Preventing Crowd Violence 9781588269416

Citation preview

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page i

Preventing Crowd Violence

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page ii

Crime Prevention Studies Volume 26 Ronald V. Clarke, series editor

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page iii

PREVENTING C ROWD VIOLENCE edited by

Tamara D. Madensen Johannes Knutsson

b o u l d e r l o n d o n

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page iv

Published in the United States of America in 2011 by Lynne Rienner Publishers, Inc. 1800 30th Street, Boulder, Colorado 80301 www.rienner.com and in the United Kingdom by Lynne Rienner Publishers, Inc. 3 Henrietta Street, Covent Garden, London WC2E 8LU © 2011 by Lynne Rienner Publishers, Inc. All rights reserved

Library of Congress Cataloging-in-Publication Data Preventing crowd violence / [edited by] Tamara D. Madensen and Johannes Knutsson. p. cm. — (Crime prevention studies ; 26) Includes bibliographical references and index. ISBN 978-1-58826-753-5 (hbk. : alk. paper) 1. Riots—Prevention. 2. Crowds. 3. Mobs. 4. Violence—Prevention. 5. Public safety. I. Madensen, Tamara. II. Knutsson, Johannes, 1947– HV6474.P74 2011 363.32'37—dc22 2010029732 ISSN 1065 7029 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library.

Printed and bound in the United States of America The paper used in this publication meets the requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48-1992. 5

4

3

2

1

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page v

Contents

Foreword, Ingelin Killengreen Acknowledgments

vii ix

1 Introduction Tamara D. Madensen and Johannes Knutsson

1

2 From Crisis to Opportunity: New Crowd Psychology and Public Order Policing Principles 7

Stephen Reicher

3 Crowd Dynamics and Public Order Policing 25

Clifford Stott

4 Initiation and Escalation of Collective Violence: An Observational Study 47

Otto M. J. Adang

5 Police Use of Active and Passive Mitigation Strategies at Crowd Events Ingrid Hylander and Kjell Granström

69

6 Policing the British G8 Protests: A Contextualized Analysis 95

David P. Waddington

7 Crowd-Related Crime: An Environmental Criminological Perspective Tamara D. Madensen and John E. Eck

v

115

00_Madensen_FM.rev.qxd:Madensen

vi

1/3/11

10:20 AM

Page vi

CONTENTS

8 The Police and Major Event Planning: A Case Study in Las Vegas, Nevada William H. Sousa and Tamara D. Madensen

139

9 Trick or Treat? Policing Halloween in Madison, Wisconsin Joel B. Plant and Michael S. Scott

159

10 Dialogue Policing: A Means for Less Crowd Violence? Stefan Holgersson and Johannes Knutsson

Bibliography The Contributors Index Books in This Series About the Book

191

217 231 235 241 243

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:51 AM

Page vii

Foreword

THE POLICE ARE EXPECTED TO UPHOLD THE LAWS THAT GUARANtee freedom of speech and the right to peaceable assembly. When dealing with crowds, the police must safeguard participants and prevent violence. Most police forces have experienced crowd violence and must be prepared to respond to harmful crowd behaviors. Whatever the circumstances, preventing crowd violence is a complex and challenging endeavor. Police forces from different countries regularly share information on a practical and operational level. This book is an important supplement to this type of cooperation. The case studies contained here illustrate how and why police sometimes fail to effectively manage crowds in spite of excellent planning and extensive experience in dealing with crowd events. Guidelines, knowledge, and understanding must be used methodically, both in planning and when responding on scene, and that is the focus here. As a police officer, I feel that the concepts described in this volume will help to improve fundamental police functions related to preventing crowd violence: good planning, understanding freedom of speech and freedom of assembly, communication with an understanding of the crowd and their legitimate aims, nonaggressive police behavior, and problem-oriented policing. I do hope that both police officers and students of policing will read and learn from the interesting case studies and guidelines proposed by the authors. I extend a warm thank you to Johannes Knutsson and Tamara Madensen, who have taken the initiative to produce this volume, and to the excellent colleagues who have contributed in writing a remarkable, important book. —Ingelin Killengreen National Police Commissioner, Norway

vii

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page viii

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page ix

Acknowledgments

WE WOULD LIKE TO EXPRESS OUR GRATITUDE TO THE STAFF OF the Norwegian Justice Sector’s Center for Training and Practice. We are also indebted to the Norwegian Police University College for funding; without its support, this project would not have been possible. Finally, we would like to thank the contributors to this volume. In particular, we appreciate their thoughtful comments on earlier drafts. We are pleased to present the results of this exchange in the chapters that follow. —Tamara D. Madensen and Johannes Knutsson

ix

00_Madensen_FM.rev.qxd:Madensen

1/3/11

10:20 AM

Page x

01_Madensen_Ch01.qxd:Madensen

1/3/11

10:21 AM

Page 1

1 Introduction Tamara D. Madensen and Johannes Knutsson

THIS VOLUME IS DEDICATED TO THE EXAMINATION OF CROWD violence and police tactics used to prevent this behavior. The theories, research, and case studies presented within this collection of chapters provide an overarching perspective from which to approach the task of developing effective crowd management strategies. We argue that effective practice, when guided by the general principles of crowd facilitation and systematic analysis, will increase positive interactions and reduce tensions between crowd members and police. This outcome requires an accurate understanding of how crowds behave. It also demands recognition of the social dynamics and environmental features that influence individual motivation in large gatherings. Crowd management is a fundamental police function in all societies. This task is complex and demanding. Police must address safety concerns in highly diverse environments and among heterogeneous collections of crowd participants, while adapting to dynamic and sometimes volatile situations with existing resources and limited information. Police must also accomplish seemingly conflicting goals when managing crowds. • Political demonstrations often bring together groups of people who strongly oppose government policies or the opinions of other political groups (often protesting in the same place, at the same time). Police in democracies must guarantee the right to free speech and assembly, while safeguarding public order during these demonstrations. • Sporting events continually attract large groups of fans. Police work to maintain a safe environment to facilitate game enjoyment, but interactions between impassioned rival fans sometimes lead to violence. • Police are required to create safe environments for large groups of visitors who attend festivities associated with alcohol consumption—for example, Midsummer’s Eve in Sweden, New Year’s Eve in the United States, Oktoberfest in Germany. However, some attendees whom police are asked to protect may engage in behaviors that threaten the safety of others. 1

01_Madensen_Ch01.qxd:Madensen

2

1/3/11

10:21 AM

Page 2

PREVENTING CROWD VIOLENCE

Peaceful protests, events, and gatherings are common, but occasional outbursts of violence and vandalism necessitate direct police intervention. Sometimes, police do not, or cannot, maintain control of crowds. Sometimes, police overreact to such outbursts and use excessive or indiscriminate force against crowd participants. Technological advances have increased the likelihood that questionable police conduct will be recorded and publicized; thus, police now face increasing pressure to appropriately manage crowds. Negative publicity arising from improperly managed events creates an impetus for examination and reform of police crowd management principles and tactics. Police conduct during the 2009 G20 meeting in London generated dissatisfaction among a considerable proportion of the general public and resulted in investigations of police performance (Her Majesty’s Inspectorate of Constabulary 2009). Likewise, as described in the last chapter of this volume, police failure to control the Gothenburg riots during the 2001 European Union summit became a national crisis in Sweden. Yet, this event served as a starting point for Swedish police to develop and implement a new national crowd management tactic. As the case studies presented in subsequent chapters will demonstrate, police failure to properly manage crowds has produced disastrous consequences, but it has also created opportunities for reform and change. Those tasked with policing crowds should develop management strategies guided by practical theory and empirical evidence. This volume provides a foundation for this process by combining descriptions of theoretical frameworks with detailed case studies of crowd events, analyses of police tactics, and strategic principles for crowd management. The following nine chapters represent an inclusive collection of crowd-based theories and research conducted over several decades, in several countries, and within two general theoretical frameworks. Chapters 2 through 6 highlight advances in social psychological perspectives on crowd behavior, and the last four introduce an alternative, yet complementary paradigm for understanding and managing crowd behavior: environmental criminology. The integration of these two academic paradigms offers a comprehensive roadmap for developing effective crowd policing strategies. In Chapter 2, Stephen Reicher reviews the G20 demonstrations in London and offers four new public-order policing principles grounded in recent approaches to crowd psychology. In Chapter 3, Clifford Stott discusses the implications of the Elaborated Social Identity Model of crowd behavior and demonstrates the need for particular strategic approaches when policing football crowds. Otto Adang identifies mechanisms that trigger and escalate crowd violence using systematic observations of sixty football matches and seventyseven protest events in Chapter 4. The relationship among the organizing processes of crowds, police reaction toward protesters, and crowd violence is examined by Ingrid Hylander and Kjell Granström in Chapter 5. In Chapter 6, David Waddington uses a major transnational protest and the Flashpoints

01_Madensen_Ch01.qxd:Madensen

1/3/11

10:21 AM

Page 3

INTRODUCTION

3

Model of Public Disorder to illustrate the benefit of a contextualized approach when examining the nature and impact of policing tactics. In Chapter 7, Tamara Madensen and John Eck systematically apply environmental criminology principles and theories to crowd-related crime and offer five crowd management principles that stem from this perspective. William Sousa and Tamara Madensen use an environmental criminology theory, situational crime prevention, to explain the success of a New Year’s Eve crowd management strategy in Chapter 8. In Chapter 9, Joel Plant and Michael Scott describe a seven-year problem-oriented policing effort that successfully reduced harm associated with an annual street party. In the final chapter, Stefan Holgersson and Johannes Knutsson discuss the function and effectivness of dialogue police officers in the newly implemented Swedish tactic for crowd management that, among other things, uses situational techniques to prevent crowd violence. As the content of this volume will demonstrate, effective crowd management requires much planning, early and continuous analysis of risk, and cooperative interactions between police and crowd members. These activities help police develop and implement interventions that reduce the potential for violence. Force, however, should not be used as a routine strategy to gain compliance or resolve conflict. Forceful police tactics, as pointed out in many of the chapters that follow, may be counterproductive and arouse aggressiveness in crowds. Research demonstrates that police can improve crowd relations and encourage peaceful events by reducing forceful appearance and conduct (Reicher et al. 2007). New crowd psychology recognizes that crowds do not make individuals go “mad,” and people do not lose control of their actions simply because they participate in a large gathering (McPhail 1991). Yet, as argued by Reicher and Stott in Chapters 2 and 3, seemingly indiscriminate police use of force may change the disposition of otherwise peaceful crowd members. Through interand intragroup processes involving the police and crowd participants, individuals’ social identity may temporarily change. If crowd participants feel police actions against the crowd are overly severe or unprovoked, individuals will be more likely to contribute to violent acts to resist the police. These processes are described and explained by the Elaborated Social Identity Model (ESIM) of crowd psychology (e.g., see Reicher 2001a). Nevertheless, some police manuals continue to promote inaccurate views of crowd behavior and support police practices that treat crowds as a homogenous and dangerous collective. Empirical studies refute the notion that crowds are collections of violent and uncontrolled individuals. Using systematic social observation, Adang shows in his chapter that violent incidents are relatively rare among gatherings of both football fans and protesters. Furthermore, when violence occurs, there are few that actually engage in acts of overt aggression. There are, as pointed out by Reicher and by Sousa and Madensen,

01_Madensen_Ch01.qxd:Madensen

4

1/3/11

10:21 AM

Page 4

PREVENTING CROWD VIOLENCE

many different groups that will participate in any particular crowd event, only some of which may have unlawful intentions. Aggressive police behavior can encourage violent intentions among groups that initially do not have them. Still, police must occasionally intervene to maintain or regain control in crowds. When police must use force to achieve this objective, they should act in ways that assure force is kept to a minimum, is directed precisely, and is perceived by crowd members as being necessary and appropriate. In Chapter 5 by Hylander and Granström and Chapter 10 by Holgersson and Knutsson, various police tactics that promote peaceful demonstrations and serve to de-escalate hostile situations are described. Facilitation and communication are emphasized as critical elements of crowd management strategies. Specially trained dialogue police officers in Sweden, as described by Holgersson and Knutsson, play a vital role in improving police-crowd relations. These officers act as liaisons to increase communication and cooperative interaction between police commanders and demonstrators (see Holgersson 2010). The importance of de-escalation tactics is further demonstrated in Chapter 6 where a multilevel framework for understanding police-protester interaction is applied by Waddington to two events in the UK. This exercise supports the ESIM crowd theory by reinforcing the importance of promoting positive interactions, for example, respectful dialogue between protesters and police. The study also recognizes the influence of situational variables on crowd behavior, for example, how physical and social environments are perceived by crowd members and managed by police. Like the ESIM perspective, and models proposed by Waddington and by Hylander and Granström, environmental criminology recognizes the relevance of situational characteristics and dynamics for understanding and controlling violence. Chapters 7 through 10 in this volume demonstrate that environmental criminology is both applicable to the problem of crowd-related crime and compatible with European research findings and social psychological theories of crowd behavior. The integration of these perspectives will allow strategic planners to better understand the dynamic processes that occur within crowds and how these processes can be managed. For example, two US policing strategies (Sousa and Madensen describing Las Vegas in Chapter 8, and Plant and Scott describing Madison in Chapter 9) and Swedish police strategy (Holgersson and Knutsson in Chapter 10) are designed to manipulate situational characteristics that provide opportunities for crime. These strategies include techniques proposed by situational crime prevention (Clarke 1997a, 2008), an approach advocated by environmental criminologists to reduce crime. They are also used extensively within problem-oriented policing. In particular, the technique of reducing provocations can, as pointed out by Holgersson and Knutsson, be seen as a direct application of the ESIM principles to prevent escalation and bring about de-escalation of crowd violence.

