The Politics of Everyday Crime in Africa

The Politics of Everyday Crime in Africa Insecurity, Victimization and Non-State Security Providers

Danielle C. Kushner

The Politics of Everyday Crime in Africa

Danielle C. Kushner

The Politics of Everyday Crime in Africa Insecurity, Victimization and Non-State Security Providers

Danielle C. Kushner Political Science St. Mary’s College of Maryland Maryland, MD, USA

ISBN 978-3-319-98094-2    ISBN 978-3-319-98095-9 (eBook) https://doi.org/10.1007/978-3-319-98095-9 Library of Congress Control Number: 2018962980 © The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland AG 2019 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover pattern © Harvey Loake This Palgrave Pivot imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

The genesis of my thinking on non-state security is rooted in my experiences coming of age in tenement housing in Baltimore city, one of the most violent cities in the United States. When I was about ten years old, my young mother, siblings, and I moved to the second floor of a 13-story, public high-rise building in the heart of east Baltimore. Lafayette projects, “the projects” as they were known, comprised several high-rise public housing buildings and some townhomes (known as low-rises) in a single complex. From the moment we entered this new home of ours, it was easy to see that the projects were a breeding ground for drugs, guns, and violence. In this new home, illegal drug transactions taking place in open-air drug markets was the norm, and it was not uncommon to see somebody being assaulted for either attempting to steal drugs or “messing up” a drug dealer’s money. In fact, the insecure reality of the projects was so palpable that it manifested itself in everyday salutations such as “be safe” or “watch yourself”. This environment was new to our young family who moved from our grandparents’ relatively comfortable single-family home. Thus, my mother naturally worried about the safety of her children as she looked for ways to cope with this new atmosphere. Initially, we coped by taking the number 15 bus to the houses of other family members and friends on a daily basis. There were even some days when there was no money for the bus, and on those days, we walked. Within a couple of months, however, we stopped this daily exodus out of the projects. Eventually, it became apparent to my mother that people did actually find a way to live in that environment. In time, our family v

vi 

PREFACE

learned to live there too. We learned the unspoken codes surrounding crime and violence. We learned that if we had a problem we could call the police, but at best they would come hours later and, at worst, not at all. We learned that it was possible to survive with criminal activity present at most corners; it was just a matter of figuring out how. Although I did not know it then, my experiences growing up in the projects provided my initial insights into how ordinary people make use of informal networks to navigate insecure locales, especially when the state is perceived as unreliable and/or apathetic. Baltimore city is no Africa, or even South Africa, for that matter, but there are certain parallels to be drawn between the realities I describe above and the everyday experiences of crime that many Africans face. For example, many Africans do not trust institutions of the state, and that distrust is often especially pronounced with law enforcement agencies. Africans’ experiences of authoritarian-style policing and corruption have led many to feel that they must rely heavily and sometimes exclusively on actors outside of the state for security and protection. Like many Baltimoreans today, many South Africans also see crime and security issues as one of the most important problems they face. The first time I visited South Africa was in 2003 and the last time was in 2012. Each of those times, and every time in between that I visited the country, people’s discussions of both their perceptions of crime and even their own experiences of victimization were just par for the course. Because everybody had a victimization story, or so it seemed, these types of stories were not considered extraordinary or remarkable in any way, but rather a normal part of everyday conversation. The only thing that has changed about these stories over the last decade that I’ve been visiting the country is the level of blame that ordinary South Africans seem willing to assign the government for crime. When I first visited the country in 2003, nine years after the end of apartheid, people talked about crime as much as they did in 2012, but most people stressed the need to be patient with the government and give it time to improve the country. Years later in 2012, sentiment toward government was much different. Many people expressed disillusionment and disappointment with the ruling ANC party. And yet for all the dissatisfaction so many South Africans feel with the ruling party on a number of performance measures, it still garners an overwhelming majority of votes. South Africa therefore presents an interesting puzzle, one which is the subject of this book, which is how citizens continue to see the state as

 PREFACE 

vii

legitimate when it fails to perform to their satisfaction and meet their basic need for security. In this book I argue that the state in South Africa, to some extent, must thank non-state security providers for its continued legitimacy. Maryland, MD

Danielle C. Kushner

Acknowledgments

This book would not have been possible without the assistance, support, and guidance of many people. First, I’d like to thank the people of South Africa. In order to write this book, I needed to spend many hours talking to ordinary South Africans about their experiences and perceptions of crime, their views on state institutions and government leaders, and their insights into how people on the ground manage to secure themselves when state security proves insufficient. I thank them for patiently sitting through my interviews, entertaining my questions, and generously giving of their time. During my time in South Africa, I was fortunate to have interviews with a number of organizations and experts who work on issues of crime and security in this context. I’d like to thank all of the men and women of the South African Police Service (SAPS) who took the time to talk to me about their work. I’d also like to thank individuals of various civil society organizations, including members of Community Police Forums (CPFs), street committees, neighborhood watch groups, and Business Against Crime (BAC). Interviews with these groups were instrumental in shaping my thinking about non-state security in South Africa. Finally, I’d like to thank the elected officials, Members of Parliament, and Local Government officials alike, who devoted time to discuss governmental perspectives on the issues I investigate here. Next, I’d like to thank my adviser and dissertation chair, Michael Bratton. There are so many ways in which he helped this book come into existence. He read countless drafts of dissertation chapters (upon which this book is based) and each time provided thoughtful and chalix

x 

ACKNOWLEDGMENTS

lenging feedback, often within just two days of receiving the draft. Mike also connected me to numerous networks within South Africa that would prove absolutely vital for successful fieldwork. And then, of course, there is the model of excellence in teaching and research that I had the privilege of observing for seven years at Michigan State. Mike’s commitment to furthering our knowledge of Africa, and his devotion to his students, colleagues, and department, is truly inspirational. As I think about the wonderful gift that Mike and the other founders have given to us in Afrobarometer, I must take time here to thank Carolyn Logan. I learned a lot from Carolyn in the years that I worked for Afrobarometer, and I thank her for sharing her wisdom, knowledge, and support over all those years. In addition to Mike, I’d also like to thank the other members of my dissertation committee: Richard Hula, Jeff Conroy-Krutz, and Mahesh Nalla. Throughout the writing of the dissertation, Rick pushed me to be more cognizant of the policy implications of the work, and Jeff provided general feedback on several areas of the work, including theory, methods, and analyses. I thank both of them for making the research stronger. Mahesh graciously agreed to serve as the external reviewer on my committee. The project greatly benefited from Mahesh’s expertise in criminology, and I’m sure that learning about the ways that criminologists approach the study of crime and security has added a level of depth to the project that would have otherwise not existed. Finally, although they were not official members of my committee, two scholars whom I greatly admire and respect have provided feedback on my work, mentorship, and encouragement. Therefore, a huge thank-you to Adrienne Lebas and Lauren MacLean is in order. While in South Africa, there were a number of institutions and individuals who went above and beyond the call of duty to help me with my research. I’d like to extend my deepest gratitude to Bob Mattes and the Democracy in Africa Research Unit (DARU) at the University of Cape Town (UCT), Clifford Shearing and the Centre for Criminology at the UCT, and the Centre for Civil Society at the University of KwaZulu-Natal in Durban. All three units provided office space, access to resources, and an atmosphere of collegiality during my fieldwork, for which I will be eternally grateful. I could not have made it through the writing of my dissertation, and, by extension, the publication of this book, without my dissertation working group. I owe a debt of gratitude to Helen Lee, Seoyoun Choi, and

 ACKNOWLEDGMENTS 

xi

Nicholas Kerr for feedback on drafts of chapters, encouragement, and friendship. I’d also like to thank Olufunmbi Elemo and Twyla Blackmond Larnell for helpful discussions about my research and friendship over the years. Last, but not least, I’d like to thank my family and close friends. People say it takes a village to write a book (okay, maybe that’s just my saying). The following people are all a part of my village. I have to start with my husband, Jeff Kushner, who supports me in every meaningful way imaginable. I thank him for being my partner, my friend, my cheerleader, my research assistant, my proofreader, and any and everything else I have called on him to be while I am conducting research and writing. I could not ask for a better spouse, and I thank him for his constant support, for his encouragement, and for all the sacrifices he endures so that I can do work that is meaningful to me. Here I will also mention my father-in-law, Gary Kushner, and mother-in-law, Lila Pride, because they are the best parents-in-law anyone could ask for and because I appreciate how well they raised the man who would eventually become my husband. I thank my mom, Felicia Wilson-Brown, and grandmother, Edith Wilson, for years of love and support. I also thank my stepmother, Synthia Carter, for her continued support and encouragement. My two sisters, Syrita and Asia, have always wanted the absolute best for me, and they speak proudly of my accomplishments to anyone who will listen. I love and appreciate them dearly. I would also like to mention my late father, Daniel Carter, Jr., and my late brother, Isaiah Simmons, III. Two of the most important men in my life did not live to see the publication of this book, but they live on in my memory forever. Placing their names here in this book is one of many ways that I try to ensure that their names live on forever. I receive an amazing level of support from my pastor, Bishop Clifford M. Johnson, III, and my broader church family at Mount Pleasant Church and Ministries, and I thank them for never-ending guidance and encouragement. Last, but certainly not least, I have seven nieces, Kyia, Najai, Nyla, Molly, Kenzie, Camouri, and Reyna, one nephew, RJ, and the best puppy anyone could ask for, Pepsi. All of these small beings bring me such indescribable joy. I thank them for being a part of my life and for continually renewing my motivation to leave the world in better shape than I entered it.

Contents

1 Introduction  1 1.1 The Research Questions  1 1.2 Why Study the Political Consequences of Non-state Security?  3 1.3 Africa: Non-state Provision in Weak States  7 1.4 Theorizing the Relationship Between Non-­state Security and Perceptions of State Legitimacy  8 1.5 The Impact of Personal Insecurity and Victimization on Popular Attitudes and Participation 14 1.6 Outline of the Book 14 References 17 2 Non-state Security in South Africa: Historical Roots, Contemporary Realities 21 2.1 Introduction 22 2.2 Crime and Security Under Apartheid 22 2.3 Crime and Security Since the 1994 Transition 27 2.4 State Responses to Crime in Post-apartheid South Africa 39 2.5 Non-state Security in Post-apartheid South Africa 42 References 51 3 The Impact of Non-state Security on Perceptions of State Legitimacy in South Africa 55 3.1 The Puzzle 55 3.2 Key Hypothesis 58 3.3 Data and Methods 60 xiii

xiv 

Contents

3.4 Measuring State Legitimacy, Reliance on Non-­state Security, and State Arranger 61 3.5 Controls 63 3.6 Explaining State Legitimacy 63 3.7 Testing the Effect of Non-state Security Reliance on Perceptions of State Legitimacy Using Afrobarometer Data 67 3.8 Conclusion 71 References 71 4 Assessing the Impact of Non-state Security, Victimization, and Insecurity on Social Capital and Collective Action in South Africa 73 4.1 Introduction 73 4.2 Existing Works Exploring the Relationship Between Service Provision and Political Participation 74 4.3 Security and Popular Participation 78 4.4 Victimization and Political Participation 79 4.5 Theorizing the Relationship Between Non-­state Security and Participation 81 4.6 Data and Indicators 83 4.7 Determinants of Political Participation 84 4.8 Explaining Security, Joining, and Collective Action in South Africa 86 4.9 Conclusion 89 References 89 5 The Impact of Everyday Crime and Security on Protest Behavior in South Africa 91 5.1 Introduction 91 5.2 Protest in Africa 91 5.3 Data and Indicators 93 5.4 A Security-Based Theory of Protest and Protest Hypotheses 95 5.5 Results 96 5.6 Conclusion102 References102

 Contents 

xv

6 Explaining the Impact of Non-state Security, Victimization, and Insecurity on Voting in South Africa105 6.1 Introduction105 6.2 Existing Literature on Voting in Africa106 6.3 Data and Indicators107 6.4 Voting Hypotheses110 6.5 Results110 6.6 Conclusion111 References113 7 Conclusion117 7.1 Introduction117 7.2 Summary of Key Findings118 7.3 Generalizability of the Findings119 7.4 Micro-Level Security Experiences and Perceptions of State Legitimacy in Africa119 7.5 Micro-Level Security Experiences and Political Participation in Africa122 7.6 Summarizing the Relationship Between Security, Legitimacy, and Political Participation in Africa124 7.7 Contributions of the Research124 7.8 Limitations of the Research126 7.9 Future Directions127 References127 Appendix A: Question Wording and Codes for Table 3.1129  Appendix B: Question Wording and Codes for Tables 3.2 and 7.1135  Appendix C: Question Wording and Codes for Tables 4.1, 5.1, 5.2, 6.1, and 7.2141 Index147

List of Figures

Fig. 1.1 Fig. 1.2 Fig. 1.3 Fig. 2.1 Fig. 2.2 Fig. 2.3 Fig. 2.4 Fig. 2.5 Fig. 2.6 Fig. 2.7 Fig. 2.8 Fig. 3.1 Fig. 3.2 Fig. 3.3 Fig. 3.4 Fig. 5.1

Conventional view of the relationship between capacity and legitimacy (state as producer) 9 An alternate path from weak state capacity to legitimacy 10 The effect of non-state security on state legitimacy in low capacity states 10 Security reliance post victimization, 2010–2012 29 Type of non-state security reliance, 2010–2012 30 Perceptions of crime in South Africa, 2003–2016/2017 36 Perceptions of personal security, 1998–2017 37 Victimization in South Africa, 1998–2012 38 Primary reasons for not reporting crime to the police, 2011 39 Growth of the commercial security industry by number of security guards, 2001–2017 44 Growth of the commercial security industry by number of security businesses, 2001–2017 44 Legitimacy of the state in South Africa, 2010–2012 57 Perceptions of crime, 2010–2012 58 Legitimacy of the state in South Africa, 2010–2012 62 Conditional effect of non-state security on perceived state legitimacy65 Predicted values of protest for state and non-state users at varying levels of state legitimacy 101

xvii

List of Tables

Table 2.1 Table 2.2 Table 3.1 Table 3.2 Table 4.1 Table 5.1 Table 5.2 Table 6.1 Table 7.1 Table 7.2

Type of non-state security reliance by race, 2010–2012 Black reliance on non-state security, 2010–2012 Pooled logistic regression output on perceptions of state legitimacy in urban, South Africa Afrobarometer logistic regression output on perceptions of state legitimacy, South Africa The impact of security variables on joining and collective action in South Africa Regression analyses of protesting, South Africa Analysis of societal non-state security on protesting (with fixed effects), South Africa The impact of security variables on voting in South Africa Logistic regression output on perceptions of state legitimacy, 32 African countries (fixed effects model) Regression analyses of joining, collective action, and voting, 32 African countries (fixed effects model)

31 31 64 69 87 97 100 112 120 123

xix

CHAPTER 1

Introduction

Abstract  This chapter presents an overview of the main research questions raised throughout the book. The book assesses how individuals’ everyday crime and security experiences affect their political participation and their attitudes toward the state. This chapter also introduces the reader to the theoretical framework used to explain the puzzle of high state legitimacy perceptions in states that have weak state security capacity. Keywords  Non-state security • Non-state provision • Weak states • State legitimacy • Africa

1.1   The Research Questions Upon clenching the South African Presidency in April 2009, now ousted head of state Jacob Zuma placed crime at the center of his policy agenda and immediately called for the resurrection of street committees.1 Initially developed during apartheid to provide a measure of protection for black

 “ANC to Focus on Crime and Education” says Zuma, Mail & Guardian Online, September 18, 2008. Street committees were largely developed by blacks under apartheid to provide a measure of protection and order at the local level. Because the South African state rarely provided security for blacks, street committees were seen as a legitimate local authority for dealing with both everyday crime and political violence that was unleashed on black communities by the apartheid regime and elements within the liberation movement. 1

© The Author(s) 2019 D. C. Kushner, The Politics of Everyday Crime in Africa, https://doi.org/10.1007/978-3-319-98095-9_1

1

2 

D. C. KUSHNER

communities, street committees were described as South Africa’s “best formula to fight crime” by the new head of state. Some interpreted Zuma’s call to restore street committees as a signal that the state continued to lack the capacity necessary to effectively decrease crime and, perhaps more importantly, assuage public fears about crime in the country. Others interpreted the new chief executive’s call as a move by the state to capitalize off of the legitimacy of these local-level non-state security actors. This project takes a fresh look at the relationship between legitimacy and capacity and, in so doing, helps to shed new light on a puzzle that persists within African politics, namely, how states remain legitimate in the eyes of their citizenry, when they are weak and underperforming. Importantly, the new pieces to this puzzle that are presented here make no reference to conventional explanations of ethnicity and patronage; rather, I argue that non-state actors attenuate the impact of weak state capacity on state legitimacy, under certain conditions. While there is no consensus on the intent and implication of Zuma’s plea for street committees to get involved in crime fighting, it is clear that his invitation brought renewed national attention to the struggle the country continues to face with generating a widespread sense of personal safety and security. It also renewed public discourse on the persistence of state and societal reliance on non-state sources of security in the post-­ apartheid era. South Africa was successful in ushering in multi-party democracy in 1994 with the end of apartheid; however, the transition has not translated into greater levels of security; state institutions and wealthy individuals continue to contract with private security companies for protection, while the poor continue to rely on informal codes, conventions, and networks as a means to navigating insecure locales. This book interrogates two questions related to everyday crime and security. The first explores the political consequences of non-state security provision. Specifically, it examines how citizens’ reliance on non-state security actors (such as commercial security companies, neighborhood watch groups, and community police forums) affects their perceptions of state legitimacy. In the context of South Africa, the key case that I examine in this book, 52% of survey respondents report turning to actors outside of state police for security.2 The second question looks more broadly at 2  This figure comes from a survey I administered during fieldwork in South Africa from October 2011–July 2012. The survey was conducted in Cape Town, Durban, and Johannesburg with a total of 432 South African citizens.

 INTRODUCTION 

3

how citizens’ everyday security experiences shape their political attitudes and behavior. It assesses how feelings of personal insecurity and experiences of victimization weigh on individuals’ political views and participation. Both of these questions contribute to an important research agenda on non-state provision and the political significance of everyday crime, with important implications for state–society relations. Across Africa, many ordinary citizens rely on non-state security provision.3 Non-state security structures such as street committees, neighborhood watches, and commercial security firms are seen as essential forces in combating crime and violence in this context. Yet for all the many calls for greater non-state provision of security, little is known about the consequences of the non-state provision of this key good.

1.2   Why Study the Political Consequences of Non-state Security? 1.2.1  The Philosophical Importance of Security Provision From a political philosophy standpoint, social contract theorists such as Hobbes and Locke argue that states exist, by and large, to provide security. These theorists, and empirical political scientists alike, argue that the state’s provision of security provides the raison d’être of states, and therefore is critical to cementing the social contract between state and society. For empirical scholars who assess how citizens’ performance evaluations affect their perceptions of the state, the key question raised here is generally how the state’s ability (or lack thereof ) to provide adequate protections for individuals shapes their attitudes toward the state (Bratton and  Chang 2006; Rotberg 2003; Weingast 1997; Wood et al. 2006). Essentially, the emphasis is a state-centered one that focuses on whether and how well the state is able to provide this good. Yet, increasingly, citizens in both the developed and developing world extensively rely on non-­ state providers to meet their security needs. This raises a new and important

3  Based on 2008 Afrobarometer survey data, Kushner and MacLean (2015) show that a substantial percentage of Africans report that non-state actors such as traditional leaders and community members assume primary responsibility for the provision of public goods. For example, close to one-third of Malawians and one-quarter of Tanzanians report that nonstate actors primarily provide education.

4 

D. C. KUSHNER

question, one that focuses on the consequences of who provides. In a world where non-state actors from commercial firms to vigilante groups participate heavily in the provision of security, the question of who provides is as essentially political as the question of the quality of state-provided services. This book therefore probes the behavioral and attitudinal consequences of non-state security provision. 1.2.2  Public Goods and Political Legitimacy There are a number of different ways that the state may boost perceptions of its legitimacy among the citizenry. Several scholars have distinguished between legitimacy or supportive attitudes that come from the provision of material goods, to ones that are derived from the way in which rules and procedures are upheld within a given society (Bratton and Mattes 2001; Diamond and Morlino 2004). Those who focus on procedurally based sources of legitimacy follow the path of Weber, emphasizing rational-legal notions of legitimacy and the importance of rule-based behavior. These works suggest that the development of rules that are perceived to be fair, as well as the fair and equal application of such rules, will lead to legitimating attitudes among the citizenry (Tyler 2003; Levi et al. 2009). Many works that examine procedurally based support of political institutions have focused on democracy, elections, and institutions of the state such as police forces and courts (Gibson and Caldeira 2003; Jackman 1993; Mondak 1993; Tyler and Fagan 2010). Citizens’ faith in the state’s legitimacy may also come about from symbolic actions undertaken by the state. The purpose of these actions is to create affective linkages with citizens’ that cause them to identify with and lend credence to the political system. These bonds may be developed through processes of socialization and/or through experience with rituals and symbols that reinforce the “rightness” or “appropriateness” of state institutions. The types of affective bonds that matter for political legitimacy are ones like ethnicity and party identification, deeply held attachments that are not contingent on elected officials’ performance. Instead, this dimension of legitimacy is more about what the political institutions represent for individuals, symbolically (Easton 1957). When considering this dimension, it becomes possible to see how individuals may continue to see that state as legitimate, even when they are dissatisfied with the provision of material goods and unhappy with how political processes unfold.

