of general, it is possible that this shaped their responses. We then posed a series of questions that focused more squarely on the role of the community in public security: “Thinking about how to organize patrols in your division, how effective do you think police patrols are in the following activities? (1) not effective at all – (5) very effective.” As Figure 8.7 illustrates, officers tended to think that police patrols were very effective in a variety of ways. Learning about the community and fostering good relationships between police, communities, and local businesses ranked the

Peru

187

Figure 8.6 Police Perceptions of Public Security Policies in Peru (2018)

Prevention programs Programs to increase employment Harsher punishments for criminals Armed forces work with police Invest in private security Authorities act on margins of law

stongly disagree

Average Support

strongly agree

Source: Survey conducted by authors of Peruvian police personnel, 2018.

highest. Four other objectives also registered high average scores, indicating that these respondents did perceive patrols as effective ways to address these different objectives: preventing crime, training new police officers, reducing fear of crime, responding to calls in the neighborhood. Overall, police officers in our survey were very supportive of patrols on foot in local communities. Substantial majorities thought patrols on foot were good investments, were better at reducing fear of crime, and allowed police to have a great deal of informal contact with members of the community. The vast majority, 85.7 percent, either agreed or strongly agreed that police can learn more about a neighborhood by patrolling on foot than in a cruiser. This high percentage contrasts sharply with the actual occurrence of foot patrols, which as noted earlier did not enjoy a great deal of police time. Interestingly, when we posed the question about anticipated police response time (from LAPOP’s national survey of Peru), police were far more optimistic about their response time. When we asked them to estimate how much time would it take to arrive at a house to respond to a robbery, 30 percent indicated that they thought police would respond in less than ten minutes, and an additional 47.2 percent stated that police response time

188

Making Police Reform Matter in Latin America

Figure 8.7 Perceived Effectiveness of Police Patrols in Peru (2018) Learn about the community Foster relationship with community Foster relationship with local businesses Prevent crime Train new police Reduce fear of crime Respond to calls in the neighborhood Manage non-criminal activities in the street

Not at all effective

Average perceived effectiveness

Very effective

Source: Survey conducted by authors of Peruvian police personnel, 2018.

would be between ten and thirty minutes. A small minority (16.7 percent) thought response time would be more than thirty minutes. When asked about their relationships with local communities, police officers reported very different perceptions. The majority (55.6 percent) said that the police were close to communities, but 36.1 percent stated that they were not close. Only two respondents thought that police and communities were “very close.” Almost all police in our sample—97 percent— indicated that the police routinely have meetings with citizens. We included a battery of questions to assess how police envisioned the role of the community in reducing insecurity. As Figure 8.8 indicates, on average police agreed that helping citizens was as important as upholding the law and that police should collaborate with communities to solve problems. There were more tepid responses when asked about the centrality of citizens’ role in security. When asked whether citizen collaboration was essential for solving crimes, or whether crime prevention was the responsibility of citizens, police respondents were less enthusiastic. They also registered low levels of interest in addressing problems in a community if they were not crimes, and they did not necessarily think that citizens knew more

Peru Figure 8.8

189

Police Perceptions of Community Role in Public Security in Peru (2018)

Helping citizens is an important as upholding the law Police should collaborate with citizens to solve neighborhood problems Without the collaboration of citizens, most crimes will never be solved The prevention of crime is responsibility of police and citizens Police should try to solve problems on patrol, even if not criminal Citizens know what happens in their neighborhood better than police Relationship between police and citizens is good on my patrol

Strongly disagree

Average

Strongly agree

Source: Survey conducted by authors of Peruvian police personnel, 2018.

about the activity in their neighborhood than police did. Even though police believed citizen collaboration was important in the abstract, it is not clear whether they believed citizens should have a large role in addressing public insecurity. Police also regarded relationships with citizens on patrol as not particularly strong. Finally, we asked police officers to identify the principal results of having a closer relationship with the residents of a municipality. As Table 8.3 reports, 57.1 percent indicated that the key result was to build more trust. Smaller percentages reported that the primary results were to identify problematic zones (20.0 percent) and have a greater police presence (11.4 percent). Smaller percentages indicated that the primary results of having closer relationships with communities was to learn about security problems, and only 5.7 percent replied that it was to involve citizens in the search for solutions. Despite their interest in engaging the community, officers in our survey identified public mistrust as a potential obstacle to such collaboration. Half of the respondents in our survey indicated that most people do not respect the police, and 83.3 percent said that most people did not understand the problems of police.