01_Madensen_Ch01.qxd:Madensen

1/3/11

10:21 AM

Page 5

INTRODUCTION

5

The situational crime prevention approach adds to the ESIM crowd management perspective by encouraging planners to consider features of the physical environment, as well as social dynamics, when seeking to reduce opportunities for violence. As described in the subsequent chapters, Las Vegas police distribute plastic cups in exchange for glass bottles to reduce the amount of glass and potential for harm on the street. Madison police create additional egress points from crowded areas by promoting the use of rear exits from bars. Swedish police remove objects that may be used as weapons from places where demonstrators travel or congregate. Manipulation of risk factors, particularly when guided by the situational crime prevention framework, reduces the opportunity for misconduct (by police or attendees) and the likelihood that police will need to use force to control crowd participants. These perspectives advocate analyses of event risk factors both during and prior to events. The importance of “dynamic risk assessment” during events by police to determine appropriate and proportional tactical response is discussed in Stott’s Chapter 3. This is one of the primary functions of dialogue police officers as described by Holgersson and Knutsson. Environmental criminology also suggests the importance of risk assessment and planned response, but at a larger time scale. Problem-oriented policing offers a framework to guide police planning efforts and pre-event analysis of risk. Plant and Scott describe how Madison police use a problem-oriented policing approach to successfully manage a previously violent and disorderly annual event. Police begin the problem-oriented policing process by identifying a specific recurring problem and carefully describing its associated harms. This is followed by a detailed analysis of the problem’s immediate causes. The next step is to develop and implement tailor-made responses to prevent the problem’s reoccurrence. Lastly, police conduct an evaluation to determine the effectiveness of their responses (Goldstein 1979, 2003; Scott et al. 2008). In Madison, this process was repeated over several years, and today, this event is a much safer and more orderly festivity. Sousa and Madensen also describe a twelve-month problem-oriented policing process used annually by Las Vegas police. As noted by Madensen and Eck in Chapter 7, problem-oriented policing is particularly useful for managing crowd events since it promotes contextspecific solutions that should decrease violence and increase cooperation and trust between police and crowd members. When choosing tactics to manage crowds, police around the world must consider their obligation to uphold the rights and freedoms guaranteed by their respective conventions and constitutions aimed at protecting human rights. Freedom of assembly is espoused as a fundamental human right in democratic societies, and police are responsible for facilitating the safe expression of this right. Furthermore, police are obliged to protect the right to life even in chaotic situations such as riots. A recent judgment by the European Court of Human Rights makes it clear that this responsibility extends beyond individual police

01_Madensen_Ch01.qxd:Madensen

6

1/3/11

10:21 AM

Page 6

PREVENTING CROWD VIOLENCE

officers; the police organization, as judged by its objectives, policies, and directives, must demonstrate a commitment to protecting human life. If not, the state may be held responsible. Additionally, police behaviors inconsistent with this purpose must be investigated in a thorough and transparent manner (European Court of Human Rights 2009). We leave the readers of this volume with four general recommendations for crowd management supported by the research that follows. First, understand the nature and goals of a crowd before attempting to manage it. Although crowds are heterogeneous collections of individuals and should be treated as such, successfully managing any gathering will be impossible without first knowing what participants wish to achieve. Second, use theory and prior evaluations of crowd interventions to guide the development of crowd management strategies. The frameworks and case studies outlined in the following chapters will help direct attention to important social dynamics and environmental characteristics that influence crowd behaviors. Third, recognize that it is neither theoretically nor practically useful to discuss crowd violence without taking into account police behavior. In a very real sense, the crowd and the police coproduce peace or violence. Finally, understand that the overarching role of police in dealing with crowds is to protect the rights and safety of participants. Sincere attempts to work with crowds will reduce police burden by eliminating a preventable source of crowd anxiety and tension: fear of unwarranted police use of force.

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 7

2 From Crisis to Opportunity: New Crowd Psychology and Public Order Policing Principles Stephen Reicher

RECENT EVENTS HAVE SERVED TO PROVIDE A POWERFUL AND timely illustration of the longstanding arguments that my colleagues and I have made about crowd psychology and crowd policing. Indeed, we couldn’t have provided a better (or, perhaps, worse) illustration if we had scripted things ourselves. The events in question concern the London G20 demonstrations of late March 2009. Since G20, public order policing has been back in the news. At first, coverage fitted a pattern that is familiar in Great Britain’s postimperial fantasy island, one summed up in the words of an old Flanders and Swann song: “The English, the English, the English are best, I wouldn’t give tuppence for all of the rest.” Italians, Germans, and the French would have used rubber bullets and tear gas, but the brave bobbies expressed typical British restraint. All was best in the best of all possible countries. Soon things changed. First there was controversy around the tactic of “kettling” (sometimes also called “coralling” or “bubbling,” it consists of surrounding the crowd and keeping them in one place without allowing anyone to leave). Was it legitimate? Was it proportionate? Did it contain trouble or create trouble? Next, news spread of a middle-aged man, Ian Tomlinson, who died during the demonstrations. The turning point came when the Guardian newspaper posted a video of Tomlinson being assaulted by a police officer shortly before he died.1 Soon more video emerged of the assault from a different angle. It showed how he had been truncheoned before being shoved to the ground.2 Within days, yet more video spread across the Internet showing more assaults. Perhaps the most graphic depicted a tall, powerful officer, his identifying numbers covered in tape, first delivering an insouciant slap across the face to a diminutive woman protester and then, with great deliberation, swiping her legs

7

02_Madensen_Ch02.qxd:Madensen

8

1/3/11

10:41 AM

Page 8

PREVENTING CROWD VIOLENCE

with his truncheon.3 A trickle became a flood, and commentators changed their tune from admiration to outrage. On April 16, the commissioner of the Metropolitan Police, Sir Paul Stephenson, announced an enquiry into policing on the day to be led by Chief Inspector of Constabulary Denis O’Connor. On April 21, a full parliamentary enquiry into the policing operation was announced. For the first time in a generation there is an opportunity to open up a debate about the very nature of public order policing, to rethink how the police view, approach, and treat crowd members. This crisis can be turned into an opportunity—and indeed the events have generated a serious and comprehensive rethinking of how crowds should be policed. But, as ever, one cannot come up with good answers without asking the right questions, and this, in turn, depends upon a candid assessment of the problems—neither understating nor overstating them. In this respect, the videos that opened up this debate may prove unhelpful in resolving them. Look at them. The individual culprit dominates the visual field. One cannot help but focus on the truncheon hand or, looking at the assault on a female protester, be gripped by the look of arrogance, of casual derision, on the face of her assailant. But look again. Don’t focus on the one officer. Look at the behavior of the others. They see what is happening but none expresses shock, surprise, or disapproval. After the assault on Tomlinson, the officer involved strolls over for a casual chat with his superior. There is no sign of reprimand. Everything seems to be taken as acceptable and normal by officers in general. I am reminded of the film The Accused, in which Jodie Foster’s character is raped. There is a trial, but the focus isn’t on the rapist, it is on the onlookers who at best did nothing, at worst egged on the rapist, and together created a moral climate in which rape became thinkable and do-able. There is another parallel with the film. Everyone present knew what had happened. All the officers saw several offenses being committed. It is their moral, professional, and legal responsibility to report any crime that they witness, but the cases came to light despite and not because of these officers. Officers (and then only the direct assailants) were only suspended after members of the public produced incontrovertible evidence. Before that, police officers either ignored or actively misrepresented what had happened on the day. What all this tells us is that we face a systemic problem, a problem of group norms and established practices, rather than just a problem of “a few bad apples” (which is not to say that some apples are not bad, or at least worse than others). That is, in a contentious crowd, everyone is fair game and deserves what they get. In this regard, consider the following comment by Peter Smyth, chairman of the Metropolitan Police Federation (which represents junior officers, those on the front line during crowd events). On the British Broadcasting Corporation’s (BBC’s) flagship morning radio news outlet, the Today program, he said some physical confrontation was inevitable during a large protest such as G20:

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 9

FROM CRISIS TO OPPORTUNITY

9

On a day like that, where there are some protesters who are quite clearly hellbent on causing as much trouble as they can, there is inevitably going to be some physical confrontation. Sometimes it isn’t clear, as a police officer, who is a protester and who is not. I know it’s a generalisation but anybody in that part of the town at that time, the assumption would be that they are part of the protest. I accept that’s perhaps not a clever assumption but it’s a natural one.4

Smyth’s observations about the nature of police perceptions certainly chime with our own evidence (Drury, Stott, and Farsides 2003; Stott and Reicher 1998a). That is, even if junior officers accept that crowds are mixed and that the majority of participants are peaceful decent citizens, they equally consider that when trouble starts, everyone left is a troublemaker. This is either because (as Smyth himself suggests) the decent majority will desert the trouble spots or else because decent people will get swept along if they do happen to remain. However, when it comes to the origins of these perceptions, Smyth both underestimates and misattributes the problem. The notion that, in a large crowd, everyone is potentially dangerous and should be treated as such is not just an individual misconception (whether natural or not), it is something that is trained into officers and given institutional support. Worse still, it is a notion that is legitimized by the stamp of science—psychological science in this instance. That is, public order training draws on certain classical ideas about crowd psychology that suggest that even the most respectable citizens lose their sense of self, of values, or restraint when submerged in a crowd and therefore become liable to get carried away and do almost anything. Everyone is potentially dangerous. Everyone needs to be treated as dangerous. No one can complain about such treatment. Thus, for instance, the material given to participants on a national public order commanders training course in 2006 cited approvingly from E. D. Martin’s 1920 text on the behavior of crowds that “a crowd is a device for indulging ourselves in a kind of temporary insanity by all going crazy together.” What is more, it suggests that such a view is the expression of an ongoing academic consensus: All psychologists seem to agree, that membership of a crowd results in a lessening of an individual’s ability to think rationally, whilst at the same time his/her more primitive impulses are elicited in a harmonious fashion with the emerging primitive impulses of all the other crowd members. The result being the establishment of a collective mind. As you can see, all these theorists have made points concerning crowd behaviour which contain a lot of truths.

So, there is strong intellectual and ideological material to support the norm that, in situations of conflict, all crowd members need to be treated as troublemakers. In addition, there is a practical dimension to such generalization. Going

02_Madensen_Ch02.qxd:Madensen

10

1/3/11

10:41 AM

Page 10

PREVENTING CROWD VIOLENCE

back to our own research, officers related how certain tactics make it necessary to treat everyone the same. You can’t sweep a street to disperse a crowd but then pause at each protester to consider if they should be forced away or not. There isn’t the time, whether there is the inclination or not. As already indicated, at G20 the tactic of choice was to contain rather than disperse the crowd, but the kettling tactic like the dispersal tactic still treats everyone the same, whether they are balaclava-wearing radicals who believe in the legitimacy of violence or a pregnant woman who is concerned about third world child mortality. Even more than the collective police response to individual acts of violence by officers, the tactic of kettling reveals a systemic issue of how the crowd is viewed and treated. It either reveals unconcern at treating all crowd members as a problem or else reveals a belief that all crowd members are a problem. More likely, homogenizing tactics intertwine with homogenizing theory to provide yet stronger support for a norm of criminalization and an associated ethic of unconcern with harm done to crowd members. So what is the problem? Treating crowds and crowd members as a problem is the problem. Ideologies and practices that warrant such treatment underpin the problem. The psychology of crowds, insofar as it is a major element of such ideologies, is therefore part of the problem. One of my core claims here, then, is that we need to review and revise this psychology in order to grasp the opportunity we have to rethink public order policy and procedures. We need a new crowd psychology for new crowd policing. There is one more aspect of the problem—or perhaps the repercussions of the problem—that needs noting. That is, when the police are seen indiscriminately to attack those involved in a protest, then the repercussions go way beyond those who have been attacked and well beyond those involved in the protest. They can have a serious effect upon the police relationship with whole communities and with the public in general. The following is just one example of the effect of “kettling” upon demonstrators: Despite our repeated requests to be searched and allowed to leave the space, we were held there for 6 hours with no access to water, food, toilets or medical care. Proudly, throughout all this, not one person in the crowd reacted with violence to any person or property. People shared the little they had and held public meetings about the aims of the G20 summit. There was little show of anger, but much unhappiness. When finally we were herded out one by one at midnight, I felt cold to the core, chilled by the unprovoked aggression of those who I had been brought up to trust. I am deeply ashamed of my state, when reasonable and calm protesters are criminalised and provoked in such a manner. I am writing this today because I grew up in this city and treasure the right to use this city space to speak out to our elected leaders in a peaceful, creative way . . . but the harm done by the police last night goes far deeper that [sic] the physical wounds inflicted; it is in the chaos of unnecessary state violence that fear is born and trust is lost.5