 INTRODUCTION 

5

More frequently, political scientists study the role that state-provided public goods and services play in boosting citizens’ support for the regime, government, and individual political leaders. We generally refer to this as the instrumental or performance-based dimension of legitimacy. The conventional wisdom has been that where the state is able to effectively provide public goods, its citizens will be more likely to see it as legitimate. This logic has been demonstrated in the literature on economic voting and voter turnout (Anderson 2000; Lewis-Beck and Paldam 2000; Powell and Whitten 1993), and in works that explore the impact of economic and political performance evaluations in spurring political legitimacy (Bratton and Mattes 2001; Bratton et al. 2002; Fernandez and Kuenzi 2009; Gilley 2006; Levi et al. 2009; Peltier 2007; Schaar 1981; Schatzberg 2001; Tyler and Fagan 2010; Wood et al. 2006). Thus, we have solid theoretical propositions and empirical support for the positive way that citizens ­ respond to political elites when they can, in fact, provide the goods. But what about when those providing the public goods are not state agents, but non-state actors? Scholars and donor agencies alike have come to recognize the important role that non-state actors play in the provision of basic goods and services throughout the world. Particularly in the developing world, non-­state actors such as corporations, non-profit organizations, community organizations, and faith-based organizations play a key role in ensuring that important goods and services are delivered to the citizenry. Yet, there are very few empirical investigations of the political consequences of non-­ state provision. Important exceptions include the work by Sacks (2012) which explores how donor and NGO provision shapes legitimacy perceptions and the works by Cammett and MacLean (2014) and Brass (2016) which both investigate the implications of non-state provision for the state in Africa and the developing world more generally. This book makes important theoretical and empirical contributions by analyzing the political consequences of non-state security provision. More so than any other good, the provision of security is seen as a key, defining characteristic of the state, and, importantly, one that shapes citizens’ attitudes toward and support for the state (Hobbes 1998; Locke 1966). This implies that attitudes toward the state will suffer when this good is not publicly provided. Yet, it is an empirical question as to whether non-state security will dampen relations between society and the state as implied by extant political science literature. Therefore, while scholars have emphasized the centrality of state-provided security, this book explores the

6 

D. C. KUSHNER

importance of this criterion from the vantage point of ordinary citizens. Do citizens’ assessments of the state suffer when they turn to non-state providers for security? By answering this question, I explore how non-state actors’ service delivery role affects citizens’ support for the state. In this sense, the book allows us to explore whether and how non-state providers may connect citizens back to the state and potentially strengthen state-­ society relations where the state’s capacity to provide security is weak. 1.2.3  Existing Literature on Non-state Security In the field of political science, non-state security provision in Africa is typically discussed at the national level for extraordinary instances of political violence such as civil war, ethnic conflict, and genocide. This literature mostly stems from the subfield of international relations and explores how non-state actors such as mercenaries and militias impact conflict (especially resource-related conflict) in various countries (Higgott et  al. 2000; Krahmann 2005; Meagher 2012). Much of this literature has centered on how the nature of warfare is no longer simply limited to fighting between states, but rather involves a wide range of non-state actors as well. This work is helpful for showing the changing nature of interstate conflict, but it does not tell us anything about everyday crime that citizens experience at the local level. This book argues that there is a need to critically examine citizens’ everyday experiences of crime and violence at the local level, and how these experiences impact their politics. The political consequences of crime become even more important to understand in Africa and other developing regions that are rapidly urbanizing and thus likely to see a rise in violent and non-violent crime in the near future. Aside from political science, much of the existing literature on non-­state security comes from the discipline of criminology. Within this discipline, there is a large body of work that speaks to the significance of a variety of non-state actors in the governance of security (Baker 2007; Bayley and Shearing 1996; Johnston and Shearing 2003; Kempa et al. 1999; Shearing 1992; Wood et al. 2006). This field focuses heavily on the private security industry but also explores other actors including vigilante groups, community police forums, street committees, and neighborhood watches. Within the criminology literature, scholars have tended to focus on the causes of non-state security, which include rising crime rates and fear of crime, the emergence of private property, and the perception that police are unable to manage crime on their own. Importantly, this literature

 INTRODUCTION 

7

demonstrates that policing and security provision is no longer the strict prerogative of the state (Kempa et al. 1999; Bayley and Shearing 1996). More specifically, the theory of “nodal governance” developed by criminologists argues that non-state actors not only participate in the provision of security, but they also help to set the tone and direction of security policy (Burris et al. 2005) While this literature is key to documenting the range of non-state providers that exist and their important role in supplying security for citizens, it stops short of asking the political questions that are of central concern to this work. The theories and insights generated by this literature, however, can be used as starting point for reassessing the political determinants of state legitimacy in Africa.

1.3   Africa: Non-state Provision in Weak States Non-state security provision is not limited to Africa, but it assumes a special importance in this context where the state is often weak and unable to provide basic goods for its citizens (Englebert 2002; Herbst 2000; Jackson and Rosberg 1982). One of the important by-products of the weak state in Africa is that non-state actors frequently fill the gap in state service delivery (Hern 2017). Donor agencies in particular have recognized the steady if not increasing importance of “non-state actors or providers”4 in the provision of goods and services in the developing world, including Africa (Batley and Mcloughlin 2009; Moran and Batley 2004; Pavanello and Darcy 2008; Rose 2007; Sacks 2012). For example, Batley and Mcloughlin (2009) have suggested that “in Nigeria and Malawi, Christian medical missions provide around 60 percent and 37 percent of healthcare services respectively, and in addition there is a myriad of for-profit providers” (15). Thus, non-state actors are key to ensuring that the basic needs of ordinary citizens in Africa and the developing world, more generally, are met. Moreover, it is not simply the weakness of the state in Africa, but also the frailty of African economies coupled with other issues such as endemic corruption and patronage, that lead a substantial number of Africans to rely on non-state actors for a range of goods, including security (Kushner and MacLean 2015). The question that remains unanswered is how this reliance affects the politics of ordinary citizens in this context.

 Moran and Batley (2004) define non-state providers as all providers existing outside the public sector, whether they operate for profit or for philanthropic purposes. 4

8 

D. C. KUSHNER

1.4   Theorizing the Relationship Between Non-­ state Security and Perceptions of State Legitimacy The conventional wisdom suggests that when political goods such as security are not adequately provided by the state, citizens’ support for the state may suffer. In fact, existing literature suggests that the lack of widespread security and a rule of law can lead to state weakness and eventually failure (Rotberg 2003; Wood et al. 2006). But when the state fails to provide a good to citizens’ satisfaction, they may turn to non-state sources, be it the market or society, for the provision of this good. The question then becomes whether individual reliance on non-state security has an impact on citizens’ attitudes toward the state. In particular, I seek to understand whether non-state security reliance strengthens or undercuts citizens’ views of the legitimacy of the state. Drawing on Lipset’s classic definition of legitimacy, I define legitimacy as the capacity of the state to “engender and maintain the belief that existing political institutions are the most appropriate or proper ones for the society” (Lipset 1959, 86). In order to illuminate the effect of non-state security reliance on perceptions of state legitimacy, I distinguish between two roles that the state may play in service delivery; that of producer or arranger. This distinction is borrowed from Savas (2000) where he explains that “the service producer directly performs the work or delivers the service to the consumer [while] … the service arranger assigns the producer to the consumer, or vice versa, or selects the producer who will serve the consumer” (Savas 2000, 64–65). Importantly, Savas suggests that collective goods may be produced and arranged by either the public or private sector (Savas 2000). While many citizens continue to see the provision of security as a government duty,5 the reality is that producers of that good are increasingly non-state agents. Thus, the important question for the purposes of this study is whether the production of security by non-state actors makes for a more or less legitimate state in the eyes of ordinary individuals. Conventional works that stress the importance of state-provided public goods, have conceived of the state’s responsibility in service provision in one, narrow way, that is, the state as a producer. Hence, under these

5  When respondents from my 2010–2012 survey in Cape Town, Johannesburg, and Durban were asked whose responsibility it is to keep people safe, 86% of respondents chose “central government” or “local government” as their response.

 INTRODUCTION 

Low State Security Capacity (Perceived)

Low State Legitimacy (Perceived)

High State Security Capacity (Perceived)

High State Legitimacy (Perceived)

9

Fig. 1.1  Conventional view of the relationship between capacity and legitimacy (state as producer)

conditions, when the state fails to adequately produce security (as we see in many countries where state capacity is weak), the legitimacy of the state is assumed to suffer. This outcome is depicted in Fig. 1.1. Conceiving of the state’s role in such a narrow way means that under conditions of weak state capacity, state illegitimacy is the only logical outcome, as the ability of the state to be perceived as legitimate by the citizenry is tied to its ability to adequately produce security (and other public goods). I suggest that the assertion that the legitimacy of the state will necessarily suffer when the state fails to adequately produce security may be overstated. Therefore, my theory proposes a third way, one in which the legitimacy of the state may remain afloat, even under conditions of weak state security capacity (Fig. 1.2). However, I argue that this third way is only possible when we account for the role of non-state actors in security provision (Fig. 1.3). The theory put forth here suggests that under conditions of weak state capacity, citizens may turn to non-state actors for this provision of security. As depicted in Fig.  1.3, individual reliance on non-state security may strengthen or undercut perceptions of state legitimacy, depending on how individuals see the proper role of the state in service delivery (producer vs. arranger). From this perspective, responsibility for service provision may be subdivided into two categories. In the first instance, the state’s burden of responsibility is heavier. When the state acts as a producer of security, it is directly involved in the day-to-day production of security via an effective public police force. In the second instance, the state is still responsible for security, but acts primarily in a facilitative role, creating the space, regulations, and policies that stipulate whether and how non-state actors may participate in the governance of security. This is the state as an arranger.

10 

D. C. KUSHNER

Low State Security Capacity (Perceived)

Low State Legitimacy (Perceived)

Reliance on NonState Security

High State Security Capacity (Perceived)

High State Legitimacy (Perceived)

Fig. 1.2  An alternate path from weak state capacity to legitimacy

duc

te

Low State Security Capacity

Reliance on Non-state Security

Sta

St

ate

Pro as

as

Ar

ra

ng

er

er

Low State Legitimacy (Perceived)

High State Legitimacy (Perceived)

Fig. 1.3  The effect of non-state security on state legitimacy in low capacity states

In the first instance, when individuals rely on non-state security, but see the state’s proper role as one of a producer, perceptions of state legitimacy suffer. This is the case when citizens see the responsibility for the production of security as largely resting with the state. For citizens who participate in political life and generate the revenue that states use to govern (payment of taxes), they may expect basic goods and services to be directly produced by the state. Thus, when citizens rely on alternate sources to fulfill what are viewed as state responsibilities, citizens’ confidence and belief in the state’s right to rule may suffer. I posit that the gap between citizens’ security demands and the state’s security supply may contribute to views of an illegitimate state. When this good is produced by non-state sources, citizens’ needs may be met, but citizens may come to view market or societal structures as

 INTRODUCTION 

11

competing sources of authority in which they vest their trust and loyalty. In this instance, state production of security provides the raison d’être for states from the perspective of ordinary citizens. Thus, when citizens rely on nonstate sources for the production of this fundamental good, the legitimacy of the state itself may be thrown into question. In the second instance, when individuals rely on non-state security and see the state as an arranger of this good, the state is able to bank legitimacy dividends. In this scenario, citizens may recognize that the state provides the constitutional framework within which non-state security actors are allowed to operate. Without the freedom of association and the legal right to bear arms and secure contracts, non-state groups would not legally be able to provide security. Moreover, citizens may recognize that the state regulates non-state actors in their capacity as security providers. Therefore, when the state is seen as an arranger of security, it may receive credit from their citizens for the moral, technical, and administrative support it provides for these actors. This may particularly be the case if these non-state security actors are viewed as effective in their provision of security. The “third way” that I introduce suggests that non-state actors may play a key role in attenuating the impact of weak state capacity on state legitimacy. Figure 1.3 accounts for the legitimacy outcomes that we should witness depending on whether individuals view the state as an arranger or producer of security. 1.4.1  Functions of the State as an Arranger The role of state as producer is the one that is probably best understood by scholars and citizens alike, since this is how we have traditionally thought of the state’s role in service delivery. In fact, we have often conflated state provision of public goods with state production of them, assuming that if states are adequately providing education, security, and healthcare, for example, then we should correspondingly observe a plethora of high-quality public schools, police forces, and clinics. However, states may ensure that the citizenry is provided for without directly ­producing the goods themselves. Because the state as arranger is a relatively new idea, it is worth explaining what functions the state might perform in this role. When the state is acting as an arranger of security, it might do one of three things. First, the state may create space for non-state actors to participate in the provision of security. This is largely a constitutional act, one

12 

D. C. KUSHNER

in which the state recognizes which entities are legally allowed to provide security. In South Africa, for example, there are constitutional provisions that allow for two main types of non-state security actors. These are the commercial security industry and individual citizens who are envisioned as participating in crime prevention through structured Community Police Forums (CPFs), formal partnerships between the police and community to combat crime. Second, state officials may issue rhetorical or symbolic endorsements of non-state security actors. Beginning with his presidential campaign in 2008,  former South African President Jacob Zuma made not one, but several references to the need to revive street committees, local community-­ based structures (initially developed under apartheid) that organize around issues of crime and violence at the street level. He has made comments such as “We reiterate the call for ANC branches to provide support to law enforcement agencies in the fight against crime, including establishing and strengthening street committees.”6 Many public and state officials in South Africa including provincial premiers, the former national police commissioner Nathi Mthethwa and other members of the South African Police Service have echoed calls for greater community involvement in the fight against crime.7 In fact, some police officials directly attribute reductions in crime to the work of ordinary citizens organized through street committees.8 In these verbal endorsements, state officials often call on ordinary citizens to get involved in crime fighting, primarily by participating in crime prevention activities. However, there is also some evidence that state officials often endorse non-state security actors in non-verbal ways. In particular, it has been suggested that the police will often allow civilian members of community police forums and street committees to harshly “deal” with alleged criminals. A member of the Social Justice Coalition, a security focused NGO based in the Khayelitsha township of Cape Town, South Africa, spoke to this tacit endorsement of violence on behalf of the police. He suggested

6  Africa News. “NGC 2010-Political Report of the President of the ANC Jacob Zuma”. September 22, 2010. 7  Sunday Times. “It’s time for a New Approach”. September 12, 2010. Agence France Presse. “South Africa May Re-Open Specialized Crime Units: Police Minister”. July, 1 2009; Daily News. “All Help is Welcome to Fight Crime: SAP”. November 12, 2009. 8  Cape Argus. “Mitchells Plain ‘Not Saturated With Drugs’ Despite Having Highest Number of Arrests”. September 25, 2009.

 INTRODUCTION 

13

that at times, the police fail to intervene and even allow the community to kill people.9 During my own field research in South Africa, a chairperson of a community police forum admitted that they would use violence to discipline alleged criminals. Members of the community police forum, in turn, are provided with extra protection from police in case alleged criminals decide to retaliate.10 Therefore, the police often tacitly endorse violence as a legitimate response to crime by allowing community members to undertake law enforcement duties by way of their role in sentencing and meting out punishment to alleged offenders. Finally, the state may act in a regulatory capacity, deciding who may have a license to provide security and empowering, monitoring, and sanctioning these actors according to a set of state-devised standards. In South Africa, the key body responsible for this task is the Private Security Industry Regulatory Authority (PSIRA). Its executive council is appointed by the minister of police, and is responsible for monitoring and evaluating the activities of private security providers within the commercial security industry. As part of this task, the PSIRA has the authority to approve and revoke licenses for commercial security providers. State regulation of societal-­based security is much more difficult to achieve. Because societal security structures are largely staffed by ordinary citizens and found throughout many local communities, the costs of effectively monitoring these structures would be very high. Thus, many of these structures function autonomously of the state. However, even in the area of societal-­ based security, the South African state has attempted to gain leverage and control. The clearest way in which it has attempted to do this is by encouraging citizens who care about crime and security issues to join CPFs. Because CPFs are officially housed at police stations, and because at least one police officer is supposed to sit on each forum, the state has attempted to use the CPF as a way of attempting to regulate the behavior of ordinary citizens and set the tone and direction of their participation in crime fighting. The above discussion provides concrete examples of the activities the state may undertake in its capacity as an arranger of security. However, it is worth noting here that the theoretical thrust of this book is not concerned with whether the state actually acts as a producer or arranger, but

9

 Los Angeles Times. “The horror of vigilante justice”. September 24, 2012.  CPF chairperson. Personal interview. December 30, 2010.

10

14 

D. C. KUSHNER

whether citizens believe it should act in one of these two different capacities. Citizens may come to see the state as an arranger of goods because of historic conditions that removed the state from or downplayed the state’s role as a producer. It’s also possible that citizens may come to see the state as an arranger because of strong norms of self-reliance that push individuals and communities to directly take up the reins of service provision.

1.5   The Impact of Personal Insecurity and Victimization on Popular Attitudes and Participation In addition to investigating the political consequences of citizens’ non-­ state security reliance, the second aim of this book is to understand how citizens’ everyday security experiences affect their political attitudes and behavior. For all the literature that examines the micro foundations of political behavior, there is surprisingly little work on the impact that citizens’ experiences of and attitudes toward crime have on their politics in Africa. Notable exceptions include Bateson (2012) who examines the impact of crime victimization on political participation worldwide, and Fernandez and Kuenzi (2009) who examine the impact of crime on support for democracy in Africa. Slightly more work has, however, been conducted on this topic in the Latin American context (Hurwitz and Peffley 2005; Pérez 2003). I argue that citizens’ everyday experiences of crime and violence matter, politically. Studying the political impact of criminal rather than political violence has never been more timely for Africa, given its rapid pace of urbanization and that fact that there are now more democracies in the region than ever before (and democracies are less prone to political violence than non-democracies). This book examines the impact of a number of crime-related variables, including violent (contact) and property ­victimization, personal insecurity, and evaluations of state and non-state security providers.

1.6   Outline of the Book This chapter has presented an overview of the main research questions raised throughout the book and my theoretical approach for explaining how non-state security shapes perceptions of state legitimacy.

 INTRODUCTION 

15

Chapter 2 introduces my key case study, South Africa, and shows why this country is an opportune place within which to study the political consequences of non-state security provision. It first lays out the historical context, showing the long history of non-state security reliance in this country. It discusses how, under apartheid, both the state and ordinary citizens relied heavily on various forms of non-state security, and how much of this reliance was connected to the politicization of crime and the police force under that regime. The chapter then moves to the post-­ apartheid and contemporary period. It suggests that even under the post-­ apartheid democratic order, crime and security remain salient issues in South Africa and the heavy reliance on non-state security persists. The chapter suggests several possible explanations for this outcome, and explores over time trends in crime using official police statistics and South African National Victims of Crime surveys. It also uses public opinion surveys to determine South Africans’ attitudes toward crime, and whether and how these attitudes have shifted over time. Chapter 3 is the first empirical chapter. Using original survey data collected on hundreds of South African citizens, and original data from interviews with elected officials, civil society organizations, and police officers, I explore the impact that non-state security reliance and South Africans’ broader experiences of crime and violence have on their perceptions of state legitimacy. I examine two types of non-state security in this context; market-based and societal-based. Market-based security is security that is provided in exchange for a fee. Commercial security companies primarily provide this form of security. Since there are a plethora of non-state actors involved in the provision of security in South Africa, however, it is important to explore the full range of non-state security providers in this context. Therefore, I also investigate the political consequences of societal-based security. Unlike market-based security, the provision of societal-based security is not contingent on financial exchange. Instead, societal-based security is largely voluntary. In this book, I account for several types of societal security actors that individuals may turn to for protection, including relatives, community police forums, street committees, and neighborhood watch groups. I find that the source of security provision (state vs. non-state) does matter for individual perceptions of state legitimacy, but that other factors, such as being a victim of crime, hold explanatory power as well. I end the chapter by conducting a robustness check of my state legitimacy model using

16 

D. C. KUSHNER

Afrobarometer data. Given that my survey was limited to urban areas of South Africa and that I lacked some variables in my survey that would allow me to successfully control for some competing explanations, I decided to run a similar model using nationally representative Afrobarometer survey data collected from October–November 2011 in South Africa. The impact of non-state security on perceptions of state legitimacy remains strong and significant. Chapter 4 is the second empirical chapter. It presents an empirical analysis of social capital and collective action. While Chap. 3 is primarily concerned with understanding citizens’ attitudes toward the state, Chap. 4 attempts to move beyond attitudes to assess political behavior. In this chapter, I explore the impact of non-state security, victimization, personal insecurity, and citizens’ evaluations of state security on joining and collective action across Africa. I find that who individuals turn to for their security needs is a key determinant of political participation, as is victimization. Importantly, the impact of victimization varies depending on the type of victimization under consideration (property vs. violent) and the type of political participation. Analyses for this chapter rely on the fifth round of Afrobarometer data collected in South Africa in 2011. Chapter 5 examines the relationship between security factors and protest behavior. While conventional determinants of protest retained their explanatory power, security variables also proved to be important in explaining protest. In particular, those who feel personally insecure in their neighborhoods and those who report being victims of violent crime are more motivated to engage in protest than their counterparts. And even non-state security reliance impacts one’s decision to protest when they believe the state is still somewhat legitimate. Chapter 6 is the final empirical chapter, and it assesses the impact of my key security variables on voting in South Africa. I find that non-state security reliance and reported experiences of violent victimization both positively impact voter turnout in South Africa. Chapter 7 concludes the book. This chapter briefly re-states the main findings and thoroughly discusses their implications. It also speaks to the generalizability of these findings and how well my theory applies across Africa and the developing world more generally. Finally, this chapter discusses the limitations of this book and suggests directions for future research.

 INTRODUCTION 

17

References Anderson, C.J. 2000. Economic voting and political context: A comparative perspective. Electoral Studies 19 (2–3): 151–170. Baker, B. 2007. Multi-choice policing in Africa. Uppsala: Nordic Africa Institute. Bateson, R. 2012. Crime victimization and political participation. American Political Science Review 106 (3): 570–587. Batley, Richard, and Claire Mcloughlin. 2009. State capacity and non-state service provision in fragile and conflict-affected states. Bayley, D.H., and C.D. Shearing. 1996. The future of policing. Law and Society Review 30 (3): 585–606. Brass, Jennifer. 2016. Allies or adversaries: NGOs and the State in Africa. New York: Cambridge University Press. Bratton, M., and E.C.C. Chang. 2006. State building and democratization in sub-­ Saharan Africa: Forwards, backwards, or together? Comparative Political Studies 39 (9): 1059. Bratton, M., and R. Mattes. 2001. Support for democracy in Africa: Intrinsic or instrumental? British Journal of Political Science 31 (3): 447–474. Bratton, M., M. Coulibaly, and F. Machado. 2002. Popular views of the legitimacy of the state in Mali. Canadian Journal of African Studies 36 (2): 197–238. Burris, S., P.  Drahos, and C.  Shearing. 2005. Nodal governance. Australian Journal of Legal Philosophy 30: 30. Cammett, Melanie, and Lauren M. MacLean. 2014. The politics of non-state social welfare. Ithaca, NY: Cornell University Press. Diamond, L., and L.  Morlino. 2004. The quality of democracy. Journal of Democracy 15 (4): 20–31. Easton, David. 1957. Approach to the analysis of political systems. Cambridge: Cambridge Univ Press. Englebert, P. 2002. State legitimacy and development in Africa. Boulder, CO: Lynne Rienner Pub. Fernandez, Kenneth E., and Michele Kuenzi. 2009. Crime and support for democracy in Africa and Latin America. Political Studies 58 (3): 450–471. Gibson, J.L., and G.A.  Caldeira. 2003. Defenders of democracy? Legitimacy, popular acceptance, and the South African Constitutional Court. Journal of Politics 65 (1): 1–30. Gilley, Bruce. 2006. The determinants of state legitimacy: Results for 72 countries. International Political Science Review 27 (47): 24. Herbst, Jeffrey Ira. 2000. States and power in Africa: Comparative lessons in authority and control. Princeton, NJ: Princeton University Press. Hern, Erin. 2017. In the gap the state left: Policy feedback, collective behavior, and political participation in Zambia. Studies in Comparative International Development 52 (4): 510–531.