190

Making Police Reform Matter in Latin America

Table 8.3

Main Result of Closer Police–Community Relationships in Peru (2018)

Main Result of Closer Relationship with Community Members in a Municipality Build more trust Identify problematic zones Have a greater police presence Learn about security problems Involve citizens in the search for solutions

Percentage Ranking Item First

Percentage Ranking Item Second

57.1 20.0 11.4 5.7 5.7

14.3 14.3 20.0 20.0 31.4

Conclusion: Learning from Nonsuccessful Cases The Peruvian experience highlights the complexities involved in police reform when there is not sustained political support or civilian technical capacity to sustain the implementation of initiatives (Huaytalla 2016). In contrast to Chile (Chapter 3), for example, chronic institutional instability has consistently undermined police reforms. Despite good intentions and political rhetoric, reforms in Peru have languished due to the volatility of political support; the low recognition of the importance of designing, implementing and evaluating initiatives; and institutional barriers. The National Police of Peru have made clear advances in terms of public investment in salaries and the infrastructure of specialized offices, but critical challenges remain. Although police and government narratives include prevention activities, in practice daily police work replicates what has traditionally been done. This situation reverts to low levels of efficiency and limited police coverage in important territories of the country. A strategic plan that defines medium and long-term policies is not enough given that in recent years, policies with that level of definition have been approved but not truly implemented. Good intentions alone cannot solve security problems. Institutional stability, sustained political will, and evidence-based institutional transformation are required to address the security problems facing the country.

Notes 1. The Obredecht scandal is one of the largest corruption scandals in Latin American history, entangling a series of politicians across the region. For an accessible overview, see BBC (2019). 2. This is similar to the community prevention models in Nicaragua before 2018 (see Chapter 6).

Peru

191

3. For more detailed information on these specific activities, see Dammert and Castañeda (2019). 4. Lima Cómo Vamos is a data observatory. They asked respondents, “What are the three most important problems that affect quality of life in Lima?” 5. Of course, victimization rates are not the only factor driving high levels of fear of crime. The literature on fear of crime finds that other factors are also important, as fear can often be a manifestation of other types of economic, political, and social insecurities (Dammert and Malone 2003). Given the findings in the literature, it is likely that Peru’s political instability could lead public perceptions of insecurity to increase. 6. This survey question read: “Do you have trust in institutions like the National Police?” 7. There is a strong correlation between trust the justice system will punish the guilty party and trust that the police will capture suspected criminals. When these two questions were included on the same survey in 2008, r = 0.589, p < 0.01. 8. “The Armed Forces ought to participate in combatting crime and violence in [country]. How much do you agree or disagree?” Responses ranged from (1) strongly disagree to (7) strongly agree. 9. These items were preceded by an introduction: “I am going to read you some things you hear on the street or in the media when people talk about ways to combat crime. Please tell me if you strongly agree, agree somewhat, somewhat disagree or strongly disagree with each one of them. The best way to fight crime is to . . .” Respondents were asked if they (1) strongly disagreed to (4) strongly agreed with the following public security options: “create prevention programs”; “be tougher on criminals”; “contract private security.” 10. Preliminary results from this analysis were reported in Malone (2023). 11. Twenty-six percent of respondents ranked training on the new penal system/oral trials first, followed closely by 23 percent who ranked laws and procedures first (an additional 17 percent listed it second). 12. Twenty-three percent of respondents ranked method of conflict resolution as second most important, and 17 percent ranked transparency and accountability as second. 13. When asked how their colleagues got along with each other, 45.7 percent said that their colleagues didn’t get along well sometimes, and another 22.9 percent were neutral on the matter. 14. Another frequent response was “by keeping up to date on my work” (19.4 percent).