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 11

FROM CRISIS TO OPPORTUNITY

11

Now consider how the events impact upon the broader community. My colleagues and I have argued elsewhere (Reicher et al. 2007) that crowds generally see themselves and are seen by others as representing general social categories: that is, they aren’t just crowds but crowds of black people, of students, of environmentalists, or those who are antiwar, or whatever. In part at least this is because these general categories are, in Benedict Anderson’s term, “imagined communities” (Anderson 1983). That is, we have a clear idea of the group and of those who belong to it, but its size and dispersal is such that one never sees an actual physical gathering in which all members of the group come together in one place at one time. The nearest one does get, we suggest, is in crowd events. That is, crowds are imagined communities made manifest. Hence what happens to the crowd is of interest to all category members, and how the crowd is treated by others (notably the police) tells all category members about their relationship with the police. This is obviously facilitated by the fact that crowd events are newsworthy events, meaning people are likely to know what happened and hence be affected by it. To take one example in recent British history, crowd events have been emblematic in defining the relationship between the police and black people. Whether in Notting Hill in the 1970s, or Brixton, Bristol, Handsworth, Tottenham, and Toxteth in the 1980s, the policing of crowds and negative relations between police and black people were clearly interlinked (see Bunyan 1981/ 1982; Institute of Race Relations 1987; Sivanandan 1981/1982, 1983). In the case of G20, however, the telling thing is that many of those who were the targets of police action were not seen as members of some specific category in society (such as environmentalists or anticapitalists or whatever) but simply as ordinary members of the public. Most strikingly, Ian Tomlinson was described in the Guardian (which broke the story of his death) as a “47-year-old newspaper seller, who had been on his way home from work.”6 He could have been anyone; he was everyman and hence his fate said something about the way the police relate to all of us. That, perhaps, explains why, according to one poll, nearly 60 percent of a national sample believed that the police used excessive violence during G20,7 while another showed that as many as 40 percent of Londoners had a worse opinion of the police after G20.8 While, in one sense, this shows that the problems that derive from public order policing are far wider than is often appreciated, that should be balanced by the realization that the opportunities that arise from public order policing may equally be far greater than is generally appreciated. Thus, if crowd-police relations impact on general police-community relations, and if conflictual crowd events sustain everyday conflicts and tensions between police and individual community members, then so the corollary should equally hold. That is, positive crowd events may be a way of turning around seemingly intractable conflicts with particular communities and improving day-to-day relations.

02_Madensen_Ch02.qxd:Madensen

12

1/3/11

10:41 AM

Page 12

PREVENTING CROWD VIOLENCE

As an example, I have just referred to the way that conflict during the Notting Hill Carnival contributed to chronic hostility between the police and black people. In the mid-1980s, however, when we studied the carnival,9 the police implemented a whole new approach to avoid conflict with carnival crowds. Police numbers were reduced, they patrolled in smaller numbers and interacted more informally with crowd members, and they specifically analyzed sources of conflict and used this to restructure their procedures accordingly. For instance, violence had often arisen when officers tried to close down sound systems (loud mobile discotheques) at night. Revellers felt they had a right to party on and resented the police stopping their enjoyment. As these officers tried to push through and stop the music, scuffles and fights often broke out. So now, the police simply told the operators that if they carried on past the agreed time, their licenses would be removed or they would get worse pitches the following year. It would take only one officer to turn up discreetly for operators to police themselves and close down. There would be a whistle and a “boo” or two, but there was no target for violence and no focus for antipolice hostility. The changes were clearly noticed and noted by the black community. I recall one incident in particular. We had interviewed one of the organizers before the carnival who expressed considerable distrust and suspicion of the police. On the day of the carnival, we ran into him. He voiced both amazement and delight at the policing and at his positive interactions with officers during the day. It is remarkable, he said, someone must have been teaching them. This takes us to a second core claim. That is, by approaching crowds as if they invariably constitute a problem, the police will necessarily miss out on the opportunities afforded by public order policing. If one presupposes that crowds will be trouble, if one anticipates events with dread and treats participants as opponents, if one considers that the best that can happen is the absence of violence, then one will never be able to use public order policing as a uniquely effective tool in solving problems for the police and the community at large. So, the aims of this chapter, and of our approach in general (Reicher et al. 2004; Reicher et al. 2007), are to challenge the traditional “problem-based” approach to crowds and to crowd policing, to outline an alternative understanding of crowds and the origins of crowd conflict, to use this understanding in order to develop principles and practices for effective public order policing, and to point to the opportunities that would be opened up by such an approach. In line with these aims, I shall first critique the old and present the new psychology of crowds. I will then lay out four principles for public order policing. I will conclude by returning to the G20 and giving one illustration of what an approach based on these principles might have looked like and could have delivered. I will, however, leave a more comprehensive account of actual prac-

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 13

FROM CRISIS TO OPPORTUNITY

13

tices and actual payoffs deriving from our approach to Clifford Stott. He has done the work and he will describe it in his chapter in this volume.

The Old Crowd Psychology and the New Crowd Psychology We have already encountered the old crowd psychology in the pages of contemporary police training manuals. Martin, who is cited directly, reproduces the ideas of Gustave Le Bon, whose book The Crowd (1895) is possibly the most influential psychology text of all time. For him, anyone who enters the crowd undergoes a fundamental psychological transformation. To be “submerged” in the mass, to become an anonymous element in the whole, leads to a loss of self. Since the self is the source of one’s judgment and values, this means that people are no longer able to judge what is right or wrong, what they should do and what they shouldn’t do. As a result, people are no longer able to filter the influences to which they are subject. On the one hand, they will respond indiscriminately to any passing idea or emotion (the idea of “contagion”), on the other, there will be no limits to how far they will go. That is why decent people can act savagely, why law-abiding citizens can act as criminals, why all those in the crowd are equally suspect. Overall, then, Le Bon insists that crowd members have descended several rungs on the ladder of civilization and that they are only powerful for destruction. In ordinary language, he proposes the idea of the “mad mob.” To borrow Lady Caroline Lamb’s description of Lord Byron, Le Bon’s crowd is mad, bad, and dangerous to know. There are many problems with such a view (see Reicher 1982, 1987, 2001a, 2001b for fuller critiques). Empirically, it simply isn’t true that everyone is subject to contagion. As S. Milgram and H. Toch (1969) pertinently observe, the police officer in the crowd hardly gets carried away by the words of a demagogue. Equally, it isn’t true that crowds are generally violent. If anything, violence is rather rare. It may be that when violence happens crowds are reported, so we only hear of crowds in conjunction with reports of violence (“Big crowd, nothing happened” is hardly likely to make the front pages), but that is a reflection of media values rather than of crowd realities. And even on the rare occasions where there is conflict, people don’t just explode and do anything. Rather there is a very clear pattern to events, to who or what is chosen as a target of violence, to the exact forms of violence meted out to them. As historians are increasingly realizing, far from being random eruptions, crowds provide a graphic illustration of the cultural understandings of crowd members. They provide a resource for those who want to understand the perspective of those involved (see, for instance, Thompson 1971).

02_Madensen_Ch02.qxd:Madensen

14

1/3/11

10:41 AM

Page 14

PREVENTING CROWD VIOLENCE

Le Bon’s approach also has a number of conceptual problems, starting with the fact that he is never quite clear what constitutes a crowd. More specifically, he fails to distinguish between a physical crowd—a set of people who simply happen to be in the same place at the same time, like shoppers in a street or travelers in a train carriage—and a psychological crowd—a set of people gathered together who have a sense of sharing something in common. An understanding of the nature and implications of this difference is the starting point for the new psychology of crowds. In order to explain it, I shall draw on an example that I have used before and which is instantly comprehensible to British people who have experience of our creaking railway system. If you come from a country where the trains always run on time, I apologize and suggest that you substitute an alternative example of your own. To start with, consider the gathering of people in a train carriage. They may well be crowded, but far from having a sense of connection with each other, they actively strive to maintain a disconnection. They seek to avoid eye contact; they stand turned away from each other; they may well all read the same paper but would never glimpse at that of their neighbor; they experience discomfort or even disgust if anyone brushes against them or even presses into them. Then the train stops. It stays stopped. There is no announcement. Perhaps, after a while, a voice explains that the train is delayed because of leaves on the line, the wrong kind of snow, or problems up the line. Now things change. Psychologically, people stop seeing themselves as separate individuals who happen to be in the same space. They now have something in common. They have a common sense of category membership. They are all commuters in contrast to the train company. Along with this psychological shift, there is a behavioral shift. People start turning to each other, talking to each other, perhaps even sharing sandwiches with each other. The emergence of psychological commonality goes along with the emergence of social solidarity. So, becoming part of a crowd does indeed involve a key transformation. But that transformation is not a matter of losing the self, it is rather a matter of shifting from an individual self to a social selfhood—myself as “a commuter” as opposed to myself as John Smith. This underpins three further transformations that affect how people act in crowds. First, just as crowd members shift the focus of identity rather than lose identity, so they shift the basis of their values and judgments rather than lose any sense of value of judgment. In other words, they no longer act on their personal values but rather on the values associated with the relevant social category. That is why one always finds constraint and pattern in crowd action, and also why it reflects collective notions of right and wrong and collective beliefs concerning appropriate action. The actions of different crowds will be highly different as a function of the social categories they represent. What a crowd of Catholics, of antifascists, of Manchester United fans will do will all be very different.

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 15

FROM CRISIS TO OPPORTUNITY

15

Second, as the train carriage example suggests, there will be a transformation in the relations between individuals in the crowd. We are more likely to agree with, trust, respect, support, and help people who are fellow in-group members (see Reicher and Haslam 2010, for a review). We want literally as well as metaphorically to be closer to them (Novelli, Drury, and Reicher 2010), and we experience their proximity with pleasure rather than disgust. We are also more likely to agree on a leader when we share psychological group membership (Reicher, Haslam, and Hopkins 2005). To use the term used previously, we both experience and provide social solidarity in the crowd. In turn, such coordination and mutual support allow members to align their efforts and achieve not only a sense of empowerment (Drury and Reicher 1999) but actual social power. This gives crowd members not only a new perspective on the world but also the ability to put it into practice. Third, there is an affective transformation. Crowds are passionate affairs. Not, as the classic theorists suggested, because the loss of judgment and intellect puts emotion in the driving seat. Quite the opposite in fact. It is because groups give people a means of orienting in the world, a sense of connection to others and—combining these two—the agency to make history (as opposed to living in a world made by others, which is our predominant experience in everyday life) that people care so passionately about the crowd (see Prayag Magh Mela Research Group 2007). It is also why they respond so strongly if they consider the group to be under attack, or its rights to be transgressed. With these general processes in mind, we can now address the question of conflict and violence in crowd events. According to the old psychology, this is simply part of the nature of crowds. The only way one can stop it is either by stopping a crowd from forming or else by repressing the crowd so completely that violence cannot be expressed. However, according to the new psychology, such a view not only is wrong, but may actively contribute to the creation and escalation of collective conflict. There may be some cases where a particular crowd has values that not only sanction but encourage violence and where crowd members have explicitly violent intentions—just as there are some cases of pacifist crowds that explicitly reject violence under any circumstances. However, in general, violence arises where crowd members believe either that their legitimate rights and entitlements have been violated or else that others have acted illegitimately toward them (Stott and Reicher 1998b). This has three important implications. First, violence is not something inherent in the crowd and explicable solely by considering the crowd. Rather it is the result of an interaction between the crowd and others such that it can only be understood by taking the actions of both parties into account. Second, what precipitates violence will vary as a function of the different conceptions of rights and of legitimacy/ illegitimacy held by different groups. Third, if the police consider all crowds and all crowd actions to be inherently suspect, and if as a result they clamp