18 

D. C. KUSHNER

Higgott, Richard A., Andreas Bieler, and Geoffrey Underhill. 2000. Nonstate actors and authority in the global system. New York: Routledge. Hobbes, T. 1998. Leviathan. Oxford: Oxford University Press. Hurwitz, Jon, and Mark Peffley. 2005. Playing the race card in the post-Willie Horton era: The impact of racialized code words on support for punitive crime policy. Public Opinion Quarterly 69 (1): 99–112. Jackman, R.W. 1993. Power without force: The political capacity of nation-states. Ann Arbor: University of Michigan Press. Jackson, R.H., and C.G.  Rosberg. 1982. Why Africa’s weak states persist: The empirical and the juridical in statehood. World Politics: A Quarterly Journal of International Relations 35 (1): 1–24. Johnston, L., and C. Shearing. 2003. Governing security. Explorations in policing and justice. London: Routledge. Kempa, M., R.  Carrier, J.  Wood, and C.  Shearing. 1999. Reflections of the evolving concept of ‘private policing’. European Journal on Criminal Policy and Research 7 (2): 197–223. Krahmann, Elke. 2005. American hegemony or global governance: Competing visions of international security. International Studies Review 7: 531–545. Kushner, Danielle C., and Lauren M. MacLean. 2015. Introduction to the special issue: The politics of the nonstate provision of public goods in Africa. Africa Today 62 (1): vii–xvii. Levi, M., A.  Sacks, and T.  Tyler. 2009. Conceptualizing legitimacy, measuring legitimating beliefs. American Behavioral Scientist 53 (3): 354. Lewis-Beck, M.S., and M.  Paldam. 2000. Economic voting: An introduction. Electoral Studies 19 (2–3): 113–121. Lipset, S.M. 1959. Some social requisites of democracy: Economic development and political legitimacy. The American Political Science Review 53 (1): 69–105. Locke, J.  1966. The second treatise of civil government and a letter concerning toleration. Oxford: Blackwell. Meagher, Kate. 2012. The strength of weak states? Non-state security forces and hybrid governance in Africa. Development and Change 43 (5): 1073–1101. Mondak, J.J. 1993. Institutional legitimacy and procedural justice: Reexamining the question of causality. Law and Society Review 27: 599–608. Moran, Dominique, and Richard Batley. 2004. Literature review of non-state provision of basic services. Paper commissioned by DFID from Governance Resource Centre, University of Birmingham, UK. Pavanello, Sara, and James Darcy. 2008. Improving the provision of basic services for the poor in fragile environments: International Literature Review-Synthesis Paper. London: Overseas Development Institute. Peltier, Jean-Philippe. 2007. State legitimacy in sub-Saharan Africa, political science. East Lansing: Michigan State University.

 INTRODUCTION 

19

Pérez, O.J. 2003. Democratic legitimacy and public insecurity: Crime and democracy in El Salvador and Guatemala. Political Science Quarterly 118 (4): 627–644. Powell, G.B., Jr., and G.D. Whitten. 1993. A cross-national analysis of economic voting: Taking account of the political context. American Journal of Political Science 37 (2): 391–414. Rose, Pauline M. 2007. Supporting non-state providers in basic education service delivery. Rotberg, R.I. 2003. When states fail: Causes and consequences. Princeton, NJ: Princeton University Press. Sacks, Audrey. 2012. Can donors and non-state actors undermine citizens’ legitimating beliefs? World Bank Policy Research Working Paper (6158). Savas, E.S. 2000. Privatization and public-private partnerships. New  York: Chatham House. Schaar, J.H. 1981. Legitimacy in the modern state. New Brunswick, NJ: Transaction Publication. Schatzberg, M.G. 2001. Political legitimacy in middle Africa: Father, family, food. Bloomington, IN: Indiana University Press. Shearing, C.D. 1992. The relation between public and private policing. Crime and Justice 15: 399–434. Tyler, T.R. 2003. Procedural justice, legitimacy, and the effective rule of law. Crime & Justice 30: 283–357. Tyler, T.R., and J. Fagan. 2010. Legitimacy and cooperation. Race, Ethnicity, and Policing: New and Essential Readings 6: 84. Weingast, B.R. 1997. The political foundations of democracy and the rule of law. American Political Science Review 91 (2): 245–263. Wood, J., B. Dupont, and MyiLibrary. 2006. Democracy, society, and the governance of security. New York: Cambridge University Press.

CHAPTER 2

Non-state Security in South Africa: Historical Roots, Contemporary Realities

Abstract  This chapter introduces the key case study of the book, South Africa, and shows why this country is an opportune place within which to examine the political consequences of non-state security provision. It first lays out the historical context, showing the long history of non-state security reliance in this country. It discusses how, under apartheid, both the state and ordinary citizens relied heavily on various forms of nonstate security, and how much of this reliance was connected to the politicization of crime and the politicization of the police force under that regime. The chapter then moves to the post-apartheid and contemporary period. It suggests that even under the post-apartheid democratic order, crime and security remain salient issues in South Africa and the heavy reliance on non-state security persists. The chapter suggests several possible explanations for this outcome and explores over time trends in crime using official police statistics and South African National Victims of Crime surveys. It also uses public opinion surveys to determine South Africans’ attitudes toward crime and whether and how these attitudes have shifted over time. Keywords  Non-state security • Crime • Security • South Africa • Apartheid

© The Author(s) 2019 D. C. Kushner, The Politics of Everyday Crime in Africa, https://doi.org/10.1007/978-3-319-98095-9_2

21

22 

D. C. KUSHNER

2.1   Introduction This chapter describes the nature of the non-state security atmosphere in South Africa. It begins with a discussion of crime and security under apartheid and examines the variety of non-state security structures that arose during this tumultuous political time. After reviewing the apartheid period, the chapter then moves on to review issues of crime and security since the transition to a democratic political regime in 1994, assessing trends and challenges using official statistics and public opinion data. Finally, the chapter explores how the state has responded to crime in the post-apartheid context and identifies the non-state entities that participate in the provision of security in contemporary South Africa.

2.2   Crime and Security Under Apartheid There are several important factors to note about crime and security under apartheid. First, many acts that would be considered ordinary forms of political participation and expression under a democratic regime were deemed crimes under apartheid. Second, policing services, like most other services provided by the state, were distributed along racial lines. Finally, non-state actors played a vital role in the provision of security during this time, with some working alongside and in cooperation with state police, and others operating in direct opposition to the state. 2.2.1  The Politicization of Crime It is impossible to understand the nature of crime under apartheid without understanding its explicit relationship to politics (Bayart et  al. 1999; Kynoch 2005; Shaw 1995, 2002). How the apartheid state defined a crime often had little to do with the objective of maintaining a safe and secure atmosphere for all. Instead, criminality was largely politicized, and those who engaged in political acts were often labeled as terrorists. One way in which the state was able to criminalize the political was by using the legal system to buttress its political goal of maintaining white minority rule (Schönteich and Louw 2011). Several scholars have shown how draconian laws were used to criminalize everyday activity and how opposition political parties that were seen as a threat to the ruling National Party, most notably the African National

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

23

Congress (ANC), were banned. Thus, any individual who expressed ­discontent with or opposition to the prevailing political order was, in effect, guilty of breaking the law. In this context, the police were much less focused on conventional policing duties such as crime prevention and investigation. On the contrary, the job of the police was to specialize in political repression. Therefore, the police in South Africa were militarized and engaged in a war with the vast majority of the South African population, who were seen as the enemy. In this climate, the police’s efforts and resources were largely dedicated to monitoring the black population, preventing political uprisings, and quelling them where they did develop. The police were also known to participate in “death squads”, which tortured and assassinated those who worked with the ANC and other opposition parties (Pauw 1991). Thus, some would argue that the state itself was criminalized, using torture and other intimidation techniques to discourage any political behavior that would contribute to the rise of a viable opposition. The politicization of crime and violence was not, however, solely the preserve of the state. The armed wing of the ANC (Umkhonto we Sizwe) and to a lesser extent the United Democratic Front (UDF, a coalition of politicized civic organizations) often used violence, both directly against state agents and as a way of achieving allegiance among the masses. In fact, political activists sometimes used the violent technique known as necklacing1 against public officials and ordinary citizens who were suspected of cooperating with the state (Buur and Jensen 2004). Violence was therefore used by the opposition as a means of securing a base of supporters in the fight against apartheid and discouraging any potential collaboration with the state. More generally, the ANC encouraged widespread mass disregard for the law in their campaign to make the state “ungovernable”. The main tactic was to withhold revenue from urban local governments (“the townships”) by engaging in tax and rent boycotts. Where these tactics failed, the UDF, acting as an agent of the ANC, was not averse to intimidating township residents and, at times, even assassinating state officials (Neocosmos 1998).

1  Necklacing refers to the practice of placing an oil-filled tire about someone’s neck and then setting the tire aflame.

24 

D. C. KUSHNER

2.2.2  The Distribution of Policing Services Aside from political violence, there were also regular forms of everyday criminal activity under apartheid. As Mayekiso (1996) notes, there is bound to be criminal activity where high levels of unemployment and alienation exist (Mayekiso 1996). So, ordinary forms of crime became ways for people to subsist in poverty stricken townships. As with most other services under apartheid, the provision of policing services was highly skewed along racial lines. According to the 1998 White Paper on Safety and Security, “in 1994, 74% of the country’s police stations were situated in White suburbs or business districts” (“White Paper on Safety and Security” 1998). This suggests that, on the rare occasions that the police did engage in normal duties of crime control and prevention, they did so primarily in white residential and business areas. The principal mission of the South African Police (SAP) force was to protect and insulate white South Africans from crime, violence, and political insurrection. Crime in African townships was rarely, if ever, investigated and punished (Reconciliation 2009). Given that crime was defined in narrow, political terms and that police services were concentrated primarily in minority neighborhoods, this created space for the involvement of various non-state actors in the provision of security. The next section will explore the various actors that became involved in non-state policing under the apartheid regime. 2.2.3  Non-state Security Under Apartheid 2.2.3.1 Market-Based Security Under apartheid, the commercial security industry thrived. In large part, the demand for commercial security was driven by the state. Since white minority rule involved restricting key freedoms of the majority black population, including restrictions on where they could travel and live within the country, massive amounts of manpower was needed to achieve this level of control. The state often turned to private security companies to supplement its understaffed public police force (Brogden and Shearing 1993). In this political climate, security companies were given extensive policing powers, with some scholars even arguing that they constituted a parallel police force. These companies not only had the power to arrest, but also full rights of search and seizure (Irish 1999).

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

25

That the commercial security industry took on the political tone of the apartheid regime and became militaristic in its approach is by now undeniable. Singh (2008) notes that commercial security training manuals from the 1980s made constant reference to “the enemy” evidences that the commercial security industry had become a key ally of the state in its efforts to repress political dissent. Not only did the state extend contracts and many policing powers to the commercial security industry, but according to Singh, it was also able to gain indirect and direct control over the industry and ensure the industry’s allegiance to its political agenda with key pieces of legislation. Indirectly, the state politicized commercial security and the expansion of the commercial security industry with the National Key Points Act of 1980 and later an amendment to the Prevention of Illegal Squatting Act in 1988. With the former piece of legislation, once a site was designated a Key Point, or of relevance to national security interests, the state could then mandate owners of the property to hire commercial security. In this way, the commercial security industry largely became the guardian of white property interests. The amendment to the Prevention of Illegal Squatting Act also allowed for the state’s indirect control of the industry, namely by shifting the onus to citizens for evicting squatters from their property. To assist in this task, those who owned property often turned to the commercial security industry. The Security Officers Act (SOA) of 1980 allowed for more direct state control of the commercial security industry. As Singh (2008) notes, “the SOA granted the Minister of Law and Order and the South African police significant authority in the control of private security, thus providing at least the potential for the identification of industry objectives with state security interests” (45). The SOA was the first piece of legislation developed to officially regulate the commercial security industry. In so doing, it established a regulatory board, appointments of which were made by the Minister of Law and Order. Thus the informal linkages that existed between the commercial security industry and the South African Police began to be formalized through the SOA, with the industry brought into an increasingly tight relationship with the state. 2.2.3.2 Societal-Based Security Aside from the commercial security industry, non-state security initiatives were also to be found more generally within society under apartheid. Most studies agree that these community-based structures reached their apex during the height of the struggle against apartheid in the 1980s (Adler

26 

D. C. KUSHNER

and Steinberg 2000; Bundy et al. 2000; Mayekiso 1996; Schärf and Nina 2001). Schärf and Nina (2001) explore the rise of non-state initiatives like self-defense units (SDUs), anti-crime committees, and people’s courts from 1984 onward and note that the development of these structures coincided with the ANC’s goal of making the townships ungovernable by the state. In their view, the exclusion of the state from these locales, led to an increasingly important role for informal institutions in the provision of security, justice and order. As Buur and Jensen (2004) note, more often than not, the emergence of vigilante formations is premised on a deep seated mistrust of the police or perceived lack of initiative by police in providing basic human and economic security. These were the case in townships under apartheid. Mistrust of the police created space for the development of non-state township structures such as SDUs, street committees, and people’s courts. The civics movement, which consisted of a range of locally based organizations led by African notables such as clergymen and clerks, was responsible for developing these community-based structures (Adler and Steinberg 2000; Seekings 2000). The street committees and SDUs served several purposes in their communities. On the one hand, they were responsible for the provision of basic goods and social services. In this sense, they played a key social welfare role. On the other hand, they were largely responsible for the provision of security and order. Because the state police, the SAP, did not control crime in townships, street committees took on the responsibility of policing black communities and protecting them from crime. Moreover, because township residents were often the direct targets of state-led violence, these structures were also responsible for protecting their communities from official state coercion. It is important to note that those involved in the various organizations that were part of the civics movement often came into direct, sometimes violent, conflict with the state. In particular, members of the civics often came into conflict with black councilors from the Black Local Authorities (BLAs) who were seen as puppets of the apartheid regime. In opposition to this form of repressive local governance, the civics boycotted local elections, burned government buildings, encouraged residents to halt the payment of rental and service charges, and sometimes injured and killed councilors (Adler and Steinberg 2000; Buur 2010). Thus, under apartheid there existed market and societal-based forms of non-state security, with the market-based version tightly aligned with the state and societal forms often standing in stark opposition to this political entity.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

27

2.3   Crime and Security Since the 1994 Transition Apartheid, a system of racial segregation and white minority rule instituted in 1948, began to unravel in 1990 as domestic and international pressure led President F.W. de Klerk to unban the opposition ANC party and release its leader, Nelson Mandela, from prison. As it became increasingly clear that political transition was inevitable, the country adopted an interim constitution and began to strike down many discriminatory apartheid laws. Four years later in 1994, the first multi-party elections were held in South Africa. The ANC won with a landslide victory and Nelson Mandela was elected as the country’s president. Even though the country successfully transitioned to democracy in 1994 and managed to abolish many draconian apartheid era laws, the politicized nature of crime under apartheid would hold two very important implications for the post-apartheid security atmosphere. The first implication is that the post-apartheid state inherited weak policing capacity. Because the police were primarily concerned with the political priority of keeping the apartheid regime in power, they lacked traditional policing skills that focused on crime prevention, detection, and investigation (Baker 2008). The second implication is that the new state had to contend with deeply entrenched levels of mistrust between the police and ordinary citizens. Thus, after the transition there was a heavy state focus on police transformation (Marks 2005; Shaw 2002). The concept of police transformation had two primary objectives. The first was for the police to learn basic policing skills needed to police citizens in a civil way as opposed to the authoritarian, militaristic style of policing carried out under apartheid. The second objective was for the police to transform in a way that would enable them to earn the trust and respect of the people (Marks 2005). To this end, the name of the police force was changed from the South African Police (SAP) to the South African Police Service (SAPS) to signal to members of society that the police had a new identity, one that was severed from the politics of the past.2 In other words, 2  In an interview with a provincial head for Business Against Crime, a major non-profit that works with government to fight crime in South Africa, he mentioned that the change to the new police force after the 1994 transition could be witnessed by the change in uniform. He smirked as he remarked that the police had “traded in their fatigues for their baby blues” referring to the blue uniforms that the police now wear. For him, this change was problematic; in his mind, the police under the new order had become soft and, as a result, people had “lost respect” for them.

28 

D. C. KUSHNER

the police force, as a key institution of the new democratic order, recognized that it not only needed to build capacity, but also legitimacy. In an effort to build legitimacy, the police force not only attempted to become more service-oriented, but it also made diversification of the upper ranks of the police force a key priority.3 In fact, South Africa adopted a national police force, rather than local forces, in part, to ensure that policing services would be distributed in a more fair and equitable manner under the new order.4 An incompetent and untrusted police force was not, however, the only challenge to creating an atmosphere conducive to democracy and the rule of law in South Africa. Another issue is that ordinary citizens, even if for understandable reasons, had become accustomed to rebelling against the law during the struggle for liberation (Gibson and Gouws 1997). Moreover, many South Africans developed a culture of self-reliance under the old regime, and were used to relying on non-state, community-based structures to solve disputes. The state, therefore, had to demonstrate to ordinary citizens the value and necessity of working through state structures to settle conflicts and achieve justice. For many citizens, the idea that one could and should turn to the state to regulate private affairs was new and would take time to embrace. These historical realities thus helped to set the tone of the security atmosphere in post-apartheid South Africa. The newly elected government inherited an environment where there was widespread criminality, weak policing capacity, and high levels of mistrust in the police. All of these factors collectively created a context that made security a primary policy issue with which to be reckoned in the post-­apartheid state. My original survey, conducted between October 2010 and July 2012  in Cape Town, Durban, and Johannesburg, South Africa, shows that non-­state security is not a thing of the past. At least among urban

3  Data on the exact composition of the police force are no longer made publicly available. However, SAPS does post the names and some pictures of high-ranking police officials on their website. From these profiles, it is clear that the upper ranks of the police force have been transformed from predominantly white officers to predominantly black ones. There also appears to be a greater degree of gender diversity within SAPS. 4  Under apartheid, the ruling National Party divided the country into 11 different states known as “homelands”. These homelands were developed so that each ethnic group could have its own territory. Under this system, black people were not citizens of South Africa; rather they were citizens of their respective homelands. Needless to say, the provision of public goods was substandard in the homelands. Under this system, there was a separate policing agency for each homeland. The interim constitution abolished the homeland system and the new constitution signed into law in 1996 established a single, national police force for all of South Africa.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

29

South Africans, reliance on non-state security not only persists, but is widespread. The survey asked respondents the following question “Please tell me who you would call on if you or someone in your family had been a victim of crime?” The response categories consisted of “police only”, “police and some other group”, “other groups only”, or “none of these”. Figure 2.1 shows that a slim majority of respondents (51%) relies solely on the police for security. This figure captures the percentage of individuals who turn to the police for help after they or someone in their family had been a victim of crime. However, there is a similar trend for individuals who turn to the police for crime prevention. For example, when asked who they would turn to keep themselves or someone in their family from becoming a victim of crime, a plurality (46%) of respondents says that they would rely solely on the police. In both instances, however, we see that at least 45% of respondents rely on non-state groups for security, either entirely, or in conjunction with state police. Given that individuals are quite supportive of non-state actors providing security, to which types of non-state security are they most likely to turn? My survey shows that reliance on market-based security dominates (Fig. 2.2). Of those who rely on non-state security, a majority of 58% report relying on a commercial security company for security services. In what I refer to as societal-based forms of non-state security, 19% of respondents 60

51

50 40

34

30 20

11

10

2

0 Police only

Police and some other group

Other groups only

None of these

Fig. 2.1  Security reliance post victimization, 2010–2012 Question: Please tell me who you would call if you or someone in your family had been a victim of crime? N = 410

30 

D. C. KUSHNER

70 60 50 40 30 20 10 0

58

19

y

e al ci om m er C

om m un

ity

se

po

cu

rit

lic

ily m Fa C

bo gh N

ei

St re

et

rh

co

oo

d

w

m m itt

at

ee

er

th

O

ch

5

2

8

7

Fig. 2.2  Type of non-state security reliance, 2010–2012 Question: In addition to or besides the police, which of these groups or organizations would you be most likely to turn to for help after the crime? N = 178

report turning to a community police forum (CPF), while 8% report turning to family and 7% to a neighborhood watch scheme (NWS). Of the various forms of non-state security, respondents are least likely to report relying on street committees (5%). Thus, in South Africa, individuals from all walks of life rely on a range of actors for security, from the conventional police force, to the market and society. But who is most likely to rely on what forms of non-state security? The data (Table 2.1) show that there is a racial gap in terms of non-state security reliance, with whites and Asians being more likely to rely on commercial non-state security and blacks and coloreds more likely to rely on societal non-state security. In fact, whites make up 56% of those who rely on commercial non-state security, although they are only about 10% of the total population. Blacks, on the other hand, comprise a majority of respondents that rely on societal forms of non-state security like NWSs, street committees, and CPFs. These figures suggest that individuals’ reliance on non-state security will depend, to a certain extent, on their socio-­economic standing. Only individuals who have achieved a certain degree of wealth can afford to hire a commercial security company to protect their person and property. In most instances, only whites and Asians earn a high enough income to afford these services. However, commercial security companies are not without some support from black consumers (Table 2.2). In fact, among blacks who rely on

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

31

Table 2.1  Type of non-state security reliance by race, 2010–2012

Commercial security Neighborhood watch Street committee Community police forum

Black

White

Colored

Asian

Total

14% 50% 100% 88%

56% 8% 0% 3%

4% 42% 0% 9%

26% 0% 0% 0%

100% 100% 100% 100%

Question: In addition to or besides the police, which of these groups or organizations would you be most likely to turn to for help after the crime? N = 178

Table 2.2  Black reliance on non-state security, 2010–2012 Commercial security Neighborhood watch Street committee Community police forum Family Other Total

24% 10% 15% 48% 2% 2% 100%

Question: In addition to or besides the police, which of these groups or organizations would you be most likely to turn to for help after the crime? N = 62

non-state security, 24% rely on commercial security, coming only second to reliance on CPFs. Most of this demand for commercial security by blacks is driven by those respondents from Johannesburg where the black business class is concentrated. The fact that such a significant number of South Africans rely on non-­ state security today of course begs the question why. Why do so many South Africans continue to rely on non-state security under the new political order? 2.3.1   Salience of Crime and Security in South Africa One reason for the widespread persistence of non-state security reliance after the transition to multi-party rule is the continued salience of crime and security issues in South Africa. With the transition to democracy, many South Africans were hopeful that the challenges of the past, including crime and insecurity, would successfully be addressed by a popularly elected

32 

D. C. KUSHNER

government. However, levels of violence remain extraordinarily high in South Africa, although this violence is now more criminal than political in nature (Harris 2003). The high levels of crime present in South Africa contemporarily may help to explain why South Africans have consistently ranked crime and security as one of their top five concerns over time.5 It is difficult to know whether crime has increased in South Africa since the transition, especially since crime statistics from the apartheid era included not only criminal, but also political offenses (Schönteich and Louw 2011). What is clear, however, is that the state continues to struggle to provide a sense of peace, security, and order for its citizens. A 2007 report compiled by the Center for the Study of Violence and Reconciliation using data from the United Nations Office on Drugs and Crime (UNODC), showed that South Africa had the second highest murder rate of all countries included in the sample.6 In fact, South Africa has often been referred to as the crime capital of the world (Altbeker 2005). This reputation, whether justly warranted or not, has serious implications for how citizens view state authority in this context. 2.3.1.1 Crime Trends in South Africa: 1994–2017 While South Africa still contends with extraordinarily high levels of crime, the overall trend is that crime has been steadily, and in some cases sharply, decreasing since the transition to democracy in 1994.7 To begin with the most violent crimes, the number of murders per 100,000 of the population decreased from 67 in the 1994/1995 reporting period to 41 in the 2016/2017 reporting period, a decrease of 39%. Attempted murder has also sharply declined. Murder and attempted murder, two of the most serious violent crimes, have been cut nearly in half since the transition to multi-party rule. But while the improvement in these two forms of crime

5  Afrobarometer, the largest cross-national public opinion survey in Africa, asks citizens what they perceive to be the most important problem facing their nation. South Africans have consistently rated crime and security as one of the tops concerns, often coming second only to the issue of unemployment. 6  Columbia had the highest murder rate at the time. Others countries in the sample were (in descending order of murder rate) Venezuela, Brazil, Ecuador, Swaziland, Mongolia, Suriname, Lithuania, Latvia, Zimbabwe, Belarus, Kyrgyzstan, Ukraine, Uganda, Estonia, Moldova, Sri Lanka, Costa Rica, Georgia, Uruguay, Peru, and the United States. 7  Data on crime trends come from annual reports released by the South African Police Service (SAPS).