9 Uruguay: Success in a Social Welfare State

ALONG WITH CHILE, URUGUAY HAS LONG BEEN CONSIDERED ONE of the safest countries in Latin America (Galiani and Jaitman 2015; PNUD 2013). While recent upticks in criminality have started to tarnish this reputation, rates of violent crime in Uruguay remain much lower than other countries in Latin America. As we noted in Chapter 2, in LAPOP’s 2018/19 AmericasBarometer, Uruguay registered the highest average levels of trust in police in the region. These high levels of trust have proven quite durable, as Uruguay recorded the highest levels of trust in 2017 as well (Malone and Dammert 2020). In the early 2010s, the Uruguayan police embarked on a reform and modernization process that led to profound institutional changes, and the police increasingly relied on patrol and prevention strategies that earned praise for their potential to improve public security and police– community relations. While not highlighted as frequently as Chile and Colombia, Uruguayan police reform has registered successes with less fanfare and has established itself as one of the most promising examples of institutional change in the region (Dammert 2016; Serrano-Berthet 2018). Like its neighbors in the Southern Cone, a military dictatorship ruled Uruguay from 1973 to 1985. The return to democracy represented a critical juncture for police reform, as democratic governments prioritized the demilitarization of police, and this institutional transformation set the stage for community-oriented policing initiatives in the late 1990s that have continued until the present day. From 2005 to 2020, the Frente Amplio (Broad Front), a center-left political coalition, governed Uruguay and introduced different approaches to the problem of criminality and public security. As with Chile (Chapter 3), this institutional stability gave reforms the time to take root. The Frente Amplio linked criminality and public insecurity to structural 193

194

Making Police Reform Matter in Latin America

problems of poverty and inequality and matched changes in policing practices with sustained investment in social services and human welfare. Even before the Frente Amplio’s electoral victory, Uruguay, like Costa Rica (Chapter 5), had historically maintained a large social welfare system, and these additional investments provided impressive resources to tackle structural inequality. Thus, changes to policing practices were supported by sustained investment in social welfare. Given the Frente Amplio’s attention to the underlying root causes of criminality, its extensive engagement with police reform, and its emphasis on preventive policing, it is perhaps ironic that it was precisely the issues of criminality and public perceptions of insecurity that contributed to the Frente Amplio’s electoral defeat in 2020. Nagging problems of criminality and fear of crime generated strong opposition to the Frente Amplio, paving the way for the electoral victory of the opposition, led by President Luis Lacalle Pou (2020–present). Upon assuming the presidency, Lacalle proposed to fight crime with longer prison sentences, transformations of the police institution, and improvements in crime control and prevention. His government has signaled that it aims to distinguish its policing programs from those of the Frente Amplio, but it is not yet clear whether these measures will become a counterreform or merely a rhetorical campaign pledge.

Institutional Transformation Since its creation in 1829, the Policía Nacional de Uruguay (PNU) has been in charge of public security; its functions include prevention, control, criminal investigation, and work in the prison system. The PNU falls under the auspices of the Ministry of the Interior, which is responsible for crime prevention and control policies. The military dictatorship (1973–1985) shifted the structures and mission of the police, along with all other security institutions. The military government redesigned the structures, dependencies, and missions of the PNU and placed it under the oversight of the Joint Chiefs of Staff. The police took on a political role, targeting opponents of the regime and dissidents and engaging in human rights violations. After the return of democracy in 1985, the civilian government took advantage of this critical juncture, succeeded in exerting more control over the police, and introduced changes to professionalize the police and reorient officers to respond to the daily problems facing citizens. The first democratic government was that of President Julio María Sanguinetti (1985–1990), who prioritized demilitarizing police headquarters and the restitution of those who were expelled from the police force during the dictatorship. Despite good intentions, the transformation processes progressed slowly. The Luis Lacalle government