02_Madensen_Ch02.qxd:Madensen

16

1/3/11

10:41 AM

Page 16

PREVENTING CROWD VIOLENCE

down on all crowd actions, then they are likely to violate the crowds’ notions of rights and hence to produce violence. But this is still too simple a picture. To flesh it out we must again return to and refine our account of physical crowds and psychological crowds. In distinguishing between the two, we also need to recognize that, in a physical crowd there may be no psychological crowds (as in the train example), but there may also be multiple psychological crowds. So, for instance, in a trainload of people traveling to a football match, one might have fans of the one club, fans of the other club, and football enthusiasts who are fans of neither. One mass of people, often all intermingled, but three separate categories of people. It is this type of situation that is commonplace in the large events that officers are called upon to police. At antiwar demonstrations, environmental protests, G20, there will be a multitude of multitudes, all with different values, beliefs, purposes, senses of rights. Some, usually a minority, will be more conflictual than others. How is it possible to control this minority without inflaming the majority? How can one stop some people from doing things that are unlawful without impeding the rights of others? That is the fundamental challenge of public order policing—and it is far from an easy challenge to meet. If all these complexities make the prospect of public order policing sound even more bleak, there is at least one silver lining. That is, those who do not (at least initially) condone violence generally see the occurrence of violence as being against their interests. Protesters are aware that, should conflict occur, it will drown out their message in the media the next day. Their cause will be branded as illegitimate, their reasons for attending a protest will be undermined. Hence, at least potentially, the majority can be a resource to control the violent few. I recall one dramatic instance of this. It was an Anti-Apartheid Movement (AAM) protest in London, ending in Trafalgar Square, in 1986. Because of the upsurge of the struggle in South Africa itself, the event attracted far more people than had been expected. As a result, people were squeezed into the square, many of them pushed up against the lines of police in front of the South African Embassy on the east side. A few people became involved in scuffles. It looked like widespread fighting might break out. At that moment, the US civil rights leader Jesse Jackson was speaking. He appealed to the crowd, saying that if fighting occurred, the AAM and its message would suffer. Only the South African regime would gain. He got people to hold hands and to chant the slogan of the AAM: “Free Nelson Mandela.” Those who had been scuffling were isolated. They were overwhelmed by the peaceful majority. The fighting stopped rather than spread. The critical question, then, is how different forms of public order policing either pull the majority to the side of the police in maintaining order or else push them toward the side of the minority in challenging order. To put it slightly differently, the issue is not so much about whether conflict occurs or not—there may be some elements in the crowd who would be conflictual

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 17

FROM CRISIS TO OPPORTUNITY

17

whatever the police do. Rather it is about whether conflict escalates or deescalates. Our critique of the old psychology is that it contributes to escalation. By responding to the few by clamping down and denying the rights of the many, one could be decreasing effectiveness by pushing the many into the hands of the few, of giving influence and power to those who advocate violence. Our claim for the new psychology is that it explains how the new psychology can contribute to de-escalation. That is, by recognizing the different groups that coexist in a crowd event, by understanding the identities of each of them and hence which groups have unlawful intentions and which have lawful intentions, by allowing or even facilitating the latter, one will not only insulate these groups from the influence of unlawful elements, but may also win them as allies in the maintenance of order. Easy as this is to say, it does involve a profound shift in the way crowd events are approached. To start with, it means approaching the crowd with a very different set of questions in mind: not so much “what dangers do they carry and how can we stop them?” but rather “what are their aims and how can we help them?” As we have indicated, a new psychology of crowds implies new principles for policing crowds. These principles are the subject of the next section.

Four Principles for Public Order Policing The principles I outline here have been outlined elsewhere both in academic publications (Reicher et al. 2004, 2007) and in the British Association of Chief Police Officers (ACPO) manual on “Keeping the Peace” (ACPO 2003). In the light of G20, this suggests that something has gone wrong somewhere. One possibility, of course, is that the principles are simply wrong or unhelpful. But first, they have been implemented to some effect elsewhere both in Southern and Northern Europe—and the work of both Clifford Stott and Otto Adang (see their chapters in this volume) speaks to that. Second, the problems of G20 did not happen despite the use of our principles, rather they happened with (and, we will argue, because of) our principles being ignored. That makes it all the more important to reassert the importance of these principles and also to provide additional clarity about them. One important point to stress is that we are not referring to specific policies. We are not referring to specific procedures. We are referring to the general grounds on which policies and procedures can be devised and revised. By the same token, however, our principles can only be effective if they are applied to all aspects of the public order process from initial training of officers to the equipment and technologies used during events—passing via the process of intelligence, the development of tactics, the structure of briefings and debriefings, and much else besides. They cannot simply be invoked as an “extra” consideration alongside a set of practices based on very different conceptions of

02_Madensen_Ch02.qxd:Madensen

18

1/3/11

10:41 AM

Page 18

PREVENTING CROWD VIOLENCE

crowds. That is a recipe for incoherence. One can only have one set of principles, one understanding of what crowds are like and how to treat them. Take them all in all ways, or else leave them all. This point does not seem to have been appreciated. Thus, in the ACPO manual, the statement of our principles is preceded by a section on crowd behavior. There are many good points in this section. It acknowledges that one might find many different attitudes and behaviors in the crowd. It states explicitly that people don’t get carried away by “crowd hysteria.” But it mixes this with traces of the old crowd psychology. For instance, it is stated that when crowd members become highly emotional or aggressive their behavior can spread and influence those all around. It does suggest that anonymity removes moral constraints. It is, in sum, something of a halfway house between the old and the new. It is certainly not a thorough rethinking of principles and practices. So, to insist on our point, let us draw on Humpty Dumpty in Through the Looking-Glass: when I use the word principle it means just what I choose it to mean—neither more nor less.10 The principles are as follows: Understanding social identities in the crowd. The core premise of the new psychology of crowds is that the different psychological groups in a crowd event will behave on the basis of their respective social identities. If you want to know how members of these groups will behave, whether they will break the law or observe the law, what suggestions they will follow and what suggestions they will not follow, how they will react to different types of police intervention, then you need knowledge of their social identities. To put it in a different way, you need an insider’s understanding of the group culture. Now, of course the police do seek insider’s knowledge of groups in the crowd. They expend a considerable amount of effort in doing so. Indeed, in the aftermath of G20, this has been the subject of considerable controversy as revelations have emerged of a combination of inducements and threats being used to convince activists to spy on their fellows.11 Both the way information is sought and the type of information that is sought derives from the old perspective of crowds as trouble, however. The focus is on obtaining criminal intelligence information about individuals and their unlawful intentions. From this perspective, the information has to be obtained clandestinely. What is being suggested here is something wholly different. It is about how the group sees the world, what it believes in, how it goes about achieving its aims. This is information that can and in fact should be obtained openly, for then the process of obtaining the information is part of the more general process of being seen to respect groups and work with (rather than against) them. To put it slightly differently, if public order policing put as much emphasis on cultural intelligence as on criminal intelligence, it would be in a far better position to preempt conflict rather than contain it.

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 19

FROM CRISIS TO OPPORTUNITY

19

Differentiating between the various psychological groups in a crowd event. Once one understands the differences in terms of values, norms and

intentions between groups in a gathering, one can appreciate the importance of treating different groups in different ways. Just because some people in the mass throw things at the police does not mean that everyone is hostile to the police or liable to get carried away. Just because some want to attack property, smash windows, and loot banks doesn’t mean others will join in. Indeed, as we have argued repeatedly, violent groups generally gain influence over others only when the police act in indiscriminate ways and hence arouse generalized hostility. At every level, then, it is important to ensure that one does not treat all crowd members as suspects because of the suspect behavior of some among them. However, as we have seen, it is one thing to appreciate this point, it is quite another to know how to implement it. If, for instance, the tactical repertoire is limited to homogenizing practices, then changing beliefs will be insufficient. If the only technologies one has are either so inaccurate or so encompassing (say, tear gas or water cannon) that anyone in the vicinity might become a target, then differentiation can be no more than a fond hope. If police equipment (scuffed Perspex shields and scuffed Perspex visors) makes it impossible to even see the crowd beyond a blur, then it becomes impossible to aim actions at appropriate targets. Differentiation is not just an idea but also a social practice, which is why its implementation does not just require a sea change in ideas but also a reappraisal of every practical aspect involved in the policing operation. Facilitating the legitimate aims of crowd members. In a way, this is the crux of the matter. It is sufficiently important to bear repeating ourselves. If there is just one lesson to be taken from this whole chapter, just one change that needs to be made, it is this: public order policing will turn from a vicious circle of distrust to a virtuous circle of trust when public order police officers start with the question “how can we empower crowd members?” rather than “how can we control them?” and when crowd members, on seeing officers, think “they are a resource” rather than “they are an impediment.” Stott and his colleagues provide a beautiful example of the turnaround that can be achieved by contrasting the treatment of English and Scottish football fans at the 1998 World Cup in France (Stott, Huchison, and Drury 2001). English fans are seen as uniformly violent and, even if they initially reject violence, are uniformly treated as such. As a consequence, peaceful fans become increasingly alienated from the police, increasingly sympathetic to the “hard men” among them, and increasingly likely to become involved in collective conflict. By contrast, Scottish fans are perceived as colorful and goodhumored, and they are well treated by the police and other authorities. Consequently, should anyone begin to act violently, other fans intervene to control them.

02_Madensen_Ch02.qxd:Madensen

20

1/3/11

10:41 AM

Page 20

PREVENTING CROWD VIOLENCE

We are not, of course, suggesting that the police should facilitate everything that any group wants to do. Understanding when to facilitate is clearly dependent upon the first two principles described above. The police need to understand the intentions of different groups, to identify those that are lawful and hence to understand where and when facilitation is appropriate. Nonetheless, we are suggesting that facilitation rather than repression should be the default option. Start by being open and friendly. If in doubt, be helpful and communicative. Only change this approach where there is good evidence that the relevant individuals are pursuing a course of disruption. There are two reasons for this. First, in terms of overall crowd dynamics, how and why the police treat crowd members as they do is important, but equally important is how they are perceived to do so. Thus, to repress without self-evident reasons will contribute to the alienation of potential allies in the crowd, but to clamp down on those who cast the first stone (metaphorically and often literally) does not run that danger. Second, if one’s overall approach serves to gain the trust of the crowd majority, then the risks of (initially) facilitating those who turn out to be disruptive are diminished. That is because one can draw upon the crowds’ selfpolicing capacities as a second line of defense. All in all, the principle of facilitation is not about being soft but rather about being effective—or rather, it is about generating the most optimal balance of forces between those who support and those who attack the public order. Communicating the strategy of facilitation. Let us return, for a moment, to

the distinction between the aims of police action and the ways police action is perceived by crowd members. Let us also return to our old study of Notting Hill Carnival. For the police, one of the major issues in the carnival was public safety. There was a real danger of crushing; there was also a danger that if, for whatever reason, people needed medical attention, there was no way through for the medics. The police therefore cleared and then barricaded streets so that, in emergencies, they could reopen them as overflow areas or as routes for ambulances. For crowd members, one of the key themes of the carnival is about community self-determination. For one weekend, the black community in particular occupies and controls the streets. They dance and carouse wherever and whenever they like. So for them, the police had no right to put up barriers and tell them where they could and could not go. The barricades became sources of tension and sites of confrontation. The point, then, is that there is a problem with the police doing things to help crowd members without negotiating and communicating what they are doing. This negotiation needs to start before events, it needs to be communicated to all crowd members as they come to events, and there needs to be continuous communication as events unfold. What is more, the most effective communication will come from inside, not outside the crowd. Let me illustrate this with another anecdote that I have used elsewhere.

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 21

FROM CRISIS TO OPPORTUNITY

21

Years ago, I was attending a football match with a friend and his father. They had both been Bournemouth fans all their lives and now, for the first time, Bournemouth (a small seaside town and a minor league club) was about to be divisional champion and promoted to a level they had never reached before—dependent on the result of the game, that is. The ground was packed. An announcement was made that, if victorious, the team would parade around the ground with the championship trophy. However, this could only happen if fans refrained from invading the pitch. Bournemouth scored. The fans erupted. A few ran on the field. Another announcement reminded people that more invasions meant no victory parade. When those who had gone on the pitch tried to evade the police and regain the safety of the crowd, people simply wouldn’t let them back in. Officers walked up to them and led them gently away, cheered on by the fans. To reframe this anecdote in analytic terms, the initial announcements communicated to the crowd how the club authorities sought to give fans what they craved. They also communicated the constraints on their ability to deliver. After the goal, the announcement communicated how those violating the regulations were acting to deprive the fans of what they craved. In this context (1) “invaders” were perceived to be impeding group goals; (2) the fan majority worked with the police to contain the invaders; (3) the fans perceived the police as facilitating their goals. Moving now to even more general terms, this anecdote illustrates how effective communication can indeed play its part in pulling the crowd majority toward the police rather than pushing them toward a disruptive minority.

Conclusion: A Counterfactual Analysis of G20 I have argued that we need a root-and-branch rethink of public order policing and that this needs to be based on new principles based on a new understanding of crowd psychology. Conversely, in outlining those principles, I have alluded to a whole series of areas in which they might impact on public order policing— everything from intelligence to tactics to technology and even to the protective equipment that officers wear. There is no corner that can remain untouched. As I have already stated, it is not my intention to go further and outline in any detail the precise practices and procedures that might flow from the new principles. First, I don’t have the space, and second, others—having done far more than I have in thinking through these issues, designing new public order practices, and evaluating their effects—can do so much more effectively. However, it is perhaps too tantalizing to leave it at that. Moreover, since I started with a critique of policing at G20, and more particularly with a critique of the kettling tactic, perhaps it is incumbent upon me to say what I think should have been done differently.