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

33

is a step in the right direction, we should, however, note that the murder rate still remains very high in South Africa. For example, over 19,000 people were murdered in South Africa in 2016/2017. This is more than the number of individuals that were murdered in the United States in 2016, even though the U.S. is six times as large as South Africa (Uniform Crime Reporting Program 2017). Therefore, while improvements have been made, there is still substantial work to be done with respect to the prevalence of murder and attempted murder in this country. South Africa has also reduced instances of assault and rape. Rates of both common assault and assault with the intent to inflict grievous bodily harm have dropped substantially over the past 20 years, over 50% and 45% respectively, to be exact. With respect to rape, there were 116 reported rapes per capita in 1994/1995. By 2016/2017, this number had decreased to 71, a decrease of 39%. Yet, it is difficult to know exactly how well official statistics represent instances of rape in South Africa, as this type of crime is grossly underreported here (as in many other places). Moreover, there have been significant changes to the definition of rape over time. Before 2007, for example, the definition of rape used by the SAPS was “unlawful and intentional sexual intercourse with a female without her consent”. Therefore, males who had been the victim of rape were not included in rape statistics prior to 2007. The problems with rape statistics notwithstanding, the overall picture shown is that rape is on the decline. As for robbery, it is worth noting that aggravated robbery is more prevalent than common robbery. A great number of robberies that are committed are done so with a weapon, contributing to the large amount of victims that are injured or killed in the commission of a robbery. As of 2016/2017, a total of 107 aggravated robberies per 100,000 of the population were reported. The occurrence of aggravated robbery has decreased by roughly 50% since the end of apartheid. Common robbery, by contrast, actually increased during this period from 85 per 100,000 of the population to 96 per 100,000 of the population. It may therefore come as no surprise that robbery is one of the crimes that citizens fear the most,8 given the difficulty of policing this kind of crime and reducing its occurrence. 8  Results from the 2016/2017 Victims of Crime Survey show that 45% of households in South Africa believe that home robbery is the crime that is feared the most followed by 42% for robbery outside the home. The figures for these two types of robbery come second only to housebreaking/burglary, since 51% of households believe that this is the crime that is feared the most. Note that individuals were allowed to choose more than one crime that they feared the most.

34 

D. C. KUSHNER

Finally, burglary (residential and non-residential combined) has decreased substantially, from 826 per 100,000 of the population in 1994/1995 to 566 as of 2016/2017. However, most of the decrease in burglary has been driven by a decline in residential burglary. Prior to 2003, data made available on burglary were only the combined figure for nonresidential and residential burglaries. From 2003 forward, however, separate burglary figures were made available. This is helpful because it allows us to see, first, that residential burglary is much more prevalent than nonresidential burglary. While individuals and businesses alike heavily rely on alarm systems and armed guards for protection, it seems that businesses are still better positioned to insulate themselves from crime. Second, while residential burglary decreased by 18% between 2003/2004 and 2016/2017, levels of non-residential burglary actually increased by 17%. According to the above tables and figures, South Africa has made progress in reducing levels of crime from 1994 to the present. According to the data provided by SAPS, the rate of each category of violent crime, with the exception of common robbery, has been substantially reduced over time. The country has also witnessed impressive decreases in residential burglaries, even if not in non-residential ones. Why then do so many people continue to rely on non-state security in South Africa? Perhaps data on perceptions of crime will provide a clue. 2.3.1.2 Popular Perceptions of Crime When assessing the crime situation in any country, it is important to examine public opinion on crime. Official reports and crime statistics may have an impact on individuals’ perceptions of crime, but perceptions of crime are rarely simply a function of actual levels of crime. In fact, there are often wide gaps between the actual prevalence of crime and individual perceptions of the pervasiveness and severity of crime. On the one hand, many crimes that occur are never reported to the police; on the other, sensational press reporting may inflate popular perceptions of the prevalence of crime. Thus, reviewing micro-level data on individual perceptions of crime may complement or provide an alternative view to state-produced statistics. I rely on Victims of Crime Survey (VOCS) data to depict citizens’ perceptions of crime in South Africa. To date, eight national VOCS have been conducted in South Africa. The first was carried out in 1998 by Statistics South Africa. The Institute for Security Studies (ISS) was responsible for conducting the 2003 and 2007 versions of the VOCS. In 2011 and 2012, Statistics South Africa once again resumed responsibility for administering the VOCS and they have continued to do so on an annual basis from this point forward.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

35

The VOCS is a nationally representative household-based survey conducted with South African residents that are 16 years of age or older. It measures citizen’s perceptions of and reported experiences with crime. It also gauges their attitudes toward key criminal justice institutions such as the police and courts. For some of the years that the VOCS was conducted, I was not able to obtain access to the raw data. In those instances I have primarily relied on reports and presentations on the data.9 2.3.1.2.1  Rate of Crime/Personal Security To begin, we can explore whether citizens’ views on the crime rate are consistent with the trend captured in official crime statistics. In 2016/2017 when respondents were asked whether they thought crime had “increased”, “decreased”, or “remained the same” in the last three years, over 40% of respondents reported that they felt crime had “increased”. As of 2017, South Africans were clearly split on their perceptions of progress on crime. An even greater percentage of urban South Africans (50%) are convinced that crime is on the rise.10 Both the Victims of Crime data and data from my survey conducted from 2010 to 2012 stand in stark contrast to official statistics, which show a decline in most categories of crime over time. At the same time, the percentage of individuals who report that crime has “increased” has dropped over time. In 2003, a majority of individuals (53%) from the Victims of Crime survey reported that crime was on the rise. Thirteen years later, this percentage has dropped by roughly ten percentage points. Therefore, even though a plurality of respondents still felt that the crime situation was getting worse in South Africa as of 2016/2017, the percentage of people who feel this way has dropped over time (Fig. 2.3). What then are the factors that determine whether people see crime as improving, remaining the same, or getting worse? When examining perceptions of crime by race (figure not shown), we see that Asians and coloreds are slightly more likely than blacks and whites to feel as if violent crime has increased and that they are significantly more likely than blacks to feel that property crime is on the rise, as of 2011. 9  Many times these reports did not include full descriptive information such as the N for the particular indicator that was being reported. 10  When respondents in my 2010–2012 survey were asked if “crime in your neighborhood is increasing, decreasing, or remaining the same?”, exactly half said “increasing”.

36 

D. C. KUSHNER

60

53

56

50 40

33

33

2011

2012

43

45

44

42

13/14

14/15

15/16

16/17

30 20 10 0

2003

2007

Fig. 2.3  Perceptions of crime in South Africa, 2003–2016/2017 Question (2011–2017): How do you think the level of property/violent crime in your area has changed in the last three years? Question (2003–2007): Do you think that crime has increased, decreased, or stayed the same in your area in the last four years? Note: Figures shown are the percentage of individuals who say that crime has “increased” Note: The 2011–2017 surveys asked about property crime and violent crime separately. The figures shown are the average of these two indicators for respondents who reported that crime had “increased”. Please note that what the VOCS refers to as “violent” crime is referred to as “contact” crime in other parts of the dissertation

Next, I explored citizens’ sense of personal security by using indicators that asked how safe they feel walking around their area during the day and at night. As of 2017, almost 90% of respondents reported that they felt safe walking around during the day. However, after dark this figures drops substantially to a little under one-third of individuals who feel safe walking around at night (29%). As of 2017, not only do more than one-third of people feel that crime is getting worse over time, but most people do not feel secure after dark. I witnessed this deep sense of insecurity during my fieldwork in South Africa, as people wrestled to secure a seat on the minibus taxi11 after work to make sure they arrived to their homes in the township before dark (Fig. 2.4). 11  The minibus taxi is a popular mode of transportation in South Africa. Each taxi carries approximately 16 passengers. The cost of transportation is relatively inexpensive (generally no more than 5 Rand [less than US$1]) and therefore comprises the mode of transportation most often used by poor people.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

37

100 80 60

Day

40

Night

20 0 1998 2003 2007 2011 2012 13/14 14/15 15/16 16/17

Fig. 2.4  Perceptions of personal security, 1998–2017 Question (2011–2017): If you had to walk, how safe would you feel walking alone in your area during the day/when it is dark? Question (1998–2007): How safe do you feel walking alone in your area during the day/after dark? Note: Figures shown are the percentage of individuals who say that they feel “very safe” or “fairly safe”

While substantial percentages of all South Africans do not feel secure outdoors after dark, again, this sense of personal insecurity is most pronounced for Asians. A little over one-third of whites, Africans, and coloreds felt safe outdoors at night in 2011, but for Asians, this figure drops to roughly one-quarter (figure not shown). These findings may suggest the need for the state to do more work to make ethnic minorities feel safe in South Africa. 2.3.1.2.2  Victimization Of all the crimes VOCS respondents were asked about in 2012, individuals most frequently reported being a victim of burglary, followed by ­robbery.12 These findings are consistent with official statistics that show burglary as the most prevalent crime in South Africa at this time. In 2003, 8% of respondents reported experiencing a burglary. As of 2012 this figure has dropped to 5%, but burglary is still, by far, the most commonly reported form of victimization in that sample. It may come as no surprise 12  Because of the various ways in which robbery has been defined over time, VOCS reports only release the robbery figures for 2011 and 2012, since these are the only two years for which the data were comparable. Victimization statistics did not appear to be readily available for the surveys conducted after 2012.

38 

D. C. KUSHNER

then that a majority of respondents (57%) said that burglary is the most feared crime followed by robbery (50%)13 (Victims of Crime Survey 2012). Data from Fig. 2.5 also align with the official statistics in that they generally depict a downward trend in experiences of crime. 9 8

Buglary

7

Robbery

6

Assault

5 4

Theft from Car

3

Car Theft

2

Sexual Offense

1

Murder

0 1998

2003

2007

2011

2012

Fig. 2.5  Victimization in South Africa, 1998–2012 Base Question: Having asked in general about your perceptions of crime, I would like to ask you about your experiences of crime over the past five years and, in particular, within the past 12 months. I am going to read out a list of crimes, and I would like you to tell me if you or any member of your household have been a victim of any of these crimes in the past five years and then in the past 12 months: Question: Housebreaking/burglary (when someone was at home)? Question: Robbery (excluding home robbery and car/truck hijackings) [1]? Question: Assault? Question: Theft out of motor vehicle? Question: Theft of car? Question: Sexual Offense (including rape)? Question: Murder? Note: Figures shown are the percentage of individuals who report that they have experienced these crimes within the past 12 months [1] For robbery, assault, and murder, the base question reads as follows: “Having asked about household crime, I would like to ask you about your personal experiences [underline in the original question] of crime over the past five years, and in particular, within the past twelve months. I am going to read out a list of crimes, and I would like you to tell me if you have been a victim of any of these crimes in the past five years, and then in the past twelve months”

13  Respondents were allowed to give more than one response to the question of which crimes were feared the most.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

39

34

Police don’t listen or care Most people do report crimes to the police

26

Victim feared reprisal from attacker

10

Police wouldn’t have been be able to do anything

9

No police or police station in the area/police station too far

6 0

5

10 15 20 25 30 35 40

Fig. 2.6  Primary reasons for not reporting crime to the police, 2011 Question: Some people say that many crimes are never reported to the police. Based on your experience, what do you think is the main reason that many people do not report crimes like thefts or attacks to the police when they occur? Source: South Africa Afrobarometer Round 5

As with official statistics, it is also possible that individuals may underreport experiences of victimization when participating in surveys. While it is impossible to know for sure whether individuals under report victimization in surveys, there is evidence to suggest that individuals underreport crime to the police. Respondents for the 2011 South Africa Afrobarometer survey were asked what the main reason is for citizen’s failure to report crimes. A plurality of respondents (34%) feel that people fail to report crimes because of police apathy. This suggests that the police in South Africa need to do a better job convincing citizens that they are actually concerned about their safety and security (Fig. 2.6).

2.4   State Responses to Crime in Post-apartheid South Africa One of the first policy documents on crime introduced by the post-­ apartheid state was the National Crime Prevention Strategy (NCPS) of 1996. The strategy included a four-pillar approach to crime prevention. The first focused on revamping the criminal justice system to make it more efficient and, in turn, a deterrent to crime. Second, the state attempted to use the environment to its advantage, focusing on the ways in which smart environmental design could decrease opportunities for crime. Third, there

40 

D. C. KUSHNER

were “public values and education” initiatives which attempted to change the way that people thought about and responded to crime. Fourth, the state saw the effective targeting of trans-national crime as a key component of crime prevention. Throughout the NCPS, the state heavily focused on mediating the social causes of crime. This task required key partnerships with other government departments such as health, education, and welfare. The thinking was that many people who engaged in crime were themselves victims, in that they had been denied access to education and other basic goods and services under apartheid (Singh 2008). Therefore, with education and increased access to opportunities, there would be less of a need for individuals to engage in criminal behavior. The approach was a multi-agency one that encouraged coordination across a range of governmental departments in order to address crime. However, partnerships were not just restricted to governmental ones. The NCPS also focused on building strong partnerships with the commercial security industry and civil society. According to the NCPS, “there are many important partners in the fight against crime. These include, among others, organizations of civil society, particularly business and community organizations, citizens who volunteer for service as Police Reservists, as well as the commercial security industry, which performs a useful role. The role of such players is, in principle, one of partnership with the State” (“National Crime Prevention Strategy” 1996). It is further noted that the role of the commercial security industry was so crucial in the fight against crime, that its role and duties would be further elaborated on in legislation designed specifically for that industry. Throughout the NCPS, there is a call for greater participation by civil society and ordinary citizens in crime prevention. At one point in the document, it says that “to effectively reduce crime, it is necessary to … facilitate real community participation” (“National Crime Prevention Strategy” 1996). This document thus sets the tone for the involvement of citizens and organizations in the prevention of crime in the post-apartheid state. Crime, in effect, became not just the business of police, but ­everyone’s business.14 14  This sentiment was echoed as recently as October 2012 when Commissioner Lamoer for the Western Cape said “Safety is not only a police problem, it needs efforts from all of us. Co-operation must be strengthened and we don’t need to be vigilantes to clear the streets.” October 19, 2012. Cape Argus (South Africa) “New station for hot spot; Second cop shop for Nyanga—the W. Cape’s murder capital.”

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

41

In this sense, the state accommodated citizens and incorporated them into the fold of policing (Buur and Jensen 2004). Moreover, the state attempted to create a clear-cut divide in security functions, making non-state actors heavily responsible for crime prevention and the state more responsible for law enforcement. In 1998, the Department of Safety and Security released the next major policy document to address issues of crime in South Africa. This document was the White Paper on Safety and Security. In the Foreword by the then Minister of Safety and Security, he notes that government’s initial agenda on crime was largely concerned with reforming the police and mobilizing people’s participation in safety and security. Further, the minister noted that while the NCPS continue to frame the development of policy within the department … the emphasis has now shifted towards improved service delivery. This means that the Department’s approach continues to be underpinned by the philosophy of community policing. These have at their heart the principle that a partnership between the police and communities is essential to effective service delivery. (White Paper on Safety and Security 1998)

The above quote shows that the emphasis on citizen involvement in policing remained a key approach to addressing crime and violence in South Africa. But the initial state emphasis on crime prevention soon gave way to a much greater focus on tough enforcement (Dixon 2004; Shaw 2002). To be fair, the White Paper states that it advocates a “dual approach to crime” by focusing on tough enforcement and crime prevention. However, in terms of spending and budget priorities, it seems that the state has mostly supported the enforcement side of the equation. One obvious priority has been on increasing the visibility of the police. In 2007/2008, for example, the South African Police Service’s (SAPS) budget was 38,371  million rand. By 2011/2012, it had increased to 60,658  million rand, an increase larger than that of any other criminal justice institution, including the courts and prisons combined. Not only has the police budget grown, but the number of personnel has also grown substantially. The number of police officers grew from 87,643  in 2002/2003 to 193,692 in 2014/2015. In just a 12-year period, the size of the police force grew by 121%. The increase in the number of officers corresponds to the priorities of SAPS, which has been on visible policing. SAPS has five sectors or programs

42 

D. C. KUSHNER

on which they spend: administration, visible policing, detective services, crime and intelligence, and protection and security services, which is responsible for providing protection for government officials. As of 2010/2011, a plurality of the total budget (43%) was spent on visible policing (Newman 2011). Even when exploring budgetary changes, we find that substantial growth in the visible policing budget has occurred over time, with the total percent increase from 2003/2004 to 2011/2012 coming second only to the percent increase for administration. Visible policing allows cops to quickly respond to crime when it occurs; however, one could also argue that having cops visibly on the beat also contributes to crime prevention. Unfortunately, funding for detective services, which could potentially help to build stronger cases for prosecution, has grown much slower. This is perhaps why even though the police have been more successful in making arrests, conviction rates have not increased much over time and conviction rates for serious crimes stood at less than 5% as of 2007 (Altbeker 2007). The shift to a greater focus on enforcement is, perhaps, understandable in a country that has come under fire for its widespread crime. Amid rising levels of crime and increasing fear of crime, enforcement approaches provide something that can be measured and sold to the public to convince them that the police are making progress. Whereas the efficacy of crime prevention is difficult to measure, enforcement activities (i.e. number of arrests) can be measured and can provide some indicator of police performance. This is particularly important in a context where the police force is under constant pressure to demonstrate that it is improving the security landscape. The official statistics shown above suggest that the police have been making progress with crime on the decline in virtually every area (with the exception of common robbery and commercial burglary). So why then does the use of non-state security remain so popular in post-­ apartheid South Africa? I turn to that subject in the next section.

2.5   Non-state Security in Post-apartheid South Africa As discussed above, the new state—with its weak security capacity, its desire to legitimate itself, and pressure to reduce crime rates—incorporated citizens into the fold of policing. This created space for members of society to participate in ensuring their own safety and security. In addition, with South Africa shifting to a free market economy, there were continued opportunities for the

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

43

operation of the commercial security industry. As Schärf and Nina note, “The irony is that the liberal state was supposed to reduce the need for non-state forms of ordering, but the inability of the transforming state to rise to the level and scope of service delivery has had the opposite effect” (2001, p. 6). This section will discuss the role that each of these types of non-state security structures is playing under the new order. 2.5.1  Market-Based Security Cross-national comparisons of commercial security suggest that the growth of this sector may be the result of three factors: the withdrawal of the state from some of its functions, the growth of mass private property, and the inability of the police to protect citizens from crime (Shaw 1995). In South Africa, the steady growth of the commercial security industry may be a result of all of these factors, but chief among these is the perception that police lack sufficient capacity to protect citizens from crime. Today, the commercial security industry is a key actor in controlling crime and protecting private as well as public spaces (Brogden and Shearing 1993). One way of measuring the persistent salience of crime and security issues in South Africa is by looking at the continued growth of the commercial security industry. Figure 2.7 depicts the total number of registered active security guards in the country from 2001 when these statistics were first recorded. It shows that the number of registered active guards grew from a total of 194,525 in 2001 to 488,666 as of 2015/2016. This represents a percent increase of 151% in 15 years, an even larger increase than 121% in the size of the SAPS over a roughly similar number of years. Similarly, the number of security businesses in operation has increased substantially over time. Figure 2.8 shows that there were 5491 registered active security businesses in South Africa in 2001, compared to over 8000 in 2015/2016, a 61% increase. So it is not simply that individual companies are getting larger over time, but that more competitors are entering the marketplace every year. Of the various types of commercial security companies that exist, those that provide guarding or patrolling services—whether commercial or residential—are by far the most popular with over 6324 such companies in operation as of 2015/2016.15 While 15  Other types of commercial security companies include, for example, those that provide bodyguarding, cash-in-transit services, security consultancy, entertainment/venue control, and car watch, just to name a few. In many cases, a business will provide several of these services and therefore may be counted more than once.

44 

D. C. KUSHNER

450000 400000 350000 300000 250000 200000 150000 100000 50000 0

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Fig. 2.7  Growth of the commercial security industry by number of security guards, 2001–2017 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Fig. 2.8  Growth of the commercial security industry by number of security businesses, 2001–2017

many people rely on technology and alarm systems as a first line of protection, it is clear that businesses and individuals alike are very dependent on guards (many of them armed) to ensure their physical safety and the safety of their property. The steady increase in commercial security guards and commercial security companies over time shows that demand for commercial security has continued to grow over the years. This demand may largely stem from persistent fear of crime, but also an increase in the number of individuals with the means to purchase market-based security.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

45

It is impossible to speak of the commercial security industry in South Africa without mentioning its relationship to the state, mainly via the formalized institution of the Private Security Industry Regulatory Authority (PSIRA). This body was created in 2001 in order to conduct oversight of the commercial security industry. According to the Private Security Industry Regulation Act of 2001, key functions of the authority include: • Granting registration to as well as suspending the registration of security service providers • Developing and maintaining standards related to security provision • Ensuring quality training of security service providers, primarily through accreditation of security training institutions • Processing complaints concerning private security providers and • Protecting private security guards from exploitation The body is governed by a council consisting of a chairperson, a vice-­ chairperson and three additional councilors, all of whom are appointed by the Minister of Police. The council members serve for a term of three years with the option to be reappointed for up to two additional terms. One of the key ways the PSIRA exercises control over the commercial security industry is through inspections. In 2011, a total of 6611 inspections of commercial security providers were carried out in South Africa. Of these inspections, the vast majority were carried out in three provinces, Gauteng (2138), KwaZulu-Natal (2144) and the Western Cape (717). Moreover, the vast majority of these inspections were carried out on security businesses that provided guarding or patrolling services (85% in Gauteng, 84% in KwaZulu-Natal, and 68% in the Western Cape). There are four different type of inspections carried out by the authority. First, the authority conducts routine inspections to ensure that service providers are in line with training and accreditation requirements. First time inspections are thorough inspections conducted for newly registered businesses. Next, the authority carries out infrastructure inspections. Finally, the authority is responsible for employing triggered inspections. Of these various types of inspections, priority is given to infrastructure and triggered inspections. Triggered inspections are important because these are spontaneous inspections that are carried out in response to complaints about security service providers. Therefore, these inspections provide a mechanism for corporations and individuals who rely on commercial security to come into contact with the state and voice their dissatisfaction with the industry.