Uruguay

195

(1990–1995) continued reforming the police and demonstrated real successes in improving the quality and organization of police work. These reforms included salary increases and improved social protections for police officers, especially after a police strike in 1992. Sanguinetti’s second term (1995–2000) faced a markedly different security context, as crime rates increased, along with public perceptions of insecurity. Institutional changes succeeded in demilitarizing and professionalizing the police, but for many citizens, these reforms did not adequately address the problems of public insecurity they encountered on a daily basis. Many were critical of the attention their government had paid to public security and were frustrated that security had not been central to previous government agendas. During this time, a series of legislative and operational changes increased criminal penalties and punishments, in addition to strengthening police capacity to respond to rising crime rates. Uruguay’s unique circumstances helped the country secure a loan from the Inter-American Development Bank to address security issues, which facilitated the formation of the Citizen Security Program (1998–2004). The program fueled a broader debate on community policing and emphasized the importance of systematizing criminal information and developing crime prevention initiatives. Frente Amplio won elections in 2005 and remained in power until 2020. During the 2005–2020 period, Frente Amplio emphasized civilian control of the police, reduced levels of police discretion, and increased human rights protection. Initially, the first government of Tabaré Vázquez (2005–2010) recognized the structural problems of inequality, poverty, and social vulnerability as the root causes of crime (Paternain and Vila 2010; Paternain 2014). The Vázquez government framed the problem of public insecurity as a consequence of structural inequalities and concentrated on improving social welfare to tackle criminality. While the emphasis was on the underlying socioeconomic factors of criminality, the government also invested in consolidating information systems and creating the National Observatory of Violence and Crime, one of the first such institutions in Latin America (Vila 2010). In addition, the government strengthened and expanded the presence of civilians in the Ministry of the Interior, creating civil servant career paths to replace the various spaces previously occupied by active or retired police officers. In 2007, the government created the position of the Director of the National Police, an office that reports directly to the minister of the Interior and can be held by a civilian. The subsequent governments of José Mujica (2010-2015) and Vázquez (20152020) advanced the agenda of “the new police,” which we analyze later in this chapter. There was a great deal of continuity between the governments of Mujica (2010–2015) and Vázquez (2015–2020), which enabled the sustainability of reforms (similar to the Chilean case). Specifically, Minister of the Interior Eduardo Bonomi held his post between 2013 and 2020, marking

196

Making Police Reform Matter in Latin America

unprecedented stability in a regional context of constant rotations. In parallel efforts, the government legalized the production, sale, and consumption of cannabis with the goal of reducing public insecurity (Queirolo et al. 2018) and overhauled the prison system (Dammert 2016). These policies of the Frente Amplio occurred against the backdrop of a unique political context. Historically, Uruguay has maintained a comprehensive social welfare system, which the Frente Amplio sought to strengthen. Uruguay’s sustained commitment to social welfare is evident in measures such as the United Nations Human Development Index, which ranks Uruguay in its top tier of countries with very high levels of human development (United Nations Development Programme 2020). Frente Amplio deepened this commitment with a series of measures. Between 2000 and 2010, enrollment in childcare services for children under three rose sharply as reforms sought to increase the universality of social welfare services (Martínez Franzoni and Sánchez-Ancochea 2016). As we noted in the introduction, when police reforms and public security policies are supported by cross-cutting investments in social welfare, it increases their viability and potential for success. Elements of Police Reform The changes and transformations of the PNU signal the depth of the reform process. Although these reforms have not always been well organized and implemented coherently, structural and budgetary changes accompanied them and have transformed the nature of police work. There have been significant increases in the size and territorial distribution of police personnel. This has been accompanied by significant budgetary changes, not only in the total amount of investment destined for the sector but also in levels of oversight. These budgetary realignments have allowed for changes in police strategies and the use of technology in the daily work of the police and the civil servants in the Ministry of the Interior. While the Uruguayan experience of police reform has not always followed a clear road map and structured planning, this method of reform could arguably allow for more flexibility in implementation and the permanent reconfiguration of priorities. The iterative process by which changes were made to the Uruguayan police institution resulted in synergistic efforts, even when they were not intentionally planned. In 2015, the Organic Police Law (Law 19.315) was approved, taking effect on January 1, 2016. It marked a substantial modernization for the PNU. It replaced the laws of 1972, which were implemented during a time of institutional instability just before the military coup, marked by militarized forms of discipline and organization. Several democratic governments had noted the limitations of the 1972 law, but it took time to reach political consensus for its eventual reform. The Organic Police Law defined the