02_Madensen_Ch02.qxd:Madensen

22

1/3/11

10:41 AM

Page 22

PREVENTING CROWD VIOLENCE

Kettling, it will be recalled, involves the containment—often over a prolonged period—of all the protesters in a given site. It is a tactic that pays little heed to the different psychological groups included in a single physical gathering. It certainly does not differentiate between them but rather treats everyone as trouble. It is the converse of facilitation—it simply stops everyone from doing whatever it was that they were about. And it involves precious little communication of any sort—whether it is a matter of explaining why people are being contained or what is going to happen to them. As a tactic it perfectly implements the assumptions of the old crowd psychology and perfectly violates all the principles of the new crowd psychology. It leads to a common experience of police illegitimacy—often exemplified by the humiliation of individuals who are patently of no threat to the police. It makes the most confrontational of individuals the most representative and hence the most influential among the group as a whole. It tilts the balance of forces strongly against the police. So, acknowledging that there may be sensitive sites in the vicinity that some protesters would be eager to attack (and supposing, at some level, there was strong political pressure upon the police not to allow any such attacks to occur [see Cronin and Reicher 2009]), what else could have been done? Well, first, any action should proceed from an understanding of the different groups among the protest, their different perspectives, and their different goals. Second, the police could communicate their intentions to facilitate groups in pursuing lawful goals but explain how this is impeded by the presence of some who have disruptive goals. Third, temporary containment could then be used along with selective filtering as a means of clearly differentiating between those allowed to pursue lawful goals and those kept back to impede their unlawful goals. What is critical here is clearly the notion of a selective filter and how the police are seen to identify and release the most vulnerable and benign sections of the crowd. To the extent that this could be achieved, then disruption inside the “kettle” would serve to impede their release and hence be isolated as action against (majority) crowd interests. Violence would become counternormative for the majority and hence those advocating violence would become isolated. Like any counterfactual analysis, my aim is not to suggest that we know exactly what would have happened if a different approach had been employed or that I have some sort of magic wand for turning violent riots into peaceful picnics. Nor am I underestimating the massive logistical difficulties involved in my proposals. Not least, communicating to a large group of people in noisy and fraught circumstances would require entirely new communications technologies (large portable sound systems or LED screens; helicopter-mounted communications, etc.) with considerable costs attached. What I am suggesting, however, is that the approach and the principles I have sketched out can help us begin to think of alternatives to present practice. They are aimed to stimu-

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 23

FROM CRISIS TO OPPORTUNITY

23

late a creative debate about how to move our approach to public order forward, how to maximize the right to protest while minimizing the threat of violence, and hence how to safeguard a vigorous participative democracy.

Notes 1. See www.guardian.co.uk/uk/2009/apr/07/ian-tomlinson-g20-death-video. 2. See www.guardian.co.uk/uk/video/2009/apr/08/g20-police-assault-ian-tomlinsonvideo. 3. See www.youtube.com/watch?v=6THfDGy1hN4. 4. Taken from news.bbc.co.uk/1/hi/england/london/7989027.stm. 5. Taken from www.londonfgss.com/thread19035-17.html. 6. Taken from www.guardian.co.uk/uk/2009/apr/07/ian-tomlinson-g20-deathvideo. 7. See www.thisislondon.co.uk/standard/article-23678442-details/G20+police+’ were+too+violent’/article.do. 8. See www.thisislondon.co.uk/standard/article-23682001-details/Four+in+10+ Londoners+don’t+trust+Met+after+G20+death/article.do. 9. This resulted in a TV program made by the BBC for the Open University’s social psychology course D317 entitled “Controlling Carnival Crowds.” 10. The way I label the terms here is different from the way we have done so before (Reicher et al. 2004, 2007). The previous labels were somewhat more abstract and hence more cryptic. Here I sacrifice brevity for clarity. 11. See www.guardian.co.uk/environment/2009/apr/27/police-spies-environmentalistsscotland-protesters.

02_Madensen_Ch02.qxd:Madensen

1/3/11

10:41 AM

Page 24

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 25

3 Crowd Dynamics and Public Order Policing Clifford Stott

IT IS IMPOSSIBLE TO ADEQUATELY UNDERSTAND THE RELATIONships between crowd psychology and public order policing without discussing the historical origins of crowd theory and its linkages to political protest and urban unrest in Western Europe. This is perhaps most evident from nineteenthcentury France, the birthplace of “scientific” crowd psychology. From the revolution of 1789 onward, France was beset by political instability linked directly to the capability of crowds to threaten or overthrow successive political regimes. It was in this context that crowd psychology emerged as a “scientific” basis for understanding, controlling, and undermining the threat that crowds were posing to the emerging industrial capitalist economy (Barrows 1981; Nye 1975). The collapse of the French military during the Franco-Prussian War of 1871 and the declaration of the socialist Paris Commune of the same year inspired one of Europe’s then most respected academic historians, Hippolyte Taine, to write a detailed social history of France. The book was much more than some neutral and objective overview of a century of political instability. It was actually inspired as an attempt to use science to try to address the malaise of the French state. Eventually the work ran to eleven volumes collectively called the Origines de la France Contemporaine. What is relevant about Taine’s treatise is that its opening volume, L’Ancien Régime, provided the first modern (i.e., post-Darwinian positivist), pseudo-scientific account of the psychology of crowd behavior (Taine 1876). Taine is now acknowledged to have been a political conservative who sought to build a sustained intellectual assault on the social theories of the Enlightenment and in particular Jean-Jacques Rousseau’s social contract, which he understood as an ideological poison injected into society (McClelland 1970, 1989). Central to this political project was a very specific and rather detailed theory that pathologized both the crowd and its psychology and placed this pathology as the primary cause of the collapse of French political 25

03_Madensen_Ch03.qxd:Madensen

26

1/3/11

10:24 AM

Page 26

PREVENTING CROWD VIOLENCE

power. In the Origines, Taine proposed that the crowds of nineteenth-century France were actually “mobs” populated by the “lower orders” who were subject to “vibration of the nervous mechanism,” “contagion,” and “feverishness,” and thus easily influenced by “criminals” and “fugitives from justice.” He proposed that this left a highly evolved hierarchical French society open to the primitive barbarity and irrationality of the mass; pathology made concrete through the actions of the revolutionary crowd. Taine’s political analysis was therefore tied fundamentally to ideas about how the evolved social “order” needed to be protected from the inherent irrationality and pathology of the crowd. Indeed, he explicitly recommended that the forces of order were required to act as a “dyke” to resist the “torrent” of the brute forces of the crowd, “despotic if need be against their despotism” (Taine 1876, 242). It is this link between a perceived pathology of the crowd and the use of force to resist such pathological intrusion that remains the fundamental problem of public order policing in democratic societies. It is understood by contemporary academics that Taine’s historical analysis was politically biased and unsubstantiated by firsthand data as would be expected of scientific research today. However, Taine’s “scientific” work was published at a time when the crowd and mass society was becoming an everincreasing theoretical and political concern throughout Europe. It was hoped by the political classes of that time that the emerging positivist sciences would not simply supply a valid understanding of the problem of the crowd but also dictate solutions (McClelland 1989). It was Taine’s social history that provided the theoretical model, assumptions, and scientific credentials for just such a perceived resolution. However, as Stephen Reicher points out (in Chapter 2 of this volume), the central figure in popularizing this form of crowd psychology was without doubt Gustave Le Bon. His classic work, The Crowd, is acknowledged as the most widely read social psychology text of all time (Le Bon 1895) and is acknowledged to be bedside reading for the likes of Hitler and Mussolini (Farrell 2004). But Le Bon’s work was influential not due to its scientific validity, since he like Taine did not draw upon data, but because Le Bon took the findings of the closed societies of “science” and disseminated them to a public and political elite eager to understand and control the crowd at a time of increasing threat of socialist revolution (Barrows 1981; McClelland 1989; Nye 1975; Reicher 1987). Le Bon essentially recapitulated in more technical form Taine’s pathologization of crowd psychology. He proposed three central mechanisms: submergence, contagion, and suggestibility. Submergence refers to the loss of the individual identity among participants through the anonymity assumed to be inherent in the crowd. Contagion refers to the uncritical social influence mechanism that subsequently emerges that allows any idea or sentiment to spread unheeded through the crowd. And suggestibility is the hypnotic psychological state induced by submergence that allows contagion to occur. Le Bon pro-

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 27

CROWD DYNAMICS AND PUBLIC ORDER POLICING

27

posed that through these mechanisms behavior within a crowd is no longer governed by individual rationality but by the “law of mental unity.” That is, in the crowd, the individual self or identity disappears, to be replaced by what Le Bon referred to as a “group mind” or “racial unconscious”—characterized by reduced intelligence, atavistic impulses, and emotionality. Thus, Le Bon proposed that through these mechanisms the civilized lone individual descends “several rungs of civilization” and in the crowd “is a barbarian” (Le Bon 1895). There have of course been various challenges to the “classic” crowd psychology promoted by Le Bon that are too numerous and detailed to deal with in this chapter.1 These critiques have primarily been built around the dominant view within the social historical literature that the actions of the crowds of nineteenth-century France were not random explosions of meaningless violence as classic theory would suggest but actually highly structured and normative. As such it is argued that these crowd events must be understood as meaningful collective responses to the actions of the state and its forces of order among those disenfranchised and disempowered by the Industrial Revolution. In other words, the psychology and behavior of these crowds were fundamentally linked to the social understandings of crowd participants and to the social and intergroup processes of the context in which these crowd events occurred (Barrows 1981; Nye 1975; Reddy 1977; Rudé 1964; Tilly, Tilly, and Tilly 1975; Reicher 1987). However, given that the scientific crowd psychology of Taine and Le Bon systematically ignores the possibility of the rationality and social determination of crowd action, it simply cannot provide an adequate conceptual framework for understanding the behavior of these, or any other, crowds. Nonetheless, this classic theoretical account of crowd psychology also attracted powerful empirical support within the US mainstream of experimental social psychology during the latter half of the twentieth century. This support came primarily from the tradition of “deindividuation” research originated by Leon Festinger, Albert Pepitone, and Theodore M. Newcomb (1952) and subsequently developed by others (e.g., Zimbardo 1969; Diener et al. 1976). Put very simply, the deindividuation model proposes that “anonymity” within crowds, a concept drawn directly from Le Bon, reduces self-awareness, and in so doing increases the tendency for individuals to express violent, aggressive, and “antisocial” behavior. However, Reicher (1987) provided a powerful critique of deindividuation, which culminated in a meta-analysis of sixty independent experimental studies of the effects of deindividuation conditions, and which demonstrated that the data from these studies does not support and therefore seriously undermines the scientific validity of deindividuation or classic theory (Postmes and Spears 1998). Therefore, the classic crowd psychology originating in late nineteenthcentury France, but developed by modern experimental social psychology,

03_Madensen_Ch03.qxd:Madensen

28

1/3/11

10:24 AM

Page 28

PREVENTING CROWD VIOLENCE

quite simply lacks any sustainable empirical support. The problems of classic theory begin because it decontextualizes crowd behavior and as such ignores any possibility of a determining role played by the actions of the forces of order or the broader social context in which crowd events occur. Consequently, it cannot explain the meaningful nature of collective action during crowd events and must be considered as seriously flawed. Nonetheless, it has produced a very specific and widely adopted theoretical model of crowd psychology that is still maintained within the vast majority of mainstream psychology textbooks, particularly those oriented toward the US market. This model assumes that crowd behavior is ultimately a pathological intrusion into modern society because mechanisms inherent to the crowd undermine individual identity and the rational psychological basis of behavior. In effect, classic theory proposes that individuals within crowds are uniformly dangerous and unpredictable because they can spontaneously coalesce into irrational and violent mobs simply due to mechanisms internal to the crowd. Moreover, as a function of these mechanisms, mobs can easily come under the influence of agitators, troublemakers, and hooligans. Given that from this perspective crowds are understood as unpredictable, volatile, and dangerous, it becomes almost self-evident that they need to be controlled, and this control must be exerted primarily through the use of force.