46 

D. C. KUSHNER

Over time, the state has called on the commercial security industry to be a key partner in the fight against crime and the commercial security industry has, in turn, presented itself as a major ally of the state. For the most part, the relationship between the state and the commercial security industry has been characterized by cooperation. However, there have certainly been and continue to be times of conflict between these two entities. One of the biggest sources of conflict stems from the criminality embedded in the commercial security industry itself. In an interview with a high-ranking police official in Pretoria, he mentioned that the commercial security industry presents lots of problems for the police in South Africa, citing the fact that you often find “fly-by-night” companies that are operating illegally and that employ illegal immigrants.16 Perhaps even more disturbing is the sentiment that many security guards are directly involved in crime, for example, helping to arrange burglaries of the very homes they are supposed to protect.17 Part of the reason for the inspections carried out by the PSIRA is to detect unregistered security business and businesses in violation of other PSIRA regulations (i.e. unregistered or improperly registered firearms). Where these are detected, criminal cases are opened with the South African Police Service. As of March 2010, 648 cases were pending with the SAPS, cases that inspectors of the authority had filed. Of the 648 pending cases, 257 had been opened in the last year. In addition to criminal cases being filed, the Authority often refuses to register security providers or revokes their registration if they are found guilty of criminal offenses. Between 2010 and 2011 alone, the authority refused 11,810 individual security service provider applications due to the applicants being guilty of disqualifiable offenses (Authority 2010/2011). Another 168 individual security service provider registrations were withdrawn after conviction of a criminal offense. In sum, the commercial security industry is a key player in the security atmosphere in South Africa. Its substantial growth over time shows that businesses and citizens alike (and at times even the state) see it as a necessary actor in protecting themselves and their property. But commercial security can be extremely costly, limiting access to the well-to-do. Therefore this service is not one to which the poor are likely to turn. To 16  South African Police Service (SAPS) Head Office Brigadier. Personal interview. March 15, 2011. 17  Member of Johannesburg Mayoral Committee on Public Safety. Personal interview. March 22, 2011.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

47

account for the full range of non-state security actors in post-apartheid South Africa, we must also examine those voluntary, community-based forms of non-state security, what I refer to throughout this book as societal-­based security. 2.5.2  Societal-Based Security Earlier in this chapter I noted that societal forms of non-state security largely arose in black communities under the banner of the “civics” before the transition to democracy. Seekings (2000) notes that most civics died off very shortly after the transition. This was, in part, because the mobilization role that they played in the struggle for liberation was no longer necessary. Yet even though the civics as a broad, overarching organizational structure may have ceased to exist after the transition, some structures that operated under its umbrella (such as street committees) did not. In fact, some scholars have argued that there has been a concerted effort on the part of the state to integrate these non-state structures into state structures, via for example, organizations such as CPFs (Buur 2005). Section 18 of the South African Police Service Act of 1995 stipulates a key role for communities to play in security provision through the creation of CPFs. According to this act, the purpose of CPFs is to formalize a ­relationship between the community and police, improve transparency and accountability of the police, and involve the community in problem identification and problem solving regarding issues of crime. In the immediate aftermath of the 1994 transition to democracy, community policing was seen as way of improving public perceptions of the police and ensuring that the police became a “people’s police”. In practice, however, some forums have often exacerbated tensions and mistrust between the police and community (Brogden 2004; Shaw 2002; Baker 2008). During my fieldwork in South Africa, I was often told by police that community members are problematic because they want to act as if they are the police. Sometimes police even suggested that they questioned citizens’ motives for joining the CPF, as it had been suspected that in some cases criminals themselves were joining in an effort to gain access to information that would be of help to them in their criminal pursuits. From the community side, I frequently heard citizens complain about the police not wanting to listen and fully take their concerns and suggestions into consideration. CPFs are established at the police station level and officially there should be one CPF for every police station. In practice, however, the establishment and activity of CPFs varies widely across regions and even

48 

D. C. KUSHNER

across individual police stations. The CPF is supposed to consist of both police officers (usually, at least, the station commander) and community members, with a community member responsible for holding the position of chairperson. Each CPF has a chairperson, vice-chairperson, and however many additional members as determined by the executive committee of the CPF. Each CPF is governed by a constitution that is supposed to be drafted by members of the executive committee within 30  days of its establishment. In many instances, if the police station covers a wide geographical area, Community Police Sub-Forums will be established in an attempt to deal with crime specific to smaller areas under the police station jurisdiction. Of all the societal forms of non-state security explored here, CPFs are by far the most institutionalized. They embody the most formal relationship between citizens and agents of the state such as the police. As a result, they are perhaps seen as the most legitimate form of societal-based security. However, this is not to say that the state is always in control of or perfectly monitors the behavior of CPFs. In an interview with a CPF chairperson, I was told that the CPF often patrolled the neighborhood with no police, although police are supposed to be a part of all neighborhood patrols. Moreover, although citizens involved with the CPF are only ­supposed to act to help prevent crime, they often react to crime and alleged criminals, taking on more of an enforcement role. In speaking to this same CPF chairperson, I was told that his CPF is effective because, off the record, “criminals will get a beating.”18 So CPFs do not necessarily act lawfully, even though they are most closely associated with those charged with upholding the law. What’s more is that in addition to community members of CPFs sometimes using violent and unlawful means for dealing with criminals, there is some evidence that the police are often aware of this behavior and tacitly endorse it.19 2.5.3  The Good, the Bad, and the Ugly of Non-state Security Until now, we have focused almost exclusively on how security is provided either by the state or through non-state actors that the state authorizes to provide security services. However, this only tells one side of the story of  CPF chairperson. Personal interview. December 30, 2010.  See Steinberg Thin Blue: The Unwritten Rules of Policing South Africa for a vivid account of how police are aware of, yet turn a blind eye to, community violence against alleged criminals. 18 19

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

49

how non-state actors become involved in the provision of security. In fact, it is a top down story that assumes that the state is completely in control of the security atmosphere. In reality, there are many organizations and individuals that participate in security provision that remain beyond the reach of the state. Some more autonomous organizations include those like street committees and neighborhood watches. Street committees originally arose under apartheid. These were local community structures that largely protected black communities from state violence, but also dealt with criminal elements within their own communities since the policed rarely patrolled black areas. While many other locally-based structures under the umbrella of the “civics” died off after apartheid, the street committee has persisted as a viable way of dealing with crime at the local level (Kempa et al. 1999). Neighborhood watches are also locally-based groups of citizens who volunteer their time in an effort to protect their neighborhoods. As the name suggests, street committees are concerned with particular streets or blocks within the community, whereas neighborhoods watches are interested in security issues relevant to a larger community. It is not uncommon, ­therefore, for those members of various street committees to band together to form a neighborhood watch. Street committees and neighborhood watches often function within the ambit of the law, though their relationship to the state is not institutionalized like CPFs. Therefore, it is conceivable that street committees and neighborhood watches may use illegal and undemocratic means to achieve their ends, especially since it is more difficult for the state to monitor their activity. In fact, in the early 1990s, street committees developed a particularly strong anti-crime focus and would use violence to bring criminals into compliance. Under the new order, the challenge is to shape these committees in ways that are consistent with democracy. To assist with this task, the new democratic state employed NGOs to train street committees in practices that were consistent with the new focus on human rights and the rule of law (Singh 2008). On the other end of the non-state security spectrum, we have organizations such as gangs and vigilante groups (Jensen 2008). These groups do not have the permission of the state to operate, nor do they seek it. The most well-known vigilante group in South Africa is People Against Gangsterism and Drugs (PAGAD) which organizes against gangsters and drug dealers in the Cape Flats area of Cape Town. Members of this organization often protest in front of drug dealers’ homes, and often severely

50 

D. C. KUSHNER

beat and/or kill those accused of gangsterism and drug dealing (Dixon and Johns 2001; Gottschalk 2011). Therefore, recognizing the plural nature of security provision in South Africa means acknowledging that security may be provided by entities and individuals that cooperate and coordinate with the state, or by entities or individuals that function autonomously of the state, and that either one of these may employ legal or extralegal means or some mix of the two for achieving their ends. Recognizing the various types of actors involved in the provision of security in this context begins to highlight the potential achievements and pitfalls of non-state security. Non-state security in its many forms can provide services to those who are either not completely satisfied with, or unable to gain adequate access to, state security services. However, non-­state provision of security can also raise several problems that should be of interest to any democratic state. The first is that it may exacerbate inequality, as only the wealthy and elite members of society can purchase and thus gain access to commercial security services (Bayley and Shearing 1996). Therefore, the poor will be left to rely solely on what is considered an under-equipped, ill-trained, and apathetic police force, especially in poor areas where stocks of social capital are not sufficient enough to allow for the production of societal non-state security. The expectation is not that non-state security will be a panacea for underperforming state security. As Baker (2008) notes, the state itself has never provided security on an equal footing, so we should not expect non-state security to achieve this great feat. However, it is worth highlighting the potential for non-state security to exacerbate inequality, and not only with regard to security, since the poor also become cut off from potential middle class networks that could help to improve their lives when the middle class and rich “fortress” themselves behind gated communities and armed security guards (Lemanski 2004). The second issue arises with the possibility of “mob justice” and vigilantism when communities are allowed to participate in security provision (Kempa et al. 1999). In this case, the concern is not with the uneven supply of security, but rather with instances of extralegal violence that perhaps unjustly target individuals and lead to harsh and violent punishment by the community. Extreme forms of punishment administered by local communities were prevalent under apartheid in the form of “necklacing” and were even witnessed recently in spates of violence against foreigners (Lubbe 2008; Mosselson 2010). Societal participation in security provision therefore opens the door to societal administration of what they perceive to be proper forms of punishment and justice, whether legal or not.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

51

Putting aside problems with both commercial and societal forms of non-state security, what is most important to note is that South Africans rely quite extensively on both forms to meet their security needs. It seems that non-state security, at least for the foreseeable future, will continue to play a critical role in the governance of security in South Africa. The question then becomes, what are the political consequences of this provision? I empirically explore this question in the next chapter.

References Adler, G., and J.  Steinberg. 2000. Introduction: From comrades to citizens. In From comrades to citizens: The South African civics movement and the transition to democracy, 1–25. New York: Palgrave Macmillan. Altbeker, A. 2005. Puzzling statistics: Is South Africa really the world’s crime capital. SA Crime Quarterly 11: 1–8. Altbeker, A. 2007. A country at war with itself: South Africa’s crisis of crime. Johannesburg and Cape Town: Jonathan Ball. Baker, B. 2008. Multi-choice policing in Africa. Uppsala: Nordic Africa Institute. Bayart, J.F., S. Ellis, and B. Hibou. 1999. The criminalization of the state in Africa. Oxford: International African Institute. Bayley, D.H., and C.D. Shearing. 1996. The future of policing. Law and Society Review 30: 585–606. Brogden, M. 2004. Commentary: Community policing: A panacea from the West. African Affairs 103 (413): 635. Brogden, M., and C.D. Shearing. 1993. Policing for a new South Africa. Psychology Press. Bundy, C., G.  Adler, and J.  Steinberg. 2000. Survival and resistance: Township organizations and non-violent direct action in twentieth century South Africa. In From comrades to citizens: The South African civics movement and the transition to democracy, 26–51. London: Macmillan Press. Buur, L. 2005. The sovereign outsourced: Local justice and violence in Port Elizabeth. In Sovereign bodies: Citizens, migrants and states in the postcolonial world, 153–171. Princeton and Oxford: Princeton University Press. Buur, Lars. 2010. Domesticating sovereigns: The changing nature of vigilante groups in South Africa. In Domesticating Vigilantism in Africa, ed. T. Kirsch and T. Gratz. Woodbridge: Boydell and Brewer. Buur, L., and S.  Jensen. 2004. Introduction: Vigilantism and the policing of everyday life in South Africa. African Studies 63 (2): 139–152. Center for the Study of Violence and Reconciliation. 2009. Why does South Africa have such high rates of violent crime? Dixon, Bill. 2004. Community policing: ‘Cherry Pie’ or melktert. Society in Transition 35 (2): 251–272.

52 

D. C. KUSHNER

Dixon, Bill, and Lisa-Marie Johns. 2001. Gangs, PAGAD & the state: Vigilantism and revenge violence in the Western Cape. Gibson, J.L., and A. Gouws. 1997. Support for the rule of law in the emerging South African democracy*. International Social Science Journal 49 (152): 173–191. Gottschalk, Keith. 2011. Vigilantism v. the State: A case study of the rise and fall of Pagad, 1996–2000. Harris, B. 2003. Spaces of violence, places of fear: Urban conflict in post-apartheid South Africa. Paper presented on the Conflicts and Urban Violence panel, Foro Social Mundial Tematico, 16–20. Irish, J. 1999. Policing for profit: The future of South Africa’s private security industry. Pretoria: Institute for Security Studies. Jensen, S. 2008. Gangs, politics & dignity in Cape Town. Oxford: James Currey. Kempa, M., R. Carrier, J. Wood, and C. Shearing. 1999. Reflections of the evolving concept of ‘private policing’. European Journal on Criminal Policy and Research 7 (2): 197–223. Kynoch, G. 2005. Crime, conflict and politics in transition-era South Africa. African Affairs 104 (416): 493. Lemanski, C. 2004. A new apartheid? The spatial implications of fear of crime in Cape Town, South Africa. Environment and Urbanization 16 (2): 101. Lubbe, Gerrie. 2008. Victims of xenophobia: African immigrants in South Africa. Escarts d’identite 112: 64–70. Marks, M. 2005. Transforming the robocops: Changing police in South Africa. Scottsville: University of Natal Press. Mayekiso, M. 1996. Township politics: Civic struggles for a new South Africa. New York: Monthly Review Press. Mosselson, Aidan. 2010. ‘There is no difference between citizens and non-citizens anymore’: Violent Xenophobia, Citizenship and the Politics of Belonging in Post-Apartheid South Africa. Journal of Southern African Studies 36 (3): 641–655. National Crime Prevention Strategy. 1996. Ed. Department of Safety and Security. Neocosmos, M. 1998. From people’s politics to state politics: Aspects of national liberation in South Africa. The Politics of Opposition in South Africa: 195–241. Newman, Gareth. 2011. Crime and policing in South Africa. Institute for Security Studies Seminar. Pauw, J.  1991. In the heart of the whore: The story of apartheid's death squads. Halfway House: Southern Book Publishers Johannesburg. Schärf, W., and D. Nina. 2001. The other law: Non-state ordering in South Africa. Lansdowne: Juta. Schönteich, M., and A. Louw. 2011. Crime in South Africa: A country and cities profile.

  NON-STATE SECURITY IN SOUTH AFRICA: HISTORICAL ROOTS… 

53

Seekings, J. 2000. After apartheid: Civic organizations in the “new” South Africa. In From comrades to citizens: The South African civic movement and the transition to democracy, 205–224. London: Palgrave Macmillan. Shaw, M. 1995. Partners in crime? Crime, political transition, and changing forms of policing control. Johannesburg: Centre for Policy Studies. ———. 2002. Crime and policing in post-apartheid South Africa: Transforming under fire. Bloomington: Indiana University Press. Singh, A.M. 2008. Policing and crime control in post-apartheid South Africa. Aldershot: Ashgate. Uniform Crime Reporting Program, FBI. 2017. Crime in the United States. Victims of Crime Survey. 2012. White Paper on Safety and Security. 1998. Ed. Department of Safety and Security.

CHAPTER 3

The Impact of Non-state Security on Perceptions of State Legitimacy in South Africa

Abstract  This chapter is the first empirical chapter. Using original survey data collected on hundreds of South African citizens, and original data from interviews with elected officials, civil society organizations, and police officers, Kushner explores the impact that non-state security reliance and South Africans’ broader experiences of victimization and insecurity have on their perceptions of state legitimacy. She finds that non-state security reliance strengthens perceptions of state legitimacy in some instances while undermining it in others. The finding critically hinges on how ordinary South Africans view the state’s responsibility and role in the provision of security. Keywords  Non-state security • State legitimacy • State arranger • State producer

3.1   The Puzzle This chapter analyzes the relationship between citizens’ non-state security reliance and their perceptions of state legitimacy. Since Hobbes, political philosophers have asserted that state security provision is key to legitimating the state-society relationship. The very notion of the hypothetical “State of Nature” suggests that physical security provision provides the

© The Author(s) 2019 D. C. Kushner, The Politics of Everyday Crime in Africa, https://doi.org/10.1007/978-3-319-98095-9_3

55

56 

D. C. KUSHNER

raison d’être of state formation. Therefore, at least in the Western world, we have taken for granted that the glue that holds the social contract together is the state’s efficient protection of citizens. The notion of state legitimacy hinging on the effective delivery of state security services is problematic in the context of the developing world for three reasons in particular. First, in many developing countries, states lack the necessary capacity to effectively provide many goods and services, including security (Fukuyama 2004; Migdal 1988). This is especially true in much of sub-Saharan Africa, where many states are endemically weak (Jackson and Rosberg 1982). Second, given the legacy of colonialism and widespread corruption in many developing countries, the police force is an oft-hated and mistrusted state institution, undermining its legitimacy. Finally, and in part because of the reasons cited above, in much of the developing world, many public goods and services are provided by a wide range of non-state actors (Kushner and MacLean 2015). This chapter focuses on the last of these three concerns by interrogating whether and how individuals’ reliance on non-state security shapes their perceptions of state legitimacy. I make use of Lipset’s classic conceptualization of legitimacy as “the capacity of the political system to engender and maintain the belief that existing political institutions are the most appropriate or proper ones for the society” (Lipset 1959, 86). Relying on a Weberian notion of the state, I am primarily interested in citizens’ perceptions of the legitimacy of those institutions that are responsible for exercising coercion. Therefore, I assess citizens’ perceptions of three state institutions that are chiefly responsible for this task: the police force, the courts, and the tax agency. Drawing from criminologists Johnston and Shearing (2003) I define security as “personal, physical safety, as well as the safety of … [individuals’] belongings from damage or depredation” (p. 1). Non-state security, then, is simply the assurance of personal physical safety and safety of property that is provided by private actors or actors outside the public sector. South Africa provides an opportune place within which to explore the political consequences of non-state security provision. First, South Africa is a democratic, developing country that has performed well across a number of social and economic indicators since the transition to democracy in 1994, but continues to struggle with crime and public insecurity. Half of respondents (50%) from my 2010–2012 survey feel that crime is increasing and a majority feel unsafe walking around in their neighborhoods at

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

57

night (71%), even though official police statistics show that crime has been decreasing over time (South African Police Service). Second, there is a long legacy of non-state security reliance in South Africa. These non-state actors are many and varied, including numerous commercial security companies, community police forums, neighborhood watch groups, street committees, and other community-based organizations. While many assumed that South Africans would come to solely or primarily rely on state police after the transition to democracy, my survey data show that this is not the case. Finally, South Africa presents a puzzle that is of central importance to this book, that is, how popular perceptions of state legitimacy remain high in spite of continued perceptions that crime and security issues persist or are getting worse. Figure  3.1 shows that the majority of urban South Africans consider institutions of the state to be highly legitimate, and these assessments have remained high over time. Yet, roughly half of urban South Africans also believe that crime is getting worse over time (Fig. 3.2). We are then left with the following question, which is: how do South Africans’ legitimacy perceptions remain afloat even when they perceive the crime and security atmosphere to be persistently perilous? 90 80 70 60 50 40 30 20 10 0

79

Police

83

Courts

78

Tax Agency

Fig. 3.1  Legitimacy of the state in South Africa, 2010–2012 Question: The police always have the right to make people obey the law Question: The courts have the right to make decisions that people always have to abide by Question: The tax department always has the right to make people pay taxes N = 413

58 

D. C. KUSHNER

60

50

50 40 30 20

30 20

10 0

Decreasing

Remaining the same

Increasing

Fig. 3.2  Perceptions of crime, 2010–2012 Question: Do you think that crime in your neighborhood is increasing, decreasing, or remaining the same? N = 414

3.2   Key Hypothesis I posit a security-driven explanation of citizens’ perceptions of state legitimacy. When the state’s capacity to provide security is low, citizens often turn to non-state sources for security. The question then becomes whether citizens’ perceptions of the state’s legitimacy are strengthened or weakened when they rely on non-state sources for this good. Recall, I argue that the state may act as a producer or arranger of security. Moreover, I suggest that individuals’ reliance on non-state security may strengthen or undercut their perceptions of state legitimacy, depending on whether they believe the state should act as a producer or arranger of this good. I argue, as Mcloughlin (2014) does, that whether non-state security reliance boosts or undercuts citizens’ legitimacy perceptions hinges critically on citizens’ normative views of the state’s role in security provision. I assert that individuals’ reliance on non-state security influences citizens’ perceptions of state legitimacy, but the relationship will vary according to how citizens view state responsibility for security provision. Not accounting for citizens normative views about the state’s role in security provision, I would expect those who rely on non-state security to be less likely than those who rely on state security, to see the state as legitimate. Because citizens overwhelmingly perceive security provision as the responsibility of the state, I would expect their legitimacy perceptions to suffer when they are unable or unwilling to turn to the state for this critical public good. However, I am more interested in comparisons among those

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

59

who rely on non-state security, since I suspect that legitimacy perceptions will also vary within this group (Fig. 1.3). When individuals view the state as a producer of security, they believe that the responsibility for the production of security largely rests with the state. For citizens who participate in political life and generate the revenue that states use to govern (e.g. taxes), they may expect basic goods and services to be directly produced by the state. Thus, when citizens rely on alternate sources to fulfill what are viewed as state responsibilities, citizens’ confidence and belief in the state’s right to rule may suffer. I posit that the gap between citizens’ security demands and the state’s security supply may contribute to views of an illegitimate state. When this good is produced by non-state sources, citizens’ needs are met, but they may come to view market or societal security structures as competing sources of authority in which they vest their trust and loyalty. Thus, when citizens rely on non-­ state sources for the production of this fundamental political good, the legitimacy of the state itself may be thrown into question. But non-state security may also improve perceptions of the state’s legitimacy. The reasons for this are twofold. In the first instance, the state provides a constitutional framework within which non-state security actors are allowed to operate. Without the freedom of association and the legal right to bear arms and secure contracts, non-state groups would not be able to coordinate and operate to provide security. Thus, it may be the case that citizens recognize that the state provides the broader legal context within which non-state groups may operate. In other words, the state may be seen as a sort of arranger, while the producers of security are non-state actors. In the second instance, governments may receive credit from their citizens because of the various ways that they may support the non-state actors who produce security. State leaders who openly support non-state forms of security may receive credit for the moral, technical, symbolic, and financial support they provide for these actors. The state may also receive credit as an arranger when it acts in a regulatory capacity. This may particularly be the case if non-state actors are viewed as effective in their provision of security. Therefore, I arrive at the following hypothesis: HYP  Among those who rely on non-state security, those who see the state as a producer, will be less likely than those who see the state as an arranger, to view the state as legitimate.