Uruguay

197

PNU as an armed body of a civil and professional nature, dependent on the Executive Branch, which according to the constitution is responsible for the “preservation of order and tranquility within” the country (Law 19.315). Its civilian nature is reflected in the functions assigned to it by the legal system, including selecting, educating, and training police personnel; the promotion system; and the disciplinary rules that govern the police. The civilian functions highlight the hierarchical and disciplined nature of the PNU. The 2015 Organic Police Law defines the police as career professionals, invested with public authority when executing their duties. To fulfill this role, officers are subject to the police statute and must receive comprehensive training from the state. The training includes technical training but also encompasses training in the legal, ethical, and human rights aspects of public security. This system recognizes the role that popular representative bodies play in creating more opportunities for broader participation in the programming, development, and teaching of police training courses. Compared with their regional counterparts, the Uruguayan police receive far more comprehensive and holistic training and professional formation. The new law included changes to the code of conduct and disciplinary processes. The 2015 reforms created additional opportunities for non-commissioned officers, allowing them to enter officer ranks to incentivize advancement in a professionalized structure. In a parallel effort, to increase the number of police officers on the streets, the Ministry of the Interior opened the opportunity to hire civilian professionals for administrative tasks in the police stations. In Latin America, there are more than a few militaristic police institutions whose organization is divided between two separate classes with different functions, missions, levels of preparation, and training. This typically leads to siloed and exclusive work environments. Uruguay’s reforms could serve as a model of a positive practice that enhances the capacity for change in other national contexts. These transformations of the police’s institutional structure and training created opportunities for innovation in policing strategies. To overhaul policing strategies, the Uruguayan government had to transform an old practice (still present in other countries in the region) that allows police officers to work in the private security sector during their rest hours. The so-called 222 program allowed this alternative time commitment, which had a negative effect on police efficacy because working in the private security sector can diminish officers’ ability to focus on security issues affecting the public as a whole and can create conflicting loyalties if police are tempted to prioritize the needs of private actors. In 2010, budget adjustments reduced the possibilities for police overtime in the private sector and offset the financial impact of this change to individual officers by increasing police salaries. Although this change allowed the police to concentrate on public security tasks, it did raise another complication, as private security forces

198

Making Police Reform Matter in Latin America

filled the vacancies left by off-duty police, and the private security industry in Uruguay is only partially regulated. The Uruguayan police is the only Latin American force that allows unionization, and the union originally opposed some of these changes. Despite police opposition, changes to the 222 program were largely sustained. In 2016, a hybrid program was created where some police officers were designated to 222 services and trained accordingly. The achievements and challenges of this change are not yet clear, as empirical analyses have not assessed whether it generated improvements in police effectiveness and dedication.1 Operational Strategies: From Random Patrol to Evidence-Based Practices In the early 1990s and through the early 2000s, the primary concerns of the police centered on increased presence, infrastructure improvements, and better salaries. Although there are still areas for improvement, it is clear that Uruguay has made progress. These changes made it possible to advance the design and implementation of new police strategies, especially those linked to crime prevention in public spaces. Two types of strategies have received special attention: community policing and problem-oriented policing (POP). In 2008, Uruguay created a specialization in community policing. The community police were envisioned as a body that would intervene in relationships in local communities, seeking cooperation and conflict resolution. The community police distinguished themselves from other types of police with their efforts to maintain low levels of staff turnover, as stability in community assignments strengthened police ties to community members and fostered higher levels of trust in police. Due to positive initial results, in 2011 the police consolidated these efforts into the National Community Police Program, which sought to unify work criteria and better define roles. The Uruguayan community policing initiative has had positive results, but its efforts were led almost exclusively by officers who specialized in community-oriented work. One of the great difficulties in implementing community policing strategies more broadly is that they are typically carried out by a specialized group of officers, rather than envisioning these strategies as a cross-cutting model for police work as a whole. This limited widespread implementation.2 Subsequently, and with the support of various international experts, Uruguay launched its POP strategy, which emphasizes the use of science and evidence for the design of police strategies. POP strategies aim to identify a specific problem, analyze it thoroughly and empirically, and ultimately develop a tailored response to the specific problem at hand. POP approaches stress the importance of empirically and periodically assessing the degree to which the response addresses the problem and readjusting the response when needed.