The Social Identity Approach to Crowd Behavior The most important thing that must be said about classic theory therefore is that it is now completely outdated and has been widely rejected within the mainstream of European social psychology (see Reicher 2001b). Given the various problems regarding its validity and explanatory power outside the world of social psychology textbooks, various alternative theoretical models have superseded the Le Bonian or “classical” account. In particular, it has been replaced in the research literature by the now leading theory of crowd psychology commonly referred to as the Elaborated Social Identity Model of crowd behavior (or ESIM).2 The ESIM has as its basis the proposition that a component part of the self-concept determining human social behavior derives from psychological membership of particular social categories (e.g., a policeman or demonstrator). Consequently, as well as having an idiosyncratic personal identity (i.e., an identity as a unique individual), crowd participants also have a range of “social identities,” which can become salient within the psychological system referred to as the “self.” Collective action becomes possible when a particular social identity is simultaneously salient and therefore shared among crowd participants. Thus, and as Reicher (this volume) points out, being in a crowd does not entail a loss of identity so much as produce a shift in the focus of self-

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 29

CROWD DYNAMICS AND PUBLIC ORDER POLICING

29

definition among crowd participants away from unique individual attributes to the more shared, group-based defining attributes of the crowd. Moreover, acting in terms of a social identity means that there will be an increased tendency among those in the crowd to adhere to the norms, values, and ideology of that social category (e.g., antiwar demonstrators acting in terms of an ideology of nonviolent demonstrative identity at the outset of a demonstration). The shared identity also determines who and what can become influential in the crowd and gives crowd participants some basis for judging whether they are able to act in particular ways by signifying if others in the crowd are likely to support particular actions (e.g., intervention to prevent or support an argument with police). Salient social identities therefore orientate crowd members toward meaningful forms of collective action in any given context, in terms of both what behavior is seen by them as appropriate or legitimate and what behavior they feel is possible. From its original proposition (Reicher 1982, 1984, 1987), the social identity approach to crowds has, unlike the classic account, been underpinned by substantial empirical evidence. This body of data began through analysis of one of the major inner-city disturbances in England in the 1980s (Reicher 1984, 1987) and gained further empirical support through studies of student “rioting” during a protest against the removal of grants (Reicher 1996), demonstrator “rioting” during a protest against the implementation of a tax (Drury and Reicher 1999; Stott and Drury 2000), a series of protests around the extension of the M11 motorway in London (Drury and Reicher 2000), and “rioting” among football fans attending football matches with an international dimension (Stott, Hutchison, and Drury 2001; Stott and Reicher 1998a; Stott and Pearson 2007). This theoretical and empirical development to the ESIM has highlighted the importance of police tactics in shaping and reshaping a crowd’s social identity and therefore in determining the nature of the collective behavior that occurs during a crowd event over time, particularly in terms of the emergence and escalation of “rioting” (Stott and Reicher 1998b). There were similar dynamics evident in each of the studies. Where police used relatively indiscriminate tactics of coercive force (e.g., baton charges), they would tend to do so against those in the crowd who saw themselves or others around them as posing very little, if any, threat to public order. As a consequence there would be corresponding increases in the number of people in the crowd who perceived the police as an illegitimate force. Such interactive social psychological processes occurring during the crowd event would then lead directly to a change in the nature of the crowd’s social identity (their shared sense of categorization of “us” and “them”) along two critically important dimensions defined by the ESIM in terms of perceptions of legitimacy and power. On the one hand, the indiscriminate use of force would create a redefined sense of unity in the crowd in terms of the perceived illegitimacy of and opposition to the actions of the police. Consequently, there would be an increase in

03_Madensen_Ch03.qxd:Madensen

30

1/3/11

10:24 AM

Page 30

PREVENTING CROWD VIOLENCE

the numbers within the crowd who would then perceive conflict against the police as acceptable or legitimate behavior. On the other hand, this sense of unity and legitimacy in opposition to the police would subsequently increase the influence of and empower those prepared to engage in physical confrontation with the police.3 Such processes could then draw the crowd into conflict even though the vast majority had no prior intention of engaging in conflict. In other words, the development of widespread rioting was not simply a product of mechanisms internal to the crowd nor of the predispositions of crowd members. Rather the psychology and social processes that made a riot possible emerged as the outcome of specific forms of group-level interactions that were largely and inadvertently initiated by police tactical responses.

Classic Crowd Psychology and Public Order Policing in the UK The ESIM is consistent with the social historical evidence cited earlier concerning the behavior of crowds in the nineteenth century (e.g., Rudé 1964). Moreover, it is also consistent with a large body of contemporary evidence and theory from criminology suggesting that there is indeed a very intimate relationship between the dynamics of crowd conflict and the tactics of public order policing (della Porta and Reiter 1998a, 1998b; Hall and de Lint 2003; Jefferson 1990; King and Waddington 2005, 2006; Sheptycki 2002; D. Waddington 2007a; P. A. J. Waddington 1987, 1991, 1993a, 1993b). Much of this literature points to a progressive global transition, post–World War II in democratic societies, from “reactive” policing primarily grounded in the threat and use of force to a more “proactive” consent-based approach that relies less on the use of force and more on communication and negotiation. This distinction is often contrasted in terms of a strategic orientation to “escalated force” versus “negotiated management” (McPhail 1997). Yet there is a strong debate within the same literature about how best to deliver the strategy of “negotiated management” tactically (McCarthy and McPhail 1998; McPhail, Schweingruber, and McCarthy 1998; Hall and de Lint 2003). Despite their differences, what the contrasting perspectives share with ESIM is the assumption that aggressive police tactics can and do have the capacity to negatively impact upon crowd dynamics. Moreover, despite evidence of these general trends in public order policing, these researchers have also highlighted how the two strategic and tactical approaches can actually be applied simultaneously by different police forces or units within the same country and even within the same event (Adang and Cuvelier 2001a, 2001b; della Porta and Reiter 1998a; Sheptycki 2002; Hall and de Lint 2003). Given this variability is acknowledged, the research literature on public order policing therefore raises two fundamental questions: What governs the application

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 31

CROWD DYNAMICS AND PUBLIC ORDER POLICING

31

of different styles of public order policing, and what are the underlying social psychological processes that determine their positive and negative impacts upon crowd dynamics? It is the ESIM that is the primary theoretical model that addresses the latter question.4 Perhaps the primary theoretical model that seeks to address the former is that developed by della Porta and Reiter (1998a, 1998b). They propose a multiplicity of interrelated factors, such as the surrounding legal framework, the culture of the police, the political context, and the pattern of interaction between the police and the crowd. They go on to argue that these variables then function together to produce “police knowledge,” which in turn determines the style of policing that is adopted. There is therefore considerable support within the literature for the suggestion that any adequate explanation of crowd behavior and public disorder must include an analysis of police knowledge and behavior. Within the literature addressing the issue of police knowledge, the evidence suggests that one of the problems facing public order policing in the UK is that police training within England and Wales deals very little, if at all, with the background scientific literature. During systematic observations of a range of training courses for public order policing in England and Wales it was evident that there was little if any formal discussion of the background literature relating to crowd psychology or public order policing (Hoggett and Stott 2010a).5 Nonetheless, the research suggests that implicit in much of the discussions about tactics were assumptions derived from the work of Le Bon. Given that classic theory has been extensively endorsed by social scientists from the nineteenth century onward, it should not be surprising that this research goes on to suggest that classic theory has become institutionalized in public order training within England and Wales (Hoggett and Stott 2010a). Indeed, supporting this contention, classic theory appears in the preclass reading material for the 2006 version of the National Police Improvement Agency (NPIA; then National Centre for Policing Excellence) Public Order Commanders Course, which states, for example, that “a crowd is a device for indulging ourselves in a kind of temporary insanity by all going crazy together.” The document goes on to state that “all psychologists seem to agree, that membership of a crowd results in a lessening of an individual’s ability to think rationally, whilst at the same time his/her more primitive impulses are elicited in a harmonious fashion with the emerging primitive impulses of all the other crowd members. The result being the establishment of a collective mind. As you can see, all these theorists have made points concerning crowd behaviour which contain a lot of truths” (p. 1). Classic theory also appears in the input regarding crowd psychology in the 2007 version of the NPIA training for the policing of events (a course designed to support police training for the 2012 Olympics). The 2008 documentation regarding learning outcomes from this course states that trainees should be able to “describe the make up of a mob.”

03_Madensen_Ch03.qxd:Madensen

32

1/3/11

10:24 AM

Page 32

PREVENTING CROWD VIOLENCE

They should understand how “based upon the research of Le Bon, a group of people may be termed either ‘crowd’ or ‘mob’” and that “classical theories of crowd formation discuss how the mob is formed from a crowd” (p. 4). However, ESIM-oriented research suggests that a central problem with this theoretical perspective on crowd psychology is not simply that it is outdated and inconsistent with the scientific literature. Rather, that it leads to a perception among the police of crowds as inherently irrational and uniformly dangerous. Indeed, the research evidence supports the idea that as a consequence of this training police of all ranks in England and Wales tend to see the general heterogeneous composition of crowds in terms of a simple dichotomy—an irrational majority and a violent minority who can easily assert influence over the crowd. Moreover, where this understanding is in place, there is also an increase in police officers’ support for and use of tactics that rely upon the use or threat of indiscriminate force. A convergence of these factors of “police knowledge” then interacts with the practical constraints of operational policing (e.g., tactical formations, protective equipment) in a way that increases the likelihood that the police will use indiscriminate force against whole crowds even when it is only a minority that are judged to be posing any threat to public order (Drury, Stott, and Farsides 2003; Hoggett and Stott 2010a, 2010b; Stott and Reicher 1998a). The ESIM and its supporting evidence also suggest that such indiscriminate use of force can then somewhat ironically contribute to a widespread escalation in the levels of public disorder. For example, during the “anti–poll tax” demonstration in central London on March 31, 1990, a small sit-down protest combined with some minor confrontation outside Downing Street led very rapidly to forceful police intervention against large sections of the demonstration. The evidence derived from a systematic study of this event suggests that the police decision to use force in this way was driven by the classic view of crowd psychology and dynamics. As such, the actions of a minority within the crowd just outside Downing Street led to the perception among police commanders that the entire crowd in this vicinity was likely to become disorderly. As a consequence, a decision was taken to disperse the crowd from the whole of Whitehall. But this research evidence also suggests that people in the crowd in Whitehall could see no threat to public order, just legitimate civil disobedience against what was seen by demonstrators as an unjust system of taxation. Consequently large numbers of demonstrators came to perceive the indiscriminately forceful intervention of the police as an attack on democratic rights. In technical terms, people in the crowd began to collectively perceive their intergroup relationship with the police as illegitimate. Thus, the relatively indiscriminate use of force interacted with crowd dynamics in such a manner that it led to a change in the crowd’s social identity. This change meant that there was an emergent psychological unity and shared perception within the crowd

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 33

CROWD DYNAMICS AND PUBLIC ORDER POLICING

33

that participants were justified and powerful enough to confront the police. Of course, such emergent and escalating hostility was likely to confirm initial police views that the crowd was becoming disorderly, which would correspond with the further increase in the scale and intensity of forceful intervention that was observed. Thus the research suggests that a process of interaction then cycled back and forth until it culminated in one of the largest riots ever witnessed in central London (Stott and Drury 1999, 2000). But it should not be assumed that there is a simple, deterministic process from crowd theory to operational practice. For example, it is evident from various observations of the policing of football crowds in England and Wales that a good deal of police good practice does occur with respect to managing crowd dynamics despite the overwhelming dominance of classic theory (Stott et al. 2008). However, it is apparent that the way this practice is understood is largely experiential (i.e., good practice that is linked to police operational experience rather than theoretical knowledge about the psychology governing collective action within a crowd). With an absence of the correct theoretical knowledge, experienced police officers are not then in a position to fully understand the general principles through which this good practice is having its positive impact. It therefore becomes more challenging to extract these principles of good practice from specific local contexts, integrate them into training, and apply them to other situations in a manner that assists the police locally and nationally in disseminating and coordinating good practice (Hoggett and Stott 2010b).

Good Practice in the Management of Public Order During Crowd Events It is a central theoretical position of the ESIM that collective disorder can and does emerge directly from the social identity processes and intergroup dynamics that materialize during crowd events, dynamics of which police knowledge and tactics are ubiquitously a component. Therefore, the central advantage of the social identity–based model of crowd psychology is that it can and does begin to identify the key processes underlying the emergence and escalation of crowd conflict—processes within both the crowd and the police (e.g., Cronin and Reicher 2006, 2009). By the very same logic, however, it is also able to provide a set of theoretical concepts and principles through which the likelihood of collective conflict can be reduced. Thus, Reicher et al. (2004, 2007; Chapter 2 in this volume) have outlined a set of ESIM-derived principles relating to the effective management of crowds; principles that have been partially incorporated within the UK into the 2003 version of the Association of Chief Police Officers’ manual of guidance for policing public order, “Keeping the Peace.”