60 

D. C. KUSHNER

3.3   Data and Methods This chapter empirically tests the relationship between citizens’ reliance on non-state security and their perceptions of state legitimacy in South Africa. It draws on original survey data collected in Cape Town, Johannesburg, and Durban, South Africa between November 2010 and July 2012. The sample consists of 414 ordinary South African citizens 18 years of age or older. Samples of this size yield a margin of error of +/− 5%. South Africa is considered a deeply divided society, one in which race, and increasingly class, substantially shapes individuals’ attitudes and experiences. In order to ensure that the attitudes of all relevant groups were adequately represented within the sample, I oversampled minorities and stratified the sample by race and income, creating a total of 8 strata. Unless otherwise specified, all the results reported in this section will be from the data collected in South Africa during the 2010–2012 time period. Given that a plethora of survey data exists in South Africa, it is worth discussing why I designed and carried out an original survey in this context. Afrobarometer, the largest cross-national public opinion survey in sub-Saharan Africa, contains a battery of questions that assess individual attitudes toward state institutions and actors. At the time that I was developing my research project, however, it did not contain questions that would allow me to capture who South Africans turn to for their security needs. Moreover, the argument that I put forth about the conditional effect of non-state security on state legitimacy (depending on whether individuals see the state as a producer or arranger) could not be tested with Afrobarometer data. The survey simply contained no indicator that would allow me to ascertain what citizens believe the state’s proper role should be in security provision. Therefore, while the Afrobarometer survey contained many questions that would allow me to measure the dependent variable of interest (state legitimacy) it contained no measures of my primary independent variables of interest (non-state security and state arranger). Conversely, the Victims of Crime Survey, which is a nationally representative survey, contained a range of indicators on crime and security (including who individuals rely on for security), but lacked sufficient indicators of legitimacy that would allow me to test the relationship between individual reliance on non-state security and their perceptions of state legitimacy. Both of these surveys, then, while providing a rich source of public opinion data, were not sufficient for an analysis of the relationship I am most

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

61

interested in explaining. Therefore, an original survey with ordinary citizens was in order. This survey not only allows me to analyze the relationship between non-state security and state legitimacy, but it also provides rich descriptive data on individuals’ micro level security experiences.

3.4   Measuring State Legitimacy, Reliance on Non-­state Security, and State Arranger The outcome of interest for this particular study is perceptions of state legitimacy. To measure this outcome, I created a binary variable using the following three indicators: “For each of the following statements, please tell me whether you agree or disagree: • The police always have the right to make people obey the law.” • The courts have the right to make decisions that people always have to abide by.” • The tax department always has the right to make people pay taxes.” The original values of these variables ranged from 0 (strongly disagree) to 4 (strongly agree). I recode the variables so that values 3 are equal to 1 or “legitimate”. This measure of perceived state legitimacy has been used by other scholars studying this topic and has been found to be a compelling way of measuring citizens’ attitudes toward the legitimacy of the state (Levi et al. 2009). Measuring state legitimacy in this way, we see that the state is perceived to be highly legitimate in urban South Africa (Fig.  3.3). Almost 70% of individuals in this context agree that the state should have a recognized right to rule. However, it is worth noting that almost a third of South African urbanites do not see the state as legitimate. I measure reliance on non-state security using the following indicator “Please tell me who you would call on if you or someone in your family had been a victim of crime?” The response categories are “police only”, “police and some other group”, “other groups only”, and “none of these”. Those who reply “police only” are coded as 0, whereas those who reply “other groups only” are coded as 1. For those who reply “police and some other group” I code them as 0 or 1 depending on how they respond to a subsequent question. For those who reply “police and some other group”,

62 

D. C. KUSHNER

Legitimate

69%

Not legitimate

31%

0%

10%

20%

30%

40%

50%

60%

70%

80%

Fig. 3.3  Legitimacy of the state in South Africa, 2010–2012 Question: The police always have the right to make people obey the law Question: The courts have the right to make decisions that people always have to abide by Question: The tax department always has the right to make people pay taxes N = 413

they are then asked: “If you were a victim of a crime, who would you FIRST call for help: the police or some other group?” For those who say they would call “the police” first, I code them as 0, and for those who say they would call “some other group” first, I code them as 1. Those who respond “none of these” are coded as missing. Because my theory suggests that the impact of non-state security on state legitimacy will be conditional on how citizens view the state’s responsibility in security provision, I also include a variable called state arranger. To measure whether citizens view the state as a producer or arranger of security, I employ the following question: “Which of the following statements is closest to your view? Choose Statement 1 or Statement 2. Statement 1: The state should take the main responsibility for protecting citizens from crime. Statement 2: The state should mainly provide support to private groups such as the commercial security industry or neighborhood watches so that they may protect citizens from crime. I code those who “agree” or “strongly agree” with statement 1 as 0, or “state ­producer”. I code those who “agree” or “strongly agree” with statement 2 as 1, or “state arranger”.

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

63

3.5   Controls In addition to the impact of non-state security, I also examine the effect of other security-related variables, including contact and property victimization and personal insecurity. My model also controls for and tests a number of competing explanations of state legitimacy. While I am primarily interested in determining the impact of non-state security and other security-­related variables on state legitimacy, prior literature suggests the need to account for citizens’ evaluations of the state, as well as the impact of affective bonds on perceived state legitimacy. I control for these two alternate explanations by including a measure of police performance evaluations and partisanship, respectively. Finally, I control for age, gender, lived poverty status, education, employment status, city, and race.

3.6   Explaining State Legitimacy To explain the impact that reliance on non-state security has on perceptions of state legitimacy, I employed a logistic regression analysis. Table 3.1 presents the pooled estimation results of perceived state legitimacy. Model 1 includes only the non-state security, state arranger, and interaction variables. The coefficient on the interaction term is positive and statistically significant, meaning that individuals who rely on non-state security and see the state as an arranger of that good, are more likely to perceive the state as legitimate. The interaction term is significant at the 0.05 level. Model 2 includes all of the security variables of interest to this study. The interaction between non-state security and state arranger retains its significance. In addition, results show that those who are victims of violent crime are also less likely than non-victims to see the state as legitimate. The full model is depicted in the final column. The interaction term remains positive and statistically significant. The key independent variable (interaction term) maintains its significance after controlling for police performance evaluations, partisanship, and a full range of demographic variables. To further support my claims, I examine the predicted probabilities for those who rely on non-state security across different values of the state arranger variable. When examining the predicted probabilities, I find that for those who rely on non-state security, the probability that one sees the state as legitimate increases from 45% to 64% when an individual moves from seeing the state as a producer to seeing the state as an arranger of security (Fig. 3.4). This is a 19 percentage point increase, which means

Table 3.1  Pooled logistic regression output on perceptions of state legitimacy in urban, South Africa Independent variables Non-state security State arranger Non-state security * state arranger Contact victimization

Model 1

Model 2

−1.166* (0.639) −0.984 (0.615) 1.439** (0.727) –

−0.355 (0.286) −0.495 (0.339) 1.029* (0.555) −2.184*** (0.472) −0.179 (0.508) 0.044 0.116 –

Property victimization

–

Personal insecurity

–

Police performance evaluations

–

ANC partisan

–

–

Age

–

–

Male

–

–

Lived poverty

–

–

Black (reference) White

–

–

Colored

–

–

Asian

–

–

Employed

–

–

Education

–

–

Johannesburg (reference) Cape Town

–

–

Durban

–

–

Constant R2 −2 log likelihood

1.705*** (0.544) 0.024 301.719

1.366*** (0.297) 0.105 446.608

N = 358 Standard errors in parentheses Statistical significance denoted by ***p ≤ 0.01; **p ≤ 0.05; *p ≤ 0.10

Full model −1.637** (0.706) −1.204* (0.694) 1.985** (0.818) −2.026*** (0.635) −1.268* (0.729) 0.125 (0.172) 0.379* (0.213) −0.046 (0.478) 0.018 (0.013) 0.201 (0.320) 0.425 (0.299) −0.792 (0.599) −1.355** (0.616) −0.417 (0.606) −0.141 (0.351) 0.188 (0.134) 1.068** (0.429) −0.137 (0.485) 0.287 (0.765) 0.166 259.475

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

Fig. 3.4  Conditional effect of non-state security on perceived state legitimacy

0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00

65

0.64 0.45

Producer

Arranger

that individuals who rely on non-state security and see the state as an arranger have a 42% greater probability of seeing the state as legitimate when all covariates are held constant. The next security-related variable, contact victimization, also has a major effect on how legitimate citizens perceive the state to be. This result suggests that when people become victims of violent crimes like rape, assault, robbery, and murder, their perceptions of state legitimacy plummet. Property victimization also suppresses perceived state legitimacy, but to a lesser degree than violent crime. Next we see that evaluations of the police also factor into citizens’ assessments of the state. The positive and statistically significant coefficient on the “police performance evaluations” variable tells us that when people feel the police do a good job preventing crime, apprehending criminals, and making people feel safe, they are more likely to feel that the state has a legitimate right to rule. Thus, while the non-state provision of security may contribute to legitimacy perceptions, the performance of the state itself still matters. Of the demographic variables included in the full model, only two are significant. In particular, I find that coloreds are substantially less likely than blacks to see the state as legitimate. In fact, the coefficients on the white and Asian variables are also negative (though not significant) suggesting that all minorities are less likely than the majority black population to see the state as legitimate. This result comes as no surprise, as members of the majority population are usually more likely to identify with institutions of the state. Finally, the logistic regression analysis shows that residents of Cape Town are significantly more likely than residents of Johannesburg to see the state as legitimate. Being a resident of Cape Town (as compared to a resident of Johannesburg) leads to more than a twofold increase in the

66 

D. C. KUSHNER

odds of seeing the state as legitimate. What’s interesting is that these results exist despite the fact that Capetonians were more likely than those from Johannesburg to report being victims of all but one crime that I account for in my survey and that they were almost twice as likely to believe that crime was increasing as of 2012. What may account for the difference, however, is the drastically higher scores Capetonians give the police on performing their duties as compared to those from Johannesburg. Before moving on to the next model, readers may be interested in knowing how different types of non-state security affect perceptions of state legitimacy. Does reliance on market versus societal forms of nonstate security make a difference for individual perceptions of state legitimacy? To test this possibility, I ran a difference of means test using a dummy variable for commercial non-state security. The purpose was to examine whether those who rely on commercial security are substantively different in their legitimacy perceptions from those that rely on societal forms of non-state security such as street committees, neighborhood watch schemes, and community police forums. The difference of means was significant at the 0.01 level, showing that individuals who rely on commercial security are more likely than those who rely on societal security to see the state as legitimate. This relationship perhaps signals something important about the legacy of state-society relations in South Africa and how different segments of society had different relations with the state in the past. Under apartheid, the commercial security industry worked very closely with the state, often to supplement its understaffed public police force. In fact, the commercial security industry was given almost parallel policing powers under apartheid, and some would argue that the distinction between the two became increasingly blurred during that time (Brogden and Shearing 1993). Because this formal and institutionalized relationship between the state and the commercial security industry persists, individuals might more readily credit the state with leveraging security resources from the security industry. In contrast to the commercial security industry, some societal-based security structures such as street committees developed in direct opposition to the state under apartheid. Because these organizations were directly involved in the struggle for liberation, their relationship to the state was inherently hostile. This legacy, along with the fact that these societal-based groups are relatively more autonomous from the state, may mean that individuals who rely on this form of security are less likely than those who rely on the market-based form to credit the state with legitimacy dividends when they do.

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

67

3.7   Testing the Effect of Non-state Security Reliance on Perceptions of State Legitimacy Using Afrobarometer Data The results discussed above lend credence to my theory on the relationship between non-state security provision and state legitimacy. But how valid and reliable are these results? The data collected during the 2010–2012 period constitutes a large-N survey with 414 respondents across three urban areas. However, one may want to know how well these findings hold when using a larger sample that is more representative of the country under study. One may also be curious as to how the “non-state security” variable performs when other variables that tap more substantive dimensions of legitimacy are included in the model. The data collected during the 2010–2012 time periods incorporates variables that test for alternative explanations, but most of the variables represent procedural determinants of legitimacy. Therefore, to check the robustness of these results, I triangulate on the same research question and run a similar model using Afrobarometer data. Afrobarometer is the largest, cross-national public opinion survey being conducted in sub-Saharan Africa. Afrobarometer has been conducting surveys in Africa for over ten years, with the most recent round of surveys being carried out in 35 sub-Saharan African countries. The survey ­measures citizens’ attitudes toward democracy, civil society, and markets in Africa. The most recent round of Afrobarometer data that contain my main independent variable were collected in South Africa from October to November 2011. Unfortunately, the Afrobarometer survey does not have a question that allows me to gauge whether individuals see the state as a “producer” or “arranger”. Thus, I am not able to test the conditional hypothesis that I propose about the interaction between the non-state security and state arranger variables and their impact on perceptions of state legitimacy. However, I am still able to look at the impact of non-state security and other security variables on state legitimacy. This data contains a nationally representative sample of 2400 adult South African citizens. Samples of this size yield a margin of error of +/− 2% and thus provide a sample large enough to test the robustness of the results discussed above. The results from the logistic regression model using Afrobarometer data validate support for security-driven explanations of state legitimacy (Table 3.2). As in the pooled results shown above, the coefficient on the non-state security variable is negative. Here that variable is also statistically

68 

D. C. KUSHNER

significant, suggesting that those who rely on non-state security are less likely to view the state as legitimate. Unfortunately, the Afrobarometer survey does not have a question that allows me to gauge whether individuals see the state as a “producer” or “arranger”. Thus, I am not able to test the conditional hypothesis that I propose about the interaction between the non-state security and state arranger variables and their impact on perceptions of state legitimacy. However, the statistically significant effect of nonstate security holds up, at least in the first two models depicted in Table 3.2. Moreover, the effect of victimization remains an important predictor of state legitimacy and also proves to be significant in the Afrobarometer model. The results from the logistic regression model using Afrobarometer data validate support for security-driven explanations of state legitimacy. Model 1  in Table  3.2 includes only security-related variables. The non-­ state security coefficient is negative and significant at the 0.01 level, suggesting that those who rely on non-state security are less likely than those who rely on state security to see the state as legitimate. This is exactly the relationship I expect to find when not accounting for individuals’ normative views of state security provision. In this model, contact victimization is also negative and statistically significant. Those who have been victims of violent crime, are therefore less likely than non-victims to look favorably upon the state, as in the urban South Africa data. The middle column of Table 3.2 depicts the full model, which includes the security variables, variables to test for competing explanations, and demographic variables. Non-state security and contact victimization both remain negative and statistically significant. Regime characteristics are also important in the model. Those who express satisfaction with democracy are more likely than those who do not to see the state as legitimate. Surprisingly, “economic performance evaluations” does not reach statistical significance, which, importantly, was included to tap into the economic bases of legitimacy. Since my original model primarily measured the supply of political goods, I thought it important to account for substantive economic goods in this model, especially given South African concerns about economic inequality and unemployment. This variable, however, fails to reach statistical significance. This suggests that, at least in current day South Africa, the legitimacy of the state largely hinges on the perceived supply of political goods. Finally, of the demographic variables included in this model, only age and employment status are significant. Specifically, older people are more likely to see the state as legitimate, while the employed are less likely. We may expect older individuals—who came of age during the struggle for liberation—to be more likely than young people to see the state as legitimate, yet

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

69

Table 3.2  Afrobarometer logistic regression output on perceptions of state legitimacy, South Africa Independent variables Non-state security Contact victimizationa Property victimization Personal insecurity Economic performance evaluations Satisfaction with democracy

Model 1 −0.354*** (0.095) −0.338** (0.142) −0.034 (0.104) −0.012 (0.031) – –

ANC partisan Age

–

Male

–

Urban Lived poverty

–

Black (reference) White

–

Colored

–

Asian

–

Employed

–

Education

–

Constant R2 −2 log likelihood

0.646*** (0.066) 0.013 3072.826

Full model w/o partisanship −0.329*** (0.098) −0.307** (0.147) −0.043 (0.106) −0.020 (0.034) −0.072 (0.069) 0.210*** (0.052) – 0.232** (0.119) −0.112 (0.088) 0.000 (0.103) 0.079 (0.064) 0.128 (0.146) 0.026 (0.133) 0.288 (0.234) −0.160* (0.097) −0.015 (0.033) 0.397* (0.215) 0.031 2983.847

Full model w/ partisanship −0.168 (0.117) −0.390** (0.168) 0.046 (0.124) −0.005 (0.041) −0.008 (0.084) 0.137** (0.061) 0.454*** (0.164) 0.244* (0.142) −0.046 (0.104) −0.043 (0.119) 0.104 (0.076) 0.597*** (0.230) 0.444** (0.185 0.229 (0.339) −0.073 (0.115) −0.030 (0.039) 0.000 (0.281) 0.028 2141.086

N = 2400 Standard errors in parentheses Statistical significance denoted by ***p ≤ 0.01; **p ≤ 0.05; *p ≤ 0.10 a The question wording for these indicators differ from the question wording for indicators in Table 3.1

70 

D. C. KUSHNER

it is not exactly clear why the employed would be less likely than the unemployed to see the state as legitimate. After all, we might expect that those who are unemployed would have more grievances against the state. But perhaps this has something to do with the oft-proposed taxation-­ representation link, and suggests that taxpayers (the employed) perhaps see institutions of the state such as the tax agency as illegitimate because it takes their revenue, but in turn does not satisfactorily provide goods and services, or represent their interests. The final model shown in Table 3.2 includes all the variables from the previous model, but it also includes a variable to account for ANC partisanship. Unlike in the urban South African sample, in this nationally representative sample, the ANC partisanship variable is positive and statistically significant, suggesting that ruling party partisans are more likely than opposition partisans to see the state as legitimate. More importantly for this research, once the partisanship variable is included, the non-state security variable ceases to be statistically significant. In this Afrobarometer sample, which includes rural respondents in addition to urban ones, it seems as though the non-state security variable does not have an independent affect apart from ANC partisanship. This may especially be the case in rural areas where those who do rely on non-state security mostly rely on societal non-state security (Kushner and MacLean 2015). This finding may therefore support the argument made in Kushner (2015) that those who rely on societal non-state security in South Africa are encouraged to support the ruling ANC. Do the results described above hold when accounting for higher-level units of analysis? To see if they do, I reran the models displayed in Table 3.2 with provincial fixed effects (not shown). I find that the results are similar with regard to the sign and statistical significance of most coefficients. Moreover, individuals in practically every province are more likely than those who reside in the Western Cape to see the state as legitimate. The results from my 2010–2012 survey show that a strong majority of individuals (69%) who reside in the city of Cape Town see the state as under the control of the ANC government (Table not shown). Because Cape Town (and the Western Cape province more generally) is ruled by the opposition Democratic Alliance party, I suggest that residents of the Western Cape are less likely to see the South African state as legitimate because they believe that institutions of the state are controlled by the ruling ANC party. Those who hold such views may withdraw legitimacy dividends from the state as a result.

  THE IMPACT OF NON-STATE SECURITY ON PERCEPTIONS OF STATE… 

71

3.8   Conclusion This research has tapped into a key dimension of the micro-foundations of statehood. While much of the literature on the state has argued that the provision of security is a paramount public good, this research has empirically tested the importance of this criterion for ordinary citizens in an important developing context. In particular, it has shown that ordinary citizens see the provision of security as an important determinant of legitimate statehood. Moreover, the results presented here move beyond conventional works that have simply looked at the importance of state security provision in legitimating the state-society relationship. Interestingly, these results highlight the importance of non-state actors in shaping individual attitudes toward the state. My research shows that individual reliance on non-state security can boost perceptions of state legitimacy, but only when individuals believe the state should act as an arranger of that good. These results demonstrate that citizens have different views about the state’s proper role in service delivery, and that their perceptions of state legitimacy are conditioned by this factor. In conclusion, this research shows that attitudes toward the state depend not just on whether key goods like security are provided, but also by whom they are provided. It shows that actors from the non-state sphere have an important role to play in shaping the relationship between state and society, and that their services may strengthen or undercut perceptions of state legitimacy.

References Brogden, M., and C.D. Shearing. 1993. Policing for a new South Africa. Psychology Press. Fukuyama, Francis. 2004. State-building: Governance and world order in the 21st century. Ithaca: Cornell University Press. Jackson, R.H., and C.G.  Rosberg. 1982. Why Africa’s weak states persist: The empirical and the juridical in statehood. World Politics: A Quarterly Journal of International Relations 35 (1): 1–24. Johnston, L., and C. Shearing. 2003. Governing security. Explorations in policing and justice. London: Routledge. Kushner, Danielle C. 2015. Non-state security and political participation: Reinforcing ruling party support in South Africa. Africa Today 62 (1): 107–135. Kushner, Danielle C., and Lauren M. MacLean. 2015. Introduction to the special issue: The politics of the nonstate provision of public goods in Africa. Africa Today 62 (1): vii–xvii.

72 

D. C. KUSHNER

Levi, M., A.  Sacks, and T.  Tyler. 2009. Conceptualizing legitimacy, measuring legitimating beliefs. American Behavioral Scientist 53 (3): 354. Lipset, S.M. 1959. Some social requisites of democracy: Economic development and political legitimacy. The American Political Science Review 53 (1): 69–105. Mcloughlin, Claire. 2014. When does service delivery improve the legitimacy of a fragile or conflict-affected state? Governance 28 (3): 341–356. Migdal, Joel. 1988. Strong societies and weak states: State-society relations and state capabilities in the third world. Princeton, NJ: Princeton University Press. South African Police Services. Official crime statistics.