Uruguay

199

Uruguay’s POP program is based on environmental theories of criminology, which focus on the situational determinants of crime to reduce its occurrence (Clarke 1995). Rather than focus solely on the perpetrators of criminal acts and responding to individual incidents, environmental theories of criminology focus on the “triangle of crime” to understand trends and patterns and ultimately interrupt them with proactive and preventive measures. The triangle of crime is composed of three parts: 1. the perpetrator who commits the crime 2. the crime victims (this can include places, spaces, and objects) 3. the guardians who engage in surveillance (this does not necessarily refer solely to police or private security, as it can include community members) In 2012, Uruguay launched a pilot POP program in Montevideo, with assistance from a $5 million loan from the Inter-American Development Bank. The program unfolded in three police districts that had consistently reported high rates of crime and violence, and it targeted the problem of juvenile delinquency. The program aimed to improve the police’s clearance rate and rehabilitate high-risk juvenile offenders. Evaluations of POP initiatives showed some positive results in crime reduction, but in most cases the police faced challenges in implementing the strategies and sustaining the model over time (Eck and Gallager 2016). When we examine the implementation of the POP in Montevideo, for example, the most obvious obstacles to its implementation were the difficulty in understanding its objectives, lack of leadership, police resistance to change, and the lack of long-term financial resources (Del Castillo 2018). The Montevideo experience with POP was not unique. As police around the world have experimented with POP strategies, multiple questions have emerged regarding the applicability and sustainability of this model given the complexities involved in implementing an evidence-based approach in institutions that have not traditionally relied on scientific approaches to public security. However, POP can be a useful tool when it is fully implemented because it offers a targeted method for solving crime and other problems of public disorder and because the benefits of this model could spread beyond the community (Braga, Papachristos, and Hureau 2014). As is the case around the world, in Uruguay most crimes are concentrated in a few places. Some criminologists have argued that public security initiatives should focus on smaller places as a unit of crime analysis (Gill, Weisburd, and Telep 2014). By concentrating on these “hot spots,” policing strategies can target the most problematic areas (Sherman, Gartin, and Buerger 1989). If crimes are not random, but concentrated in certain places for extended periods of time, it makes sense to focus public security initiatives on

200

Making Police Reform Matter in Latin America

a small number of places and ultimately address a large proportion of crimes. A focus on hot spots makes it possible to optimize scarce resources, particularly police personnel, in specific places to reduce crime. The traditional geographic spaces used by the police such as quadrants, blocks, or radiuses are much larger in size than a hot spot. A strategy focused on hot spots should ideally involve specific addresses, buildings, or areas that are only a few meters long. Evaluations of hot spot interventions show modest reductions in the occurrence of crime, but it generates the diffusion of crime control benefits in the areas immediately surrounding the intervention area (Braga, Papachristos, and Hureau 2014). In fact, various studies have emphasized the need to move to a police strategy that is mostly based on hot spots (Braga and Weisburd 2010). Civilian and police leaders in Uruguay absorbed this criminological debate and incorporated these ideas into their reform process. In 2013, Uruguay embarked on a territorial redefinition of police work, coupled with a specialized criminal analysis unit in 2016. This ultimately led to a 2016 pilot program on problem-based strategies, called the Programa de Alta Dedicación Operativa (PADO). The PADO focused on patrolling in areas of high crime concentration. Its implementation encouraged the strengthening of the police’s information collection system and the use of technology to better deploy police presence and patrol routes. In line with Uruguay’s earlier emphasis on police preparation, officers received thorough training on these new strategies. PADO focused on decreasing the prevalence of violent robberies in targeted territories, and evaluations showed promising results (El Observador 2017). In fact, because of some achievements in crime reduction, the government announced its plans to expand the program in 2018 and increase the number of police working in it.3 It remains too early to draw firm conclusions about the success of the program, and some problems with implementation have persisted. For example, the program does not have a clear mechanism for managing institutional memory, which affects the construction of knowledge and the training of police officers and analysts. Cooperation with various state institutions is not clearly defined, and protocols for established tasks can be vague. Power is concentrated in the officials who oversee the program, leading to high levels of discretion. With the change of government, the PADO was transformed. In March 2020, the new minister of the Interior, together with the president, confirmed that the main objective of the government was to have more visible police on the street. With this objective, the program grants police department chiefs the power to define how and when to deploy personnel, without necessarily relying on the data collected by the information systems. Given the short amount of time the new government has been in power, it is too early to see how much a change in party control will translate into a change in policing practices on the ground.