03_Madensen_Ch03.qxd:Madensen

34

1/3/11

10:24 AM

Page 34

PREVENTING CROWD VIOLENCE

These principles propose that prior to a crowd event it is important for police forces to educate themselves about the cultural norms of crowds and the legitimate intentions for participants that flow from these. It is recommended that the subsequent police strategy is not oriented exclusively toward the control of the crowd through the use or obvious threat of force but also the effective facilitation of the legitimate intentions underpinning crowd action. The police tactics for managing the event should then be capable of communicating police intentions to facilitate crowd members’ rights, for example, to protest. Most important of all, if and when the police are required to use force, they should differentiate between groups and individuals within the crowd and above all avoid the indiscriminate use of force (Reicher et al. 2004). A central feature of the development of the ESIM has been its application to the management of crowd dynamics in relation to football matches with an international dimension (see Stott and Pearson 2007 for a comprehensive overview). For example, analysis of English and Scottish fan collective behavior during the 1998 Football World Cup Finals in France showed that while perceptions of illegitimacy underlay conflict escalation in crowds, perceptions of legitimacy underlay absences of collective conflict (Stott, Hutchison, and Drury 2001). What these studies also reveal is the complexity of the relationship between police tactics and crowd behavior, particularly in terms of crowd participants’ views of the legitimacy of police tactics and the appropriate junctures for police use of force. Nonetheless, this research confirms that perceptions of police legitimacy are critical because they affect the crowd’s internal dynamics, facilitating or undermining the ability of those seeking conflict to exert social influence upon others in the crowd (Stott et al. 2007). What these studies of football crowds also highlight is the need to conceptualize risk to public order as a process arising from the operation of the group-level dynamics that occur during crowd events (both within and between the crowd and police). In turn, the effectiveness of public order policing can be understood and analyzed in terms of its ability to proactively manage these group-level dynamics in a way that minimizes the potential for large-scale incidents of disorder to emerge. Put slightly differently, the ESIM suggests that the effectiveness of police strategy and tactics for policing crowd events should be evaluated by their impact upon the psychological crowd dynamics of legitimacy and power. However, fully testing the value of this ESIM approach to public order policing requires (1) systematic research on the policing tactics and strategies that influence public order dynamics, and (2) the application of ESIM principles to develop specific guidelines for public order policing that are then successfully implemented at an operational level. An important step toward generating such links between research and practice was taken in 2001, when the UK Home Office began to fund a program of studies of the policing of English football fans traveling abroad. This

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 35

CROWD DYNAMICS AND PUBLIC ORDER POLICING

35

research, led by the University of Liverpool, paid particular attention to developing a model of “good practice” in public order policing with regard to the policing of high-risk fan groups traveling to football matches with an international dimension. The intention was to subsequently make this model available such that it could be adopted and implemented by police forces across Europe, but in particular by the Portuguese authorities during the 2004 European Football Championships. The subsequent study observed crowd management practices at more than thirty-five fixtures with an international dimension, all categorized as high risk, which took place in nine European states. Data were gathered from police, fans, and other relevant agencies and were then analyzed to develop a theoretical model of “dynamic risk” to public order and a model of a “balanced” or “graded” tactical police profile (Stott and Adang 2009). Given that the identities that drive collective action are open to change, the model of dynamic risk proposes that threat or risk to public order during a crowd event should be conceptualized as a continuum. Movement along this continuum from low to high risk can be affected by the perceived legitimacy of the intergroup relationships surrounding the crowd. The analysis suggested that a key factor governing movement along the risk continuum was police tactical profile. In particular, where a “balanced,” or “low-profile,” form of public order policing was in place, this appeared to be most effective at minimizing widespread hooliganism. The critical underlying process was the police capability to match or balance their tactical profile proportionately to the actual levels and sources of risk to public order that were present during the event. To achieve this, the host force would need to develop a graded tactical profile and ensure that their profile of policing was capable of rapid change in a manner that accurately reflected the levels and sources of risk. But this required the host police to undertake constant “dynamic risk assessment” and to address their information flow and their command and control capability to ensure that these assessments of risk subsequently informed tactical responses. On this basis, a model of police “good practice” was developed. First, the initial stage of the operation should involve background preparation characterized by research not just upon those fans known to pose a risk but also on the underlying culture of the fan group to be policed (e.g., What kind of behavior do they engage in? What will be their primary legitimate motivations and intentions?). This understanding of fan culture should then be mapped onto the local context, and, most important, consideration should be given to how the situation can be organized to facilitate the fans’ legitimate intentions. Priority should be given to communicating with fan organizations about what behaviors will and will not be tolerated or where fans might like to gather to be together in a safe environment (e.g., establishing “fan zones” with access to alcohol in popular areas). The second stage involves initial contact with the mass of visiting fans, a stage that should be characterized by low-impact visibility, information gather-

03_Madensen_Ch03.qxd:Madensen

36

1/3/11

10:24 AM

Page 36

PREVENTING CROWD VIOLENCE

ing, and monitoring. This stage is very important for establishing the perceived legitimacy of the relationships between the local police and fans, for defining the intention of the police to facilitate fans’ legitimate behavior, and for identifying those seeking to be confrontational. To this end, it was suggested that police officers should be deployed in pairs or in small groups in standard uniform, preferably dispersed widely across and within crowds; if riot squad officers were to be used, they should be deployed also in pairs or small groups without helmets or shields and with batons sheathed. The police should engage in high levels of positive interpersonal interaction with fans (nonaggressive posture, smiles, nods, accommodating requests for photographs, etc.). Where language is not a barrier, officers should try to communicate with fans to gather information about their demeanor, intentions, concerns, and sensibilities. This initial stage would then be effective at assisting the host force in actively constructing perceptions of police legitimacy within the crowds and gathering information about the overall level of risk and in particular the activity and intentions of groups and individuals within crowds who may or may not be posing threats to public order. There would, of course, be situations in which risk is identified. Depending upon the nature of these risks, the model of good practice proposed that certain forms of escalation in police deployment would be effective. This second stage of the graded tactical profile should be characterized by the firm but targeted communication of tolerance limits, and some increased visibility of the police’s capability to use force. Certainly, officers monitoring fan behavior in stage one will have identified those who are judged to be posing a potential risk. Further to this, a process of communication should be engaged defining the situation (so that all those present understand the police’s awareness of the problem) and communicating to those posing the risk that they are provoking the potential for police sanction. Moreover, those fans acting legitimately should be allowed to leave the vicinity and given some time to impose “selfpolicing”—so that those within the crowd may begin to control those posing the problem. If a riot squad is used at this stage, they should still be in standard or “peace-level” uniform (i.e., not wearing helmets or holding shields at this stage and batons in sheaths). Should these measures not deal with the situation, then it was proposed that further escalation may be required. This third stage should normally be characterized by targeted intervention and the removal of situational risk factors (i.e., not simply targeting known hooligans, but making an intervention that is fully informed of the factors that are actually causing the problem). The objective of police deployment at this stage is to minimize risk, and it is therefore essential that actions are not conducted that actually serve to escalate tensions. Therefore, as this escalation of force takes place, it is vital that valid intelligence about the source of the risk be communicated clearly to the intervention squads being deployed. It was made clear that it is simply not sufficient to act against a whole crowd of fans who happen to be present, unless

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 37

CROWD DYNAMICS AND PUBLIC ORDER POLICING

37

there is validated situational evidence that they are uniformly seeking to provoke disorder. Beyond this, there were, of course, options for increased use of force including baton charges, water cannon, gas, and rubber bullets—but it was made clear that throughout the study the researchers had yet to witness a situation in a football context in which such weaponry had been used either appropriately or effectively. It was made clear once again that indiscriminate intervention would only initially increase the level of risk, but if police use of force was evidence led and accurately targeted, the intervention, even at this level, would not be seen as illegitimate by the majority of the crowd. Therefore, police would be able to deal effectively with those issues threatening public order without themselves escalating the disorder. Having achieved this, there should then be a de-escalation policy in place to ensure an early restoration of normal-level policing. Thus, the research suggested that by continuously acting to gather information during events police were more capable of intervening in ways that achieved a differentiated use of force and engaged with risks to public order much earlier in the potential cycle of escalation. Moreover, where there was a strategic orientation of the police toward the facilitation of fans’ lawful intentions and behaviors, this tended to be combined with the capability to use nonconfrontational tactical options (such as interacting positively with fans). This facilitated police ability to gather “real-time” and valid intelligence of any emerging risks and to react rapidly to them in a manner that avoided the indiscriminate use of force. These tactics in turn appeared to be associated with shared perceptions among fans of the legitimacy of their relationship with the police. Moreover, there was evidence of lower levels of disorder and fans marginalizing hooligans and even seeking to prevent disorder by self-policing those fans behaving in a manner likely to provoke aggressive police response (Stott and Adang 2004, 2009). In other words, a graded tactical profile was effective at managing the group-level and social psychological dynamics of crowd events occurring within the context of football matches with an international dimension in Europe.

Application and Validation at Euro2004 This model of good practice and underlying theory made its most significant impact upon policy and operational practice through a series of consultations with the Portuguese Public Security Police (Polícia de Segurança Pública, or PSP), who then utilized it in the development of their “use of force” strategy for the 2004 Union of European Football Associations European Championships. Therefore, the security policy for the tournament was in line with the research evidence and ESIM psychological theory.6 The PSP’s public order

03_Madensen_Ch03.qxd:Madensen

38

1/3/11

10:24 AM

Page 38

PREVENTING CROWD VIOLENCE

strategy and tactical model focused upon a four-stage graded intervention policy that was oriented toward facilitating fans’ legitimate intentions and avoiding unnecessary or indiscriminate use of coercive force. Subsequently, the PSP cooperated with a major research study into the policing of the tournament and its impact on crowd psychology and behavior, the data from which confirm the successful implementation of their strategy and tactics on the one hand, and extremely low levels of “disorder” on the other.7 More specifically, the data from this research suggests that where crowds were gathered (even during events defined by the authorities as posing increased risk), there was often no obvious uniformed police presence. Instead, police in standard uniform, working in pairs and small groups, would integrate themselves into crowds of fans. The role of these officers was to communicate with fans, facilitate (e.g., provide directions or advice), monitor for emergent risk, and, if necessary, to provide a capability for dealing with low-level problems (e.g., minor arguments between individuals in the street). Simultaneously, small teams of plain-clothed officers would operate within crowds, monitoring for and intervening against activity likely to provoke (e.g., groups or individuals acting in openly hostile ways). If larger-scale problems emerged, the PSP would move to a level-two formation. This involved larger groups of police in standard uniform who would communicate with fans and negotiate potential solutions to problems. These groups could if necessary quickly escalate to level-three formations by donning protective equipment carried with them on their belts. Throughout, officers from the Corpo de Intervenção (the PSP’s “riot units”) would be stationed nearby but out of sight in order to provide a capability for high-level use of force if required. Perhaps most remarkable of all, however, the data suggest that the visibility of paramilitary-style police was virtually zero throughout the tournament and that there were no major incidents of “disorder” in match cities and only one England fan arrested for a violence-related offense, despite the presence of approximately 150,000 England fans during the tournament. Indeed, the Corpo de Intervenção for Northern Portugal did not record a single circumstance where they were required to draw a baton for the entire tournament (including for the high-risk fixture between the Netherlands and Germany). Thus, the evidence suggests that where a graded tactical approach is utilized there are fewer requirements to use police units, who by their very purpose and training are likely to employ tactics involving indiscriminate force against crowds. This research therefore demonstrates that during the tournament a form of graded tactical profile in line with the ESIM-derived theoretical models of police good practice was successfully implemented. This approach was highly proactive rather than laissez-faire, utilizing a wide range of nonconfrontational tactical options that allowed police to engage with fans at an early stage prior to any threat to public order. This meant that they were more capable of sup-

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 39

CROWD DYNAMICS AND PUBLIC ORDER POLICING

39

porting perceptions of police legitimacy and effectively differentiating between those within the crowd who were and who were not posing a public order risk. Data gathered from among England fans confirm that these tactics corresponded with widespread perceptions of police legitimacy—perceptions that were punctuated by critical moments of self-policing that helped in the prevention of any major incidents of disorder in match cities despite the presence of hooligans. In contrast, subjectively illegitimate policing, in areas outside of the PSP’s jurisdiction, was associated with the generation and escalation of conflict combined with the apparent empowerment of hooligan fans. Perhaps most striking, prior to the tournament the data suggest that strong identification as an England fan implied dissimilarity to police in match cities whereas following the tournament it implied similarity. The research therefore suggests that the experience of these policing tactics positively transformed a previously antagonist relationship between England fans and the police, perhaps over the longer term. This finding suggests that the route to successful public order outcomes may be via the creation of common bonds of social identification between crowd participants and the police that emerge during the crowd event itself as an outcome of the perceived legitimacy of police tactics (Stott et al. 2007, 2008). The widely recognized success of the PSP approach has also begun subsequently to impact upon policy at a European level. For example, in 2005, the Police Cooperation Working Party of the Council of the European Union adopted policies based upon this research as addendums to the European Union Handbook on International Police Cooperation and Measures to Prevent and Control Violence and Disturbances in Connection with Football Matches with an International Dimension.8 These relate specifically to the concept of “dynamic risk assessment” (8241/05) and “police tactical performance for public order management in connection with international football matches” (8243/05). This research and theory also form the scientific input to a proposed curriculum in a 4.7 million euro bid from the United Kingdom Football Policing Unit, along with forty partners from twenty-nine European states, to the European Commission to fund a Pan-European Police Training Programme for police officers involved in public order management at football matches with an international dimension.