CHAPTER 4

Assessing the Impact of Non-state Security, Victimization, and Insecurity on Social Capital and Collective Action in South Africa

Abstract  This chapter is the second empirical chapter. This chapter tests the influence of non-state security, victimization, and insecurity on joining and collective action in South Africa. While all three security factors play some role in shaping popular participation, the impact of these variables differs according to the type of participation under consideration. Keywords  Non-state security • Social capital • Collective action • Victimization • Insecurity

4.1   Introduction The previous chapter examined how individuals’ reliance on non-state security shaped their political attitudes. Specifically, it found that those who rely on non-state security may be more or less likely to see the state as legitimate, depending on what they perceive the proper role of the state to be in service delivery. That chapter therefore established the key role that non-state actors may play in shaping popular perceptions of state legitimacy. The empirical chapters that remain shift to a focus on political behavior and examine the ways in which security factors influence individuals’ political participation. I explore three key security variables—non-­ state security provision, victimization, and insecurity—and how they affect the ability and/or will of citizens to build social capital, act collectively, protest,

© The Author(s) 2019 D. C. Kushner, The Politics of Everyday Crime in Africa, https://doi.org/10.1007/978-3-319-98095-9_4

73

74 

D. C. KUSHNER

and vote. The following chapters interrogate whether the ­non-­state provision of a key public good (security) encourages individuals to withdraw from or engage more fully with other citizens and the state. They also explore how feelings of personal insecurity and becoming a victim of crime shape individuals’ willingness to be a part of political associations and processes. I emphasize the impact of non-state security because very little is known about the political consequences of non-state security provision, even though it is widely used across Africa (see Chap. 1). Studies by political scientists and donor organizations have probed the political consequences of the non-state provision of health, education, and water, but there is virtually no work that examines how the non-state provision of security shapes political behavior (Batley and Mcloughlin 2009; Moran and Batley 2004; Rose 2007; Sacks 2012). Scholars have often implied that the extent and quality of state-provided goods (including security) could impact citizens’ political behavior and attitudes (Levi et al. 2009). But it is a completely open question as to how the non-state provision of goods shapes individuals’ political engagement. Likewise, victimization becomes important in places like South Africa where crime is widespread (see Chap. 2). Yet, most studies do not examine the effect of reported victimization, but rather focus on how perceptions of crime shape individuals’ attitudes and behavior (Lemanski 2004; Louw 1997; Louw). Therefore, it seems both timely and relevant to investigate the political consequences of reported victimization in this context. Chapter 3 showed that individuals who report being a victim of a violent crime are less likely to see the state as legitimate. This chapter will explore whether this factor also has an impact on political behavior. Finally, I assess whether and how a sense of personal insecurity affects political participation.

4.2   Existing Works Exploring the Relationship Between Service Provision and Political Participation There is limited extant research on the behavioral consequences of non-­ state service provision. Most of the empirical work that examines the impact of service provision focuses on how state service provision affects political participation. For example, MacLean (2011) explores how the retrenchment of the state in Africa shapes political behavior. In particular,

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

75

she investigates the effect of individual reliance on public education and health services on citizens’ political participation. MacLean examines electoral and non-electoral forms of participation, including voter registration and turnout, political contacting, and joining. Using Afrobarometer survey data from 18 countries, she finds that individuals who report having more frequent experiences with public schools and clinics are more likely than those who have no experience to participate in both electoral and non-electoral politics. According to MacLean, the mechanism linking citizens’ experience with public services and their greater levels of participation is the quality of experience citizens have when making use of these services. She argues that “the experience of the declining quality of publicly provided health and education services … mobilized greater citizen participation and engagement” (MacLean 2011, 1256). While it is not possible to do with the Afrobarometer data that MacLean uses, it is important to disaggregate this “no experience” group, as individuals’ reasons for not having experience with public services may be multiple and varied. Survey respondents may have no experience with public services because they choose to rely strictly on private ones, or they may have no experience because they are locked in a structural and/or physical position that makes access to public services (and private ones for that matter) prohibitive. Citizens in both instances may view the state as being responsible for their basic needs, but I would expect there to be a difference in how they react to and engage with the state based on whether their needs are not met at all, or if they are met, but by actors outside of the state. In the latter case, it would be interesting to know how citizens’ political actions are affected by their reliance on non-state sources of provision. On one hand, citizens may withdraw from participation in the political sphere altogether, convinced that no amount of public participation will yield the quality of public services they want. One the other hand, citizens may continue to press the state to deliver quality service that they may one day be satisfied to use. Finally, MacLean speaks of a two-tiered social service system, one in which urban citizens have largely exited from state services and primarily have their services provided through private entities, and another tier in which rural citizens who lack the financial means to turn to private companies, begrudgingly rely on public services. So while we find a public/ private, rural/urban dichotomy here, what seems to be common to both the urban and rural citizens in this story, is that they are largely dissatisfied with the state. In the case of the rural, largely poor citizenry, the argument

76 

D. C. KUSHNER

is that this dissatisfaction leads to increased political participation. But in the case of the urbanite, this dissatisfaction leads to an exodus from public services and, by extension, political engagement. The question of why urban dwellers—who continue to pay taxes to finance public services— would not advocate for better services that they could potentially benefit from in the future, remains open. Why would the more well-to-do not be motivated to mobilize and “throw out the rascals” who are perceived to be squandering their tax dollars and providing poor public services? This might especially be the case when the politicians in power are from a different ethnic group than those who are wealthy enough to secure services outside of the state through the private sector. More recently, Hern (2017) researches the impact of state service provision on collective behavior and political participation in Zambia. She finds that the fewer government services that individuals have access to, the more likely they are to engage in collective action. Unlike in developed states where inclusive policies spur collective action, Hern argues, citizens in weak African polities engage in collective action when they have been excluded from public services. In short, people pick up and provide for themselves in areas where the state cannot or will not. The idea that citizens might band together and pool resources to meet their own needs when the state fails to is intuitive. What is less clear, is why citizens who provided their own services, in private, because of state neglect, would then go and participate in the formal politics of the state. Is it that in joining together to meet their own needs citizens build a sense of efficacy that they believe they can then use to pressure the state into providing the services they once did? Do citizens engage simply to maintain what few services they are already getting from the state? There are many reasons why individuals might do this, but the explanations don’t fully come through in this article. In both MacLean (2011) and Hern (2017), people are driven by need; in the former, the need for better quality public services, in the latter, the need to fill “in the gap the state left” and provide those services the state did not. In MacLean’s account, citizens join together to lay collective claims on the state for improved service delivery. In the context of Zambia, collective action is how citizens directly ensure that their full range of basic needs is met. In a seemingly hybrid approach to service delivery, Zambians supplement weak state supply of basic services by pooling their skills, labor, and resources to provide services their communities find lacking. Yet one can’t help but to wonder whether all citizens in Zambia must fill

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

77

in the gap left by the state, or just some. In other words, if some citizens are able to rely more fully on the state to meet their basic needs, while others must rely to a much greater extent on kith and kin to do so, it would be worth knowing how the participation levels of the former compare to the latter. In recent years, there has been some important empirical work that has begun to directly assess the political impact of non-state service delivery. Cammett and MacLean (2014) was perhaps the first political science text to truly develop a theoretical framework of non-state provision, and examine the consequences of such provision in developed and developing contexts. Their edited volume examines the impact of non-state social welfare on equitable access to social services, accountability, and state capacity. While the impact of non-state provision varies across different policy areas and geographic locations, at the individual level, Cammett and MacLean show that the fragmented service delivery context can make it difficult for citizens to know whom to hold accountable. In a similar vein, Kushner and MacLean (2015), focusing specifically on non-state provision in Africa, show that this type of provision can often lead to unequal outcomes, and further relegate vulnerable populations such as ethnic minorities and the poor, to the margins of service delivery access. The papers in their edited special issue collectively show that even non-state provision generally works for well-connected and well-to-do citizens in Africa, and that it really only encourages the types of political participation that maintains the status quo. Finally, Brass (2016) makes an important contribution to this emerging area of research with her book Allies or Adversaries: NGOs and the State in Africa. Brass investigates how NGOs’ service delivery work affects legitimacy, state capacity, governance and territoriality. She finds that NGOs play such a central role in governance and in expanding the territorial reach of the state, that the dividing line between NGOs and the state has actually become blurred. With a more positive assessment of state-NGO relations than has typically been the case, Brass states that “if we consider NGOs to be part of the organizational form of the state, it is logical that NGOs do not undermine state legitimacy” (2016, 44). Brass shows that NGOs generally enhance each dimension of stateness that she explores in her research. She demonstrates that citizens who benefit from the services that NGOs provide typically do not arrive at a different view of the state at all, or, in cases where they do, they tend to see it more favorably. But as Brass briefly acknowledges in quoting Jelinek

78 

D. C. KUSHNER

(2006), it is possible that the impact on citizens attitudes toward the state will vary based on the type of non-state service they receive. She adds that in the case of Afghanistan, the area under study by Jelinek, individuals “came to expect security from the state, but other services from NGOs” (Brass 2016, 214). Therefore, the non-state delivery of services would only undermine individuals’ perceptions of the state when they believe those services should be provided by the state.

4.3   Security and Popular Participation The vast majority of literature explaining political participation has ignored citizens’ everyday security experiences. There are a few works where scholars have shown how fear and feelings of public insecurity can suppress political behavior (Pérez 2003; Salamon and Van Evera 1973). For example, Salamon and Van Evera (1973) demonstrate how fear of physical and economic reprisal suppressed black political participation in the American South during the era of Jim Crow, while Perez, working in the context of El Salvador and Guatemala, has shown how public insecurity increases support for authoritarianism (and, by extension, decreases support for democracy and democratic practices). Similarly, most of the work on Africa that addresses the relationship between violence and participation focuses on people’s experiences with extraordinary instances of political violence. For example, Blattman (2009) shows that ex-combatants who were conscripted via abduction in Uganda were much more likely to be engaged politically after the conflict ended. He shows that those who were abducted were more likely than those who weren’t to vote and to engage in community leadership, suggesting that victimization encourages political participation. Along similar lines, Olivier (1991) shows how South Africans engaged in collective action in response to the repressive apartheid state. Thus, there has been some work which has looked at the impact of security-related factors on participation, but not much. The work by Blattman in particular is instructive, but it focuses on experiences of violent political conflict whereas my interests lie in understanding the political consequences of everyday forms of crime and violence. The discipline is therefore in need of research that investigates how victimization during more ordinary political times might shape political behavior. This research agenda is especially germane to Africa, a region that is rapidly urbanizing and democratizing, and will no doubt begin to witness conflict and violence that is more criminal than political in nature in the future.

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

79

The work that has addressed this question most directly and thoroughly within political science is Bateson (2012). Bateson (2012) examines the effect of crime victimization on political participation in countries spanning five continents: Europe, Asia, Latin America, Africa, and North America. She finds that there is a worldwide association between crime victimization and political participation (Bateson 2012). In particular, Bateson argues that being a victim of crime makes individuals more likely to participate in all forms of politics included in her study. Importantly, Bateson argues that all crime victimization, whether serious or petty, violent or non-violent, increases all forms of political participation. Bateson hints at several reasons why victimization might cause individuals to be more politically active, including post-traumatic growth theory, instrumental reasons, and emotional and expressive factors. Bateson’s work presents the first attempt, to my knowledge, to systematically and empirically explore the effect of everyday violence on political behavior, and she does so in a way that is far-reaching and compelling. Therefore, I would like to build on Bateson, delving more deeply into how we might account for the relationship between victimization and participation, but also generating a broader security-driven model of participation that accounts for other security concerns, most notably, the impact of non-state security provision.

4.4   Victimization and Political Participation When it comes to the relationship between victimization and participation, I generally expect that victimization will make individuals more likely to engage in political acts. However, I argue that individuals’ decisions to join associations may be contingent on trust. Participation in associational life is about citizens working together through organizations to form and address their collective interests. Because this form of participation requires extensive interaction with fellow citizens, issues of trust become important and perhaps key to understanding the relationship between victimization and participation. For example, crime victimization has been shown to suppress trust (Delhey and Newton 2003). Therefore, we may expect citizens who have been victims of crime to be less trustful of others, and, in turn, less likely to join in local organizations with them. The extremely important work by Putnam  et  al. (1994) has shown the significance of trust and other forms of social capital for democracy. In this same vein, I suggest

80 

D. C. KUSHNER

that trust becomes an essential element for developing and sustaining a key feature of modern democracies; political participation. As Putnam notes, “the greater level of trust in a community, the greater the likelihood of cooperation” (1994, 171). In a deeply divided country like South Africa, I would argue that it is not only important to explore whether trust exists, but also the locus of that trust. I suggest that understanding this factor will be important for whether victimization will lead people to be more or less likely to join. Putnam hints at the importance of the locus of trust in his discussion of networks and whether they are bridging or bonding. Importantly, he notes that segregated networks (bonding) foster cooperation within the group. Because of the racially segregated nature of life in South Africa, it would seem that most organizations would be of the bonding type. However, I suggest that intra-group and intra-community trust is not a given, and that this type of trust must be present for individuals to join local associations. To illustrate the importance of the locus of trust, let’s examine a public opinion question about criminals. When asked where they believe criminals reside, many white South Africans report that criminals reside outside their communities, whereas blacks largely report that criminals reside within their communities. So those who feel as if criminals come from within their communities (and therefore may have less trust in neighbors) may be less likely to join in local organizations and engage in collective action. Conversely, those who believe that crime comes from outside their communities (and are therefore more trusting of their neighbors) might be more likely to band together in organizations to keep criminals out. Therefore, I expect social trust to mediate the relationship between victimization and joining. My expectations for the relationship between victimization and participation are as follows: Hyp. 4.1  Individuals who have been victims of crime will be less likely than those who have not to join associations, unless they trust their neighbors, then they will be more likely. Hyp. 4.2  Individuals who have been victims of crime will be more likely than those who have not to engage in collective action.

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

81

I expect victims of crime to be more politically engaged in collective active than non-victims, using their activism as a way to express grief, to demonstrate discontent with the state, and to push for substantive changes in the crime and security policy environment.

4.5   Theorizing the Relationship Between Non-­ state Security and Participation Early political philosophers saw the provision of state security as essential for cementing the state-society relationship, but they are not the only ones. The idea that security is a chief public good to be provided by the state is born out in people’s everyday expectations of public service provision. For example, Bratton and Chang (2006) interrogate the aspects of stateness that matter most for democratization in Africa. They find that state capacity, and in particular law enforcement capacity, is a critical determinant of the quality and supply of democracy at the macro level. From their individual level analysis, however, Bratton and Chang conclude that “if the state can attain political stability, regulate conflict within its borders and protect the citizenry from criminals, then people will conclude that democracy is being supplied. [They also note that] contrary to conventional wisdom, Africans do not refer to the delivery of official welfare services in forming positive opinions about the desirability of democracy” (2006, 24). This finding demonstrates the importance of security provision from the vantage point of ordinary citizens. I argue that ordinary citizens have well defined ideas about the kinds of obligations and responsibilities for which the state should be chiefly responsible. Assuming responsibility for security and maintaining a strong state security apparatus allows the state to make credible commitments, including a commitment to keeping individuals safe from external threats, but also internal threats such as the ones that manifest in the form of everyday crime and violence. Whether the state can or cannot credibly demonstrate its ability to maintain a safe environment for all has implications for whether and how people participate in politics. A strong, yet restrained state security apparatus takes on special importance in the context of Africa where citizens and the state must contend daily with the legacy of colonialism, authoritarianism, and weak state capacity. Under colonialism, Africans had no meaningful opportunities for

82 

D. C. KUSHNER

participation in the political sphere. Under authoritarian or hybrid regimes, past or present, participation may easily be undercut by electoral fraud, manipulation, or, in more extreme cases, political violence. Under both, and even in democratic systems with resilient clientelistic networks, those who were not and are not well connected may not place much faith in the state to meet their security needs. Therefore, African states have more work to do than some others to convince their people that they have transformed into citizen-centered, service-oriented ones. As citizens have taken advantage of more opportunities for participation that have opened up over time, African states must continually work, as all states really should, to cultivate, nurture, and encourage participation by fostering and remaining committed to a healthy state-society relationship. This is achieved, above all, by taking responsibility for the broad-based provision of security: convincing citizens that it can protect them from other citizens that mean to do them harm, but also getting them to trust that the state itself will not prey on the citizenry through bribery, brutality, or corruption. In the context of weak capacity states, contending with the legacies I describe above, I expect the following relationship between non-state security, social capital, and collective action: Hyp. 4.3  Those who rely on non-state security will be less likely than those who rely on state security to join associations. Hyp. 4.4  Those who rely on non-state security will be as or more likely than those who rely on state security to engage in collective action. My argument about non-state security reliance and joining is that in an atmosphere of weak state security capacity, individuals who turn to non-­ state security may not have faith that the state can produce an atmosphere that is conducive to broadly realized security. Facing what they perceive to be a reality of insecurity, non-state security users will be less assured that they can go into the world and participate in a civil society that is safe and secure for all. Therefore, they will tend to steer clear of embedding themselves in associations or groups. Similarly, I expect users of non-state security to be less trusting of those around them. With the state being the only institution that can legally sanction behavior and provide broad security, at least in theory, those who rely on non-state security will likely be reluctant to trust those around them when the state has not convinced them that it can provide reasonable checks on the predatory behavior of individuals.

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

83

Finally, I expect non-state security users to be as or more likely than state security users to engage in spontaneous collective action and contentious politics as a vehicle for the expression of state discontent. These temporary, ad-hoc situations that bring people together for a short moment in time are qualitatively different from associations of civil society. They serve a key expressive function, while not requiring the bonds of association and trust that are necessary for associational membership. Lastly, I expect the following hypotheses regarding personal insecurity: Hyp. 4.5  Individuals who feel a greater level of personal insecurity will be less likely than those who do not to join associations. Hyp. 4.6  Individuals who feel a greater level of personal insecurity will be less likely than those who do not to engage in collective action. As suggested above, I expect high levels of personal insecurity to cause people to avoid interaction with others as much as possible, thereby ­suppressing participation in associations and collective action.

4.6   Data and Indicators The following analyses rely on data from Round 5 Afrobarometer surveys carried out from October–November 2011 in South Africa. Afrobarometer is the largest, cross-national public opinion survey in Africa. The survey measures citizens’ attitudes toward social, economic, and political issues in their country. All surveys are based on random samples that are nationally representative of the adult population 18  years old or older. The South Africa survey consisted of a sample of 2400 such citizens. Samples of this size yield a margin of error of +/− 2%. In describing the measures employed in this study, I first begin with the dependent variables of interest. 4.6.1  Political Participation One of the ways that citizens participate as active members of society is through joining organizations. In Africa, joining organizations has been an important way of exercising political voice, with political activity often being organized through key organizations such as voluntary associations, trade unions, and churches. Because this book is chiefly concerned with issues of security, ideally I would examine the extent to

84 

D. C. KUSHNER

which individuals join societal-based security organizations. In the absence of such a measure, I employ the closest proxy that asks individuals the following: “Now I am going to read out a list of groups that people join or attend. For each one, could you tell me whether you are an official leader, an active member, an inactive member, or not a member: Some other voluntary association or community group?” I dichotomize this variable, coding as 1 those who say that they are “an official leader” or “an active member” and coding as 0 those who report that they are “an inactive member” or “not a member”. Next we explore the extent to which citizens connect with one another by engaging in collective action. Here I combine two questions that tap the extent to which citizens’ fellowship with each other over matters that are important to them as a community. The questions are: “Here is a list of actions that people sometimes take as citizens. For each of these, please tell me whether you, personally, have done any of these things during the past year. If not, would you do this if you had the chance: (1) Attended a community meeting? and (2) Got together with others to raise an issue? Values range from 0 (“no, would never do this”) to 4 (yes, often). I created a construct labeled “collective action” by taking the mean of these two variables.

4.7   Determinants of Political Participation The main explanatory variables of interest are non-state security, victimization, and personal insecurity. I measure individual reliance on non-state security by using the following question: “If you were a victim of crime in this country, who, if anyone, would you go to first for assistance?” Response categories include “the police”, “a security service or security company that you pay for”, “a traditional leader or traditional court”, “a street committee or local security organization”, “a powerful local person or local gang”, “you would personally take revenge”, “you would join with others to take revenge”, “your own family or friends”, and “the family of the perpetrator”. I code as 0 those who report that they would go to the police and all others as 1. Therefore, my measure of non-state security is a broad one that includes both market and societal forms (see Chap. 3). I measure two types of victimization; contact and property victimization. My measure of contact victimization asks: “During the past year, have you or anyone in your family: Been physically attacked?” My measure of

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

85

property victimization asks “During the past year, have you or anyone in your family: Had something stolen from your house?” The contact victimization variable is essentially gauging whether respondents or someone in their families have been assaulted in the past year, whereas the property victimization measure assesses whether respondents or someone in their families have been victims of theft. I dichotomize these variables, coding as 0 those who respond “no” to those questions and coding as 1 those who respond “yes, once”, “yes, twice”, and “yes, three or more times”. Because I argue that social trust may mediate the relationship between security and collective action and joining, I also include a measure of social trust that asks: “How much do you trust each of the following types of people: Your neighbors?” I also create a binary variable here, coding as 0 those who say they trust their neighbors “not at all” or “just a little” and coding as 1 those who say they trust their neighbors “somewhat” or “a lot”. In addition to the above security factors, I control for two additional security considerations; personal insecurity and evaluations of government performance on the provision of security. These are important considerations as they may directly impact an individuals’ decision to participate, particularly between elections. For example, individuals may be less likely to attend community meetings and join associations that meet after work when they feel it is not safe to go outdoors after dark. Pérez (2003) has shown that greater levels of insecurity decrease support for democracy. Similarly, I would expect higher levels of insecurity to have a negative effect on political participation. The personal insecurity indicator asks: “Over the past year, how often, if ever, have you or anyone in your family: Felt unsafe walking in your neighbourhood?” Values range from 0 (never) to 4 (always). For evaluations of state security, I utilize a question that asks citizens how well they believe government is reducing crime. Values range from 0 (very badly) to 3 (very well). In addition to the security-related variables, I also control for other factors that are known to influence political participation, including ­economic performance evaluations, personal efficacy, partisanship, political sophistication, media exposure, and social structure. Scholars have long shown that the provision of economic goods (i.e. a strong economy) have a substantial influence on voting. Therefore, we should expect economic evaluations to have an impact on whether and how individuals participate. To measure performance on economic goods, I create an index from four items that asks individuals how well the government is doing with inflation,

86 

D. C. KUSHNER

providing jobs, providing food, and narrowing the gap between the rich and poor. Values range from 0 to 3, with higher numbers representing better performance on these issues. Next, I create a measure of political efficacy. Existing research has shown that individuals will be more likely to participate in politics when they feel elected officials are responsive to their demands (Finkel 1985). Therefore, I include a construct of political efficacy using the mean of the following two indicators: “How much of the time do you think the following try their best to listen to what people like you have to say: (1) Members of Parliament, and (2) Local government councilors?” Values range from 0 (never) to 3 (always). I also control for partisanship, which has been shown to be one of the most important predictors of voting behavior (Bartels 2000; Campbell et al. 1986; Bratton et al. 2012). I code those who identify with the ruling party as 1 and those who identify with an opposition party as 0. The extent to which citizens participate in political life has also been shown to be influenced by their levels of political sophistication and access to information (Kenski and Stroud 2006; McLeod et al. 1999). Two different measures of political sophistication are explored here. The first asks: “When you get together with your friends or family, would you say you discuss political matters: Frequently, occasionally, or never?” Values range from 0 (“never”) to 2 (“frequently”). The second measure asks: “How interested would you say you are in public affairs?” (0 = not at all interested; 3 = very interested). I measure media exposure by creating an additive index from the following four indicators: “How often do you get news from the following sources: (1) Radio, (2) Television, (3) Newspapers, and (4) Internet?” Values range from 0 (never) to 4 (everyday). Finally, I control for the impact of demography by including variables for age, gender, race, urban/rural residence, level of poverty, employment, and education.