Uruguay

201

Crime Trends and Public Fear of Crime Until recently, Uruguay’s rates of violent and nonviolent crime have been among the lowest in the region. During the late twentieth and early twentyfirst century, homicide rates remained very low, fluctuating from 5.8 to 7.6 per 100,000 people.4 As Figure 9.1 illustrates, in 2012, homicide rates climbed to 7.9 and steadily began to increase until they reached historical highs in 2018, at 12.2 per 100,000. Though still well below the Latin American regional average, homicide rates in Uruguay in the twenty-first century have risen by 85 percent. This is a sizable increase in a relatively short time, and these kinds of sharp changes in a state’s insecurity context tend to engender a much stronger response from citizens than more gradual increases in long-standing trends (Hale 1996). Indeed, the Uruguayan experience with violent crime is very similar to that of Costa Rica’s (see Chapter 5), where citizens are not typically reassured by the fact that their county registers low homicide rates for the region but react with alarm to the fact that the contemporary status quo is very different from what people were accustomed to in the 1980s and 1990s.

Figure 9.1

Homicide Rates in Uruguay (2005–2020)

Source: World Bank, World Development Indicators.

202

Making Police Reform Matter in Latin America

Behind these national averages, homicide rates tend to be concentrated geographically, as some neighborhoods register higher levels of homicides than others. In 2018, three districts of Montevideo had a homicide rate of more than 30 cases per 100,000 inhabitants. Besides homicides, citizens’ main concern tends to be violent robbery (known as rapiñas).5 As Figure 9.2 illustrates, rates of reported violent robberies in 2018 were almost nine times as high as those of 1989. In fact, during 2018, reported rapiñas reached 29,904 cases (actual and attempted) with high territorial concentration in Montevideo (23,495 cases) and an increasing use of guns (19,754 cases). Although typically it can be difficult to rely on official crime statistics because of disparities in reporting levels, in this case, these data are valid and reliable because rates of crime-reporting in Uruguay tend to be high (Boidi and Queirolo 2008). Generally speaking, rates of reporting armed robbery tend to be more reliable than those of other types of crimes, as people will often report these crimes to satisfy requirements of insurance agencies (Malone 2014). Self-reported victimization rates confirm these trends. When LAPOP asked survey respondents whether they had been victimized by crime in 2006, 21.6 percent replied that they had. That percentage increased to 29.0 percent in 2019. In 2019, average rates of self-reported victimization were higher than the regional average, belying Uruguay’s reputation as one of the safest countries in the region. Indeed, although the José Mujica government had argued that legalizing marijuana would help reduce crime (Queirolo et al. 2018), legalization did not bring about the desired result, at least in the short run. Since the legalization of marijuana in 2013 and its widespread availability through pharmacy sales starting in 2017, rates of homicides, rapiñas, and self-reported victimization have notably increased. We do not argue that marijuana legalization caused crime to increase, as the available data do not allow us to substantiate causality. However, it is clear that the legalization and state regulation of the marijuana market did not coincide with a decrease in crime rates. The 2017 National Victimization Survey of Uruguay documents similar levels of victimization: 24 percent of the adult population.6 In addition, this survey measured the number of households indicating that at least one member had been victimized by crime in the past twelve months, which reached a total of 26 percent of respondents in 2016. In Montevideo that number rose to 32 percent, whereas in the rest of the country around 20 percent of households had at least one person who was a victim of a crime. Data on perceptions of insecurity are slightly mixed. For example, Uruguay’s 2017 National Victimization Survey found that 48 percent of respondents indicated that they thought Uruguay was a safe or very safe country, and 62 percent perceived their neighborhoods as either safe or very