4.8   Explaining Security, Joining, and Collective Action in South Africa To begin, I first examine the political impact of security variables on joining. In the model of joining that is depicted in the first column of Table 4.1, non-state security, property victimization, nor personal insecurity reach statistical significance. The only security variable that is statistically significant

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

87

Table 4.1  The impact of security variables on joining and collective action in South Africa Independent variables Non-state security Contact victimization Property victimization Personal insecurity Social trust Political order Economic goods Political efficacy Discuss politics Interest in public affairs Media exposure Age Male Lived poverty Black (reference) White Coloured Asian Urban Employed

Joining

Collective action

−0.060 (0.157) 0.515** (0.211) 0.046 (0.163) 0.063 (0.053) 0.199 (0.142) −0.131 (0.089) 0.098 (0.126) 0.255*** (0.088) 0.453*** (0.114) 0.164** (0.080) 0.162* (0.075) 0.592*** (0.183) −0.096 (0.138) 0.119 (0.097)

−0.077 (0.049) 0.009 (0.075) 0.173*** (0.053) 0.036** (0.017) 0.116*** (0.045) 0.035 (0.028) −0.171*** (0.041) 0.118*** (0.029) 0.269*** (0.036) 0.099*** (0.025) 0.103*** (0.025) 0.104*** (0.029) −0.036 (0.044) 0.178*** (0.032)

−0.201 (0.231) −0.278 (0.224) −0.118 (0.359) −0.454*** (0.156) 0.050 (0.152)

−0.808*** (0.074) −0.454*** (0.067) −0.661*** (0.114) −0.158*** (0.052) −0.058 (0.048) (continued)

88 

D. C. KUSHNER

Table 4.1 (continued) Independent variables Education Constant R square −2 Log likelihood

Joining 0.082 (0.053) −3.997*** (0.3723) 0.093 1494.328

Collective action 0.002 (0.017) 0.705*** (0.127) 0.211 –

N = 2400 Standard errors in parentheses Statistical significance denoted by ***p ≤ 0.01; **p ≤ 0.05; *p ≤ 0.10

in this model is contact victimization. The sign on the contact victimization coefficient is positive, suggesting that those who have been victims of crime are more likely than non-victims to join associations. While I expected victimization to suppress joining, victims of violent crime are more likely to join, perhaps in an effort to fight back against crime in their communities. Importantly, social trust does not reach statistical significance in this model. Aside from the security variables, political efficacy, discuss politics, interest in politics, and media exposure are positive and statistically significant indicators of joining. Therefore, it seems that people will join when they feel that doing so will make a difference, when they have access to information, and when they are genuinely interested in political affairs. Finally, older people and rural dwellers are more likely to join associations. This is consistent with a number of other studies which assess social capital in the context of Africa. Turning to the collective action model, yet again, non-state security is not statistically significant. In this model, however, property victimization and personal insecurity are both positive and statistically significant. Therefore, victims of property crimes, and those who often feel insecure are more likely to engage in collective action. These results suggest that victims of violent crime may attempt to deal with their victimization through systematic and sustained engagement in voluntary associations, while those who experience property crimes like theft are more likely to express their outrage by participating in more sporadic episodes of collective action. Moreover, while a sense of personal insecurity has no significant impact on joining, it seems that those who feel unsafe in their neighborhoods find solace in attending meetings and joining with others to raise issues.

  ASSESSING THE IMPACT OF NON-STATE SECURITY, VICTIMIZATION… 

89

4.9   Conclusion This chapter has shown that non-state security reliance does not impact citizens’ decision to join in associational life or engage in collective action. Victimization, however, does have a positive relationship with participation. Victims of violent crime are more likely to connect with other citizens in community groups, while those who have reported experiencing property victimization are more likely to attend occasional meetings or engage in other forms of collective activity. Finally, those who feel personally insecure or unsafe in their own communities are more likely to engage in collective action as well. Therefore, the empirical evidence presented here shows that while conventional determinants of participation remain important, explanations of participation that exclude citizens’ micro-level security experiences are incomplete.

References Bartels, Larry M. 2000. Partisanship and voting behavior, 1952–1996. American Journal of Political Science: 35–50. Bateson, R. 2012. Crime victimization and political participation. American Political Science Review 106 (3): 570–587. Batley, Richard, and Claire Mcloughlin. 2009. State capacity and non-state service provision in fragile and conflict-affected states. Blattman, Christopher. 2009. From violence to voting: War and political participation in Uganda. American Political Science Review 103 (2): 231–247. Brass, Jennifer. 2016. Allies or adversaries: NGOs and the state in Africa. New York: Cambridge University Press. Bratton, M., and E.C.C. Chang. 2006. State building and democratization in sub-­ Saharan Africa: Forwards, backwards, or together? Comparative Political Studies 39 (9): 1059. Bratton, Michael, Ravi Bhavnani, and Tse-Hsin Chen. 2012. Voting intentions in Africa: Ethnic, economic or partisan? Commonwealth & Comparative Politics 50 (1): 27–52. Cammett, Melanie, and Lauren M. MacLean. 2014. The politics of non-state social welfare. Ithaca, NY: Cornell University Press. Campbell, James E., Mary Munro, John R. Alford, and Bruce A. Campbell. 1986. Partisanship and voting. Research in Micropolitics 1: 99–126. Delhey, Jan, and Kenneth Newton. 2003. Who trusts?: The origins of social trust in seven societies. European Societies 5 (2): 93–137. Finkel, Steven E. 1985. Reciprocal effects of participation and political efficacy: A panel analysis. American Journal of Political Science: 891–913.

90 

D. C. KUSHNER

Hern, Erin. 2017. In the gap the state left: Policy feedback, collective behavior and political participation in Zambia. Studies in Comparative International Development 52 (4): 510–531. Jelinek, Emilie. 2006. A study of NGO relations with government and communities in Afghanistan. Afghanistan research and evaluation unit. Kenski, Kate, and Natalie Jomini Stroud. 2006. Connections between Internet use and political efficacy, knowledge, and participation. Journal of Broadcasting & Electronic Media 50 (2): 173–192. Kushner, Danielle C., and Lauren M. MacLean. 2015. Introduction to the special issue: The politics of the nonstate provision of public goods in Africa. Africa Today 62 (1): vii–xvii. Lemanski, C. 2004. A new apartheid? The spatial implications of fear of crime in Cape Town, South Africa. Environment and Urbanization 16 (2): 101. Levi, M., A.  Sacks, and T.  Tyler. 2009. Conceptualizing legitimacy, measuring legitimating beliefs. American Behavioral Scientist 53 (3): 354. Louw, A. 1997. Surviving the transition: Trends and perceptions of crime in South Africa. Social Indicators Research 41 (1): 137–168. MacLean, Lauren M. 2011. State retrenchment and the exercise of citizenship in Africa. Comparative Political Studies 44 (9): 1238–1266. McLeod, Jack M., Dietram A.  Scheufele, and Patricia Moy. 1999. Community, communication, and participation: The role of mass media and interpersonal discussion in  local political participation. Political Communication 16 (3): 315–336. Moran, Dominique, and Richard Batley. 2004. Literature review of non-state provision of basic services. Paper commissioned by DFID from Governance Resource Centre, University of Birmingham, UK. Olivier, Johan L. 1991. State repression and collective action in South Africa, 1970–1984. South African Journal of Sociology 22 (4): 109–117. Pérez, O.J. 2003. Democratic legitimacy and public insecurity: Crime and democracy in El Salvador and Guatemala. Political Science Quarterly 118 (4): 627–644. Rose, Pauline M. 2007. Supporting non-state providers in basic education service delivery. Putnam, Robert D., Robert Leonardi, and Raffaella Nanetti. 1994. Making democracy work: Civic traditions in modern Italy. Princeton, NJ: Princeton University Press. Sacks, Audrey. 2012. Can donors and non-state actors undermine citizens’ legitimating beliefs? World Bank Policy Research Working Paper (6158). Salamon, Lester M., and Stephen Van Evera. 1973. Fear, apathy, and discrimination: A test of three explanations of political participation. The American Political Science Review 67 (4): 1288–1306.

CHAPTER 5

The Impact of Everyday Crime and Security on Protest Behavior in South Africa

Abstract  This chapter assesses the relationship between security factors and protest. While there is no evidence to support a relationship between non-state security reliance and protest, victims of crime and those who often feel a great sense of personal insecurity are more likely than their counterparts to protest. Keywords  Non-state security • Victimization • Insecurity • Protest

5.1   Introduction This chapter examines the impact of non-state security, victimization, and insecurity on protest behavior.  The previous chapter demonstrated that security issues have an impact on citizens’ collective behavior. This chapter is the last to focus on non-electoral forms of participation, exploring what, if any, impact citizens’ everyday security experiences have on their decisions to engage in protest.

5.2   Protest in Africa South Africa has a reputation for being a protest nation. Historically, popular protests and strikes were instrumental in bringing the apartheid era to a close, and protests remain a visible facet of political life in South Africa today. For example, Mottiar and Bond (2012) note that in recent years © The Author(s) 2019 D. C. Kushner, The Politics of Everyday Crime in Africa, https://doi.org/10.1007/978-3-319-98095-9_5

91

92 

D. C. KUSHNER

South Africa has averaged over 8000 “gathering acts”, as they are called by the SAPS. According to research done by Vally (2009), it is clear that a majority of these gathering acts are political in nature and involve protest. There are many reasons why citizens might take to the streets in protest. According to Mueller (2010), Africans engage in protest out of frustration with declining living conditions and low chances for upward social mobility. MacLean (2011) is in line with this argument, showing that participation in Africa is motivated by citizens’ experiences of deteriorating public service provision, especially as state retrenchment spread across Africa in the late twentieth century. In the South African context in particular, the general consensus is that protests are largely driven by concerns about service delivery. Booysen (2007) argues that South Africans who support the ANC do so in a dual-­ pronged way, by both voting and protesting and, in so doing, keep their party in power while at the same time putting pressure on it to deliver the goods. The idea that citizens use protest to win socio-economic victories is echoed in Mottiar and Bond (2012), who focus on social protests in Durban. They posit that in a time when the state has failed South Africans in many ways, citizens have secured material improvements in living conditions by taking to the streets. Taking a slightly different focus, Atkinson (2007) argues that protests are caused by three factors in particular. These factors include a lack of municipal effectiveness (capacity to deliver services), a lack of responsiveness, and perceived corruption on the part of local government councilors. In fact, other scholars have found that South Africans’ assessments of local government in general are driven in large part by how responsive and capable they believe their local councilors to be (Bratton and Sibanyoni 2006). Importantly, though, Atkinson (2007) does not place the blame for poor local government performance solely on the shoulders of local councilors, but instead argues that the provincial and national levels government are, at least, partly to blame for failing to adequately fund and equip local governments to do their jobs. The problem with much of the prior work that has been discussed is that it primarily examines how citizens pressure elected officials for economic goods. With extraordinarily high rates of general unemployment, youth unemployment, poverty, and one of the highest levels of inequality in the world, it may be tempting to assume that South Africans are primarily, if not solely, mobilized to display dissatisfaction with the level and

  THE IMPACT OF EVERYDAY CRIME AND SECURITY ON PROTEST BEHAVIOR… 

93

quality of socio-economic goods provided by the state (Masipa 2018). However, South Africans have and do protest over political goods as well. For example, Alexander et al. (2010) explains how South Africans have protested over poor policing, the eradication of cross-border municipalities, and corruption. He suggests that governance issues may have featured more prominently in Zuma-era protests, pointing to how citizens in some of the largest protests have demanded information about the use of funds they believed were misappropriated. Alexander et  al. note  that “accountability has emerged as a major theme in the unrest” (2010, p. 32). Still others have pointed to protests in South Africa that have taken place over councilor unresponsiveness and nepotism (Nleya et al. 2011). Therefore, in investigating the linkage between service delivery and protest, it is important to not only assess the impact of economic goods, but political goods as well.

5.3   Data and Indicators The following analyses rely on data from Round 5 Afrobarometer1 surveys carried out from October–November 2011 in South Africa. Afrobarometer is the largest, cross-national public opinion survey in Africa. The survey measures citizens’ attitudes toward social, economic, and political issues in their country. All surveys are based on random samples that are nationally representative of the adult population 18 years old or older. The South Africa survey consisted of a sample of 2400 such citizens. Samples of this size yield a margin of error of ±2%. In describing the measures employed in this study, I first begin with the dependent variable of interest. 5.3.1  Measuring Political Protest To measure protest, I created an additive index to gauge the extent to which individuals go on strike, participate in demonstrations or protests, or engage in political violence. The questions used to create this index are as follows: “Here is a list of actions that people sometimes take as citizens. For each of these, please tell me whether you, personally, have done any of these things during the past year. If not, would you do this if you had the chance: (1) Refused to pay for services provided by government, like water,

1

 See www.afrobarometer.org for more details on these data.

94 

D. C. KUSHNER

electricity or property rates? (2) Refused to pay a tax or fee to government? (3) Attended a demonstration or protest march? (4) Gone on strike to demand a higher salary or better working conditions? and (5) Used force or violence for a political cause?” Values range from 0 to 4, with four indicating that individuals have undertaken these actions “often”. It may be easy to see how questions three through five tap into more contentious forms of political behavior, but questions one and two may deserve further explanation. In addition to protests and the use of political violence, South Africans often used tax, rent, and rate boycotts as a form of protest under apartheid (Adler and Steinberg 2000; Buur 2010). In fact, nonpayment of fees was a strategic form of political participation under apartheid, one whose intention was to cripple the economy of the apartheid state. Thus, South Africans have a long legacy of withholding payments of fees to make political statements. Therefore, I examine these forms of participation along with strikes, protests, and political violence. 5.3.2  Determinants of Protest The main explanatory variable of interest is non-state security. I measure individual reliance on non-state security by using the following question: “If you were a victim of crime in this country, who, if anyone, would you go to first for assistance?” Response categories include “the police”, “a security service or security company that you pay for”, “a traditional leader or traditional court”, “a street committee or local security organization”, “a powerful local person or local gang”, “you would personally take revenge”, “you would join with others to take revenge”, “your own family or friends”, and “the family of the perpetrator”. I code as 0 those who report that they would go to the police (71%) and all others as 1. The percentages of individuals who rely on each type of non-state security are depicted in Fig. 2.2 in Chap. 2. In addition to non-state security, I include three other security-related variables that have been shown to affect political participation: contact and property victimization and personal insecurity. I also control for other determinants of protest discussed above, including economic and political performance evaluations and corruption perceptions. Conventional factors that affect political participation, such as political interest, partisanship, and demographic variables, are also included in the forthcoming models. The full question text and coding of all variables can be found in the appendices. Unless otherwise stated, all “don’t know” or “missing” responses have been dropped from the analysis.

  THE IMPACT OF EVERYDAY CRIME AND SECURITY ON PROTEST BEHAVIOR… 

95

5.4   A Security-Based Theory of Protest and Protest Hypotheses I expect victimization to increase protest. As victims of political violence are shown to be more politically engaged than non-victims, I also expect victims of criminal violence to be more likely than non-victims to protest. Victims of crime may protest for expressive reasons, joining in solidarity with other victims to demonstrate frustration with the crime and security atmosphere. But in a society where citizens are also accustomed to using protest to secure concessions on public goods, citizens may also protest to try to improve the public safety context. Finally, I expect those who feel a greater sense of personal insecurity to engage in protest as well. These individuals may find solace in the numbers typically generated by protest, and this could potentially be a more attractive form of participation since it doesn’t require the trust necessary for organizational participation. Hyp. 5.1  Individuals who have been victims of crime will be more likely than those who have not to protest. Hyp. 5.2  Individuals who feel more personally insecure will be more likely than those who feel less insecure to protest. The relationship between non-state security and protest is more complicated. For those who primarily rely on non-state actors for security, they may have lost faith in the state’s ability to secure them. These individuals may mainly seek to protect themselves, their families, and their property through neighborhood watches, street committees, relatives, or commercial security corporations. For those who believe the state lacks any real capacity to provide protection, they may feel it is not worthwhile to lay claims on the state via protest. Hyp. 5.3  Those who rely on non-state security will be as or more likely than those who rely on state security to engage in protest. On the other hand, non-state security users may be more motivated than those who primarily rely on the police for protection to protest. Non-­ state security users, depending on their normative views of public service delivery, may still feel that it is the state’s responsibility to provide safe and secure conditions for all, and so may still lobby the state for adequate

96 

D. C. KUSHNER

public safety. Non-state security users, especially those who pay for their services and pay taxes to the state, may be especially motivated to try to secure quality public safety, so as to not be taxed twice for the same good. If non-state security users still have enough faith in the state’s capacity to improve the security environment, the costs of protesting may be worth it to them.

5.5   Results The first column (model 1) of Table  5.1 shows the multiple regression analysis of protesting with the security variables of interest to this book. In this model, non-state security does not reach statistical significance, but two other security variables do. Specifically, for every one unit increase in violent victimization, individuals are 0.247 more likely to protest, all else held constant. Moreover, those who feel more personally insecure are also more likely to protest. Model 2 in the table, includes variables to test for several competing hypotheses of protest. In particular, I test the impact of citizens’ evaluations of economic and political goods, as well as corruption perceptions. In this model, contact victimization and personal insecurity remain positive and statistically significant, but performance evaluations also matter. Specifically, when people feel that the police are doing a good job managing crime, they are more likely to protest. In the context of South Africa where crime is perceived to be extraordinarily high, citizens may feel the need to keep the pressure on, even when the police are doing well, to continue to bring crime down to what is perceived to be a more tolerable level. Alternatively, when South Africans believe the state is performing well in one area, this may motivate them to protest in an attempt to improve service delivery in other areas. Alexander et  al. (2010) suggest that protest can increase when the government is seen as legitimate; that people will take to the streets more because they believe there is someone listening to them. Perhaps this idea helps to explain the positive relationship between police performance evaluations and protest. In model 3, I introduce three additional variables; political efficacy, interest in public affairs, and ANC partisanship. The sign and significance of all the previous variables remain the same with the exception of police performance evaluation. That variable is no longer significant, but the coefficient on economic goods is significant. The result suggests that those who are more satisfied with the provision of economic goods are less likely

  THE IMPACT OF EVERYDAY CRIME AND SECURITY ON PROTEST BEHAVIOR… 

97

Table 5.1  Regression analyses of protesting, South Africa Independent variables Non-state security Contact victimization Property victimization Personal insecurity Police performance evaluations

Model 1

Model 2

Model 3

Model 4

−0.010 (0.027) 0.220*** (0.039) 0.025 (0.029) 0.054*** (0.009) –

−0.009 (0.027) 0.216*** (0.040) 0.033 (0.029) 0.058*** (0.009) 0.040* (0.015) 0.014 (0.021) −0.001 (0.008) –

−0.053 (0.037) 0.232*** (0.054) 0.009 (0.039) 0.075*** (0.012) 0.029 (0.021) −0.050* (0.031) 0.007 (0.012) 0.051*** (0.009) 0.074*** (0.016) 0.116*** (0.040) –

−0.044 (0.038) 0.223*** (0.055) 0.010 (0.039) 0.073*** (0.012) 0.028 (0.021) −0.048 (0.031) 0.006 (0.012) 0.050*** (0.009) 0.069*** (0.017) 0.015 (0.057) −0.001 (0.001) 0.042 (0.034)

Economic goods

–

Local Government Councillor corruption Political efficacy

– –

Interest in public affairs

–

–

ANC partisanship

–

–

Age

–

–

Male

–

–

–

Black (reference) White

–

–

–

Colored

–

–

–

Asian

–

–

–

Urban

–

–

–

Employed

–

–

–

Education

–

–

–

Constant R square

0.303*** (0.019) 0.037

0.240*** (0.031) 0.043

N = 2400 Standard errors in parentheses Statistical significance denoted by ***p ≤ 0.01; **p ≤ 0.05; *p ≤ 0.10

0.047 (0.055) 0.099

−0.213* (0.083) −0.134** (0.067) −0.051 (0.136) 0.036 (0.037) 0.010 (0.037) 0.012 (0.013) 0.102 (0.103) 0.109

98 

D. C. KUSHNER

to protest. This finding is in line with the economic voting literature which demonstrates that people are more likely to participate when economic times are bad. All three additional variables that were added in model 3 are positive and statistically significant. The results show that those who have an interest in public affairs, those who feel that their participation will yield a response, and those who are ANC partisans, are all more likely to protest. It comes as no surprise that those who are perhaps more politically sophisticated would be more likely to protest. It is also not shocking that ANC supporters are more likely to engage in protest than non-ruling party partisans, especially given the long legacy of protest use by those within the ANC. The final model in Table 5.1 is the full model. Here, demographic variables measuring race, age, gender, urban/rural status, employment status, and education level are included in the model. The results of the full model remain the same as model 3 with one exception: partisanship is no longer statistically significant. Essentially, by incorporating race into the model, partisanship loses its statistical significance. This is not shocking as race is correlated with partisanship in this context. For example, more blacks tend to support the ruling ANC and more whites tend to support the opposition Democratic Alliance (DA), though there is some support for the DA among black South Africans as well. The fact that the non-state security variable does not reach statistical significance in any of the above models is surprising, so I considered the possibility that this outcome may have to do with the coding of my non-­state security variable. When I created this variable, I coded as “1” all individuals who primarily rely on an entity other than the police for security. This includes commercial security users. The problem with this approach is that it conflates all non-state security users, regardless of the specific type of nonstate actor to which citizens turn. As discussed in Chap. 2 of the book, historical differences occur in the way that the commercial security industry aligned itself with the state relative to other types of non-state security actors. Perhaps even more importantly, stark differences can be found among users of different types of non-state security, with whites in particular relying much more heavily on the commercial security industry for protection, and blacks relying more on informal community networks for this good. Significantly, white South Africans are much less likely to engage in protest behavior. Therefore, the lack of statistical significance on the non-state variable in this model may be driven by the inclusion of commercial security users who are overwhelmingly white and unlikely to engage in protest.

  THE IMPACT OF EVERYDAY CRIME AND SECURITY ON PROTEST BEHAVIOR… 

99

To account for this possibility, I drop from the analysis those who report that they would turn to “a security service or security company that they pay for” and rerun the protesting model with a measure I call “societal non-state security”2 (Table  5.2). In addition, I include a relevant interaction term. I argue that non-state security users’ perceptions of the state may mediate the relationship between non-state security and protest; I include a measure of state legitimacy and explore the interaction between non-state security and state legitimacy. The state legitimacy measure is created using the following three questions: “Please tell me whether you agree or disagree: (1) The police always have the right to make people obey the law, (2) The courts have the right to make decisions that people always have to abide by, and (3) The tax department always has the right to make people pay taxes.” Original values ranged from 0 (“strongly disagree”) to 4 (“strongly agree”). Because I am interested simply in those who see that state as legitimate versus those who do not, I then create a binary variable with values