Reported Armed Robberies in Uruguay (1989–2019)

Source: Ministry of the Interior, Uruguay, 2019.

Figure 9.2

203

204

Making Police Reform Matter in Latin America

safe. Still, 21.6 percent of respondents identified insecurity as the most pressing problem facing the country, and 45 percent thought the country’s security could deteriorate in the next two years. A longitudinal analysis of LAPOP data finds that the percentage of the population that felt insecure in their neighborhood remained fairly constant and in 2019 was on par with the regional average (Figure 9.3). Such perceptions of insecurity reflect the increase in violent crimes such as homicide and armed robbery, as well as higher levels of self-reported victimization more broadly. Uruguay’s reputation as one of the safest countries in the region is at odds with these heightened levels of insecurity. Despite the sustained efforts of the Frente Amplio to address the socioeconomic root causes of public insecurity and criminality, it is clear that crime and public fear of crime have increased in recent years. One cannot attribute this increase to government policies, as the causes of crime are multidimensional. However, one could see how these trends, particularly the increase in the violent crimes of rapiñas and homicides, opened the door to reexamining public security policies under a new government.

Figure 9.3

Fear of Crime in Uruguay (2006–2019)

Source: LAPOP’s 2006–2018/19 AmericasBarometer.

Uruguay

205

Perceptions of Police Against this backdrop of community-oriented police reforms and an uptick in criminality, we turn to examine public attitudes toward the police. Although internally the debates on the quality and effectiveness of the police institution have become one of the main axes of political confrontation, in the regional perspective Uruguayans tend to hold favorable perceptions of the police. As we noted in Chapter 2, Uruguay ranked first in average levels of trust in police in the region in the 2019 wave of the AmericasBarometer and registered one of the highest levels in 2017 as well (Malone and Dammert 2020). As Figure 9.4 illustrates, trust in the police has consistently been in the positive range of the scale. From 2010 to 2014, trust in the police was significantly higher than trust in other political institutions, on average. While trust in police has been relatively stable, a small, gradual decline began in 2014 and continued through 2019. In 2014, trust in police dipped lower than average levels of trust in other domestic political institutions and

Figure 9.4

Trust in Police in Uruguay (2006–2019)

Source: LAPOP’s 2006–2018/19 AmericasBarometer.

206

Making Police Reform Matter in Latin America

remained lower through 2019. This decline corresponds to higher levels of homicide, armed robbery, and overall self-reported victimization. The 2017 National Victimization Survey of Uruguay corroborates these moderately strong levels of trust in police, as 38 percent of interviewees were satisfied with police performance. In addition, 60 percent believed that the police strive to fight crime, and 60 percent have a lot of or some confidence in police action. These positive attitudes toward police could be linked partially to their treatment of crime victims. In the national victimization survey, 41.3 percent of those who had made a police report said that they were satisfied with how the police had worked their case and 70.7 percent acknowledged that they received good treatment at the hands of police. Another reason that trust in police remains strong could be attributed to low levels of bribery. As Figure 9.5 demonstrates, reported levels of police bribery are extraordinarily low in Uruguay, ranking far below the regional average. After a slight increase in 2012, rates of self-reported bribe solicitation declined in Uruguay, to the point that in 2019, these rates were even lower than their 2006 levels, as only 2 percent of respondents reported that police solicited a bribe from them. Although we cannot address the issue of causality, we find that reported levels of police corruption are extraordinarily low in Uruguay, and this is linked to public trust in police; respondents who indicated that police had solicited a bribe from them reported significantly lower levels of trust in police each year (r = −0.107, p