From Football to Protest Crowds There is now substantial evidence that a social identity approach to understanding crowd psychology and behavior can be fruitful both theoretically and practically. The available evidence derived from the partnerships between science and police practice in the domain of football-related public order policing therefore starts to build a very convincing empirical case for (1) the ESIM

03_Madensen_Ch03.qxd:Madensen

40

1/3/11

10:24 AM

Page 40

PREVENTING CROWD VIOLENCE

analysis of crowd dynamics and psychology and (2) police good practice. Accordingly, it is evident from the wider body of research that risks to public order are not just a feature of the crowds or those within them but can emerge from the dynamics of the events themselves, dynamics of which the police can and do play an integral role. The available research and theory suggests that the effective and efficient policing of protest crowds involves the proactive management of the crowd dynamics of legitimacy and power. While there is no universal panacea, there are clearly underlying principles that can be of use for those developing and implementing strategic and tactical models at a local level. In order to achieve this, however, it is essential that the correct background theoretical understandings of crowds are in place among police. It would appear that the most effective means of managing crowd dynamics is by ensuring that ongoing threat assessments are linked as closely as possible to graded, dynamic, specifically targeted, information-led, and rapid tactical deployments. Within police organizational structures within the UK, it is evidently Bronze commanders that are best positioned to perform this function because they can adapt both the operational protocols and the tactical plan to meet the emergent risks. (In the UK, public order command teams are structured into three categories: Gold is the senior commander responsible for developing and setting strategy; Silver is responsible for determining the tactics for delivering the strategy; and Bronze are the forward commanders responsible for the operational delivery of tactics.) This would suggest that Bronze-level autonomy is an important feature in allowing police tactics to adapt efficiently and effectively to the dynamics of risk that emerge during an operation itself. The research suggests that it is therefore important to empower Bronze commanders. But at the same time it is essential to train them with the necessary strategic competencies and theoretical knowledge regarding crowd dynamics (Stott et al. 2008). This tactical approach should also reflect a strategy of facilitating fans’ legitimate intentions and the early use of nonconfrontational tactical options, such as communication and positive interpersonal interaction. Interacting with protesters during the event positively assists the police in risk assessment, the accurate targeting of their resources, and developing channels for liaison, and it actively constructs views of police legitimacy in the wider crowd. In this respect it is also evident that there would be some benefit in developing the “community policing” function, perhaps provided by special police units. It is not at this point clear who could provide such tactical capability, but it is certainly already a feature of the UK’s foreign police delegations traveling abroad. For example, during the Football World Cup Finals in Germany in 2006, English uniform police were used to create an easily identifiable liaison and dialogue capability with English fans. Moreover, in Frankfurt during the visit of England fans to the city, special units of German police were deployed

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 41

CROWD DYNAMICS AND PUBLIC ORDER POLICING

41

as “communication officers.” These police were observed successfully managing some very high-risk crowd incidents in coordinated interventions with Bereitschaftspolizei (riot police) (Stott and Pearson 2007, 238–297). Perhaps the strongest example to date of good practice with regard to the development of a dialogue capability has occurred in the Stockholm County Police Department in Sweden. As Stefan Holgersson and Johannes Knutsson (Chapter 10 in this volume) describe, there were a series of protests surrounding the European Union meeting in Gothenburg in June 2001. Around 25,000 people, including several “anarchist” organizations, took part in protests surrounding the event. While the majority of these protests were peaceful, some were also extremely violent and this disorder led to a sense of political crisis within Sweden, very similar to that experienced in the UK following the London G20 protests in 2009. A subsequent official report then led to the fundamental reform of public order policing in Sweden via the development of a common national tactical concept for policing crowds—the Special Police Tactic (SPT). The SPT is based around highly mobile units with protective equipment capable of working together in a manner very similar to the Police Support Unit in the UK. The Swedish National Police Board then began to implement the SPT via training at the National Police Academy. But it is important to emphasize that the analysis underpinning the strategy recognized the importance of being nonconfrontational and avoiding mass arrests via the differentiated use of force (Holgersson and Knutsson, this volume). It is also important to acknowledge that the training at the National Academy in Sweden is informed by and also formally linked to the ESIM via lectures and pre-training reading materials. A key issue then is that the ESIM and other research helps emphasize the manner in which interactions between demonstrators and police are critical in terms of governing the processes of escalation and de-escalation. But it is also relevant that this theoretical emphasis corresponds with an additional tactical capability that relies upon dialogue. Specifically, that in addition to the repressive tactics focused on the use of force, the SPT includes units of “dialogue police.” It is evident from the analysis provided by Holgersson and Knutsson (this volume) that the dialogue police play a particularly important function within the policing of crowd events that assists in managing their dynamics in a manner that offsets the emergence of “riot psychology.” Moreover, in line with the level-one formations of the PSP during Euro2004, the Swedish dialogue police work in pairs and normally in civilian clothes but are distinguishable by yellow bibs that display the words “Dialogue Police.” Their primary role is to act as a communication link between demonstrators and police commanders. However, their goal is to avoid confrontation through genuine dialogue, communication, identifying potential risks to public order, the facilitation of protesters’ legitimate intentions, and creating self-policing among the

03_Madensen_Ch03.qxd:Madensen

42

1/3/11

10:24 AM

Page 42

PREVENTING CROWD VIOLENCE

crowd (see Holgersson 2010 for a comprehensive historical overview and analysis of the dialogue police). What also seems relevant is that dialogue police were drawn in part from a background of being negotiators, and some had prior experience of negotiation from peacekeeping missions in the Balkans. They carry many of the principles with them into their role in public order management. In this respect, they work before, during, and after events to establish links to radical protest groups over extended periods such that during events they know and are known to key figures within such groups. In effect, the dialogue police adopt a “community policing” orientation to crowd participants. They are then able to build links of mediation and negotiation between police commanders and influential protesters during crowd events.9 Moreover, because they understand and have points of contact with the groups, they are able to assist both commanders and protesters by providing advice on and negotiating the potential impacts of different courses of action. Since their inception, no information from dialogue police has ever been used in the conviction of anyone arrested during a crowd event. It seems evident then that the effectiveness of their role is achieved partly through their explicitly “nonrepressive” function. What is also interesting is how (as Holgersson and Knutsson discuss) initially there was a great deal of hostility to the dialogue units, not just from protesters who had a history of antagonistic relationships to the police but also from police colleagues. This nonrepressive function along with not playing an “intelligence” role (e.g., gathering information that would be used to compile intelligence files or secure prosecutions) assisted the dialogue units in gaining trust with protesters. In order to maintain their links to protesters, however, they would be required to acknowledge that the indiscriminate and inappropriate use of force by police was itself a factor increasing the levels of hostility. Consequently, given the defining roles and responsibilities of the police, a common reaction from other officers was that the dialogue police had compromised themselves as police officers, taken the side of demonstrators, “gone soft,” and were in and of themselves a sign of police weakness. As such, the dialogue units were initially ostracized as “traitors” by colleagues. However, within a short period of time, police commanders began to recognize that dialogue had positive effects operationally in terms of reducing disorder, and they are now an established and highly effective component of the Stockholm Police Department. Consequently, the analysis provided by Holgersson and Knutsson begins to draw out the difficulties of managing what is essentially a complex array of intergroup relationships that occur during crowd events. Nonetheless, what seems clear from this is that the effective management of crowd dynamics is as much about mediating and managing the relationships between police and the crowd as it is about managing those within the crowd itself.

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 43

CROWD DYNAMICS AND PUBLIC ORDER POLICING

43

Conclusion It is evident from this overview of the research literature that there is a fundamental link between the psychology and policing of crowds. As Reicher (this volume) makes clear, this relationship should allow us to begin to forge a more optimistic understanding of the opportunities that crowds provide police for addressing their relationships with the communities that they serve. There is ample evidence to support this contention from the research we have been developing in the context of football matches with an international dimension. And this evidence begins to demonstrate how useful football crowds are as an arena within which to address more general issues. This research on football crowds also begins to point out that during crowd events police use of force should be informed by the immediate circumstances in which it takes place so as to avoid interventions directed at crowds or others who happen to be present in the vicinity of the incident. As the PSP in Portugal make clear in their strategy document for Euro2004, this use of force “must be targeted only at those individuals who have brought about an incident and who have adopted aggressive and violent attitudes.” In other words, and as one of the organizing principles of the ESIM approach, police use of force should not be exercised on the basis of a presumption that crowds pose a uniform threat to public order when only a small number within the crowd are judged to actually be posing a risk. In this respect, policing should be based upon the actual behavior and motivations of individuals within the crowd, but whole crowds should not then be subjected to the use of coercive force because groups within them seek to or have already transgressed limits of acceptable behavior as defined by the police. An important feature of this research is that it begins to develop upon the principles proposed by Reicher, Stott, Cronin, and Adang (2004) to suggest that facilitating the peaceful intentions of crowds, avoiding the disproportionate and indiscriminate use of force, and adopting a “graded” tactical approach are effective because they increase the range of tactical options open to the police prior to the deployment of indiscriminate coercive force. This increased range of tactics seems to allow policing to move from the reactive controlling of public disorder to the proactive management and maintenance of public order. What is critically important from a theoretical perspective is that the research evidence suggests that such proactive management is achieved through the maintenance of the perceived legitimacy of the intergroup relationships with, and thereby helping to create and maintain bonds of psychological identification between, police and crowd participants (Stott et al. 2008). In other words, police ability to construct and maintain identification between the majority of crowd participants and themselves is the psychological tool that can be utilized to assist in the maintenance of public order.

03_Madensen_Ch03.qxd:Madensen

44

1/3/11

10:24 AM

Page 44

PREVENTING CROWD VIOLENCE

Thus, while reactionary forms of policing are effective at restoring and asserting police authority through occupying physical space, such tactical approaches appear less capable of building the bonds with communities that are useful in maintaining public order over the longer term. In this sense, the evidence emerging from football crowds and elsewhere suggests that it is critically important that there are universal reforms of public order policing in democratic societies. To achieve this, it is necessary to make sure police commanders are more accurately informed concerning the scientific evidence and theory regarding crowd dynamics. It is also necessary to provide them with the tactical capability to negotiate and communicate with crowds so they are more able to create policing approaches within their immediate locus of command that avoid escalating the risk to public order. Where this approach is adopted, the science suggests that a situation is more likely to emerge in which (1) those within the crowd will perceive their intergroup relationships with the police as legitimate, (2) those seeking disorder are marginalized both physically and psychologically, (3) there is “self-policing” within the crowd, and (4) there will be overall reductions in the emergence, scale, and intensity of disorder. One of the central debates within the literature on public order policing is how best to achieve the avoidance of disproportionate use of police force within the policing of protest. Peter Waddington has consistently argued in favor of the use of paramilitary policing as a means of assisting in the maintenance of public order during protest events (P. A. J. Waddington 1987, 1991, 1993b, 1994). His contention is that the discipline, combined with the strong command and control, of paramilitary policing is effective at undermining tendencies that officers have to invoke disorder through uncoordinated and indiscriminate discretionary actions (e.g., such as officers “lashing out” at protesters). Crucially, he also argues that paramilitary interventions against the crowd must be underpinned by information, such that the police are able to accurately target the use of force. In contrast, Tony Jefferson (1987, 1990) argues that paramilitary policing carries inherent dangers of precipitating and exacerbating collective violence. He articulates a four-stage process where police expectations of disorder can lead inevitably to disproportionate and indiscriminate use of force by paramilitary police, a process through which widespread disorder emerges as a form of self-fulfilling prophecy (Stott and Drury 2000; Stott and Reicher 1998a). The literature from crowd psychology tends to be consistent with the positions of both Jefferson and Waddington. In particular, the data and analysis arising from Euro2004 point toward the important role played by squads of standard and nonuniformed officers who were operating within crowds gathered in match cities. These officers were able to gather real-time information on, and react quickly and sensitively to, any emerging sources of risk to public order. That study also shows that large squads of riot police were present but being kept deliberately out of sight, ready for rapid deployment. Thus, while the low-

03_Madensen_Ch03.qxd:Madensen

1/3/11

10:24 AM

Page 45

CROWD DYNAMICS AND PUBLIC ORDER POLICING

45

profile approach of the PSP may have avoided the obvious use of paramilitary police, it still retained the capability to utilize them. But it was the early and proactive deployment of nonparamilitary police that appears to have increased the capacity for the PSP to engage the type of information-led targeted use of force central to Waddington’s model. In other words, it was the deployment of nonparamilitary tactics, to locate and deal with emergent problems with dialogue or low-level force in the first instance, that meant that public order was maintained and that the paramilitary police—with their increased tendencies for the use of indiscriminate force—were hardly, if ever, required. In this sense, the central lesson from the scientific research is that there is a requirement to build upon the paramilitary tactics and formations that in the UK grew out of the major disturbances of the 1980s. The literature suggests that such development should focus upon how to unite the existing approach with a range of nonconfrontational tactical options that do not rely upon but still work in tandem with police capacity for the coercive use or threat of force. Perhaps the best model of this is reflected within the dialogue police pioneered by the Stockholm Police Department. And it should not be lost on us that Sweden is itself widely acknowledged to be one of the most democratic nations in the world. How such tactical capability could be achieved in other countries is less clear. In the UK, for example, it may well be that such tactics could be provided by Police Support Units themselves or by special squads that operate in similar formations to, but have fundamentally different roles from, what are currently referred to in the UK as Forward Intelligence Teams.10 But whatever tactical advances were chosen, they would then need to be supported through the development of public order training. It will not be sufficient for such training to simply report upon the scientific knowledge. As Reicher points out in Chapter 2, it will be necessary to integrate the implications of the science into all aspects of training at the level of strategy, tactics, and operational practice. It may also be useful to train officers involved in riot units to develop skills of conflict management via techniques of verbal and nonverbal communication, in a manner similar to training currently provided to some police negotiators. But whatever advances are made, it should be seen as selfevident that in partnership with science and education, a research capability should be developed such that any developments can be fully tested and the evidence fed directly back into police training and scientific knowledge as part of an ongoing strategy for the development of best practice.

Notes This chapter is based on a report submitted by the author to an inquiry in the UK by Her Majesty’s Inspectorate of the Constabulary (HMIC) into the policing of the London G20 protest in 2009.

03_Madensen_Ch03.qxd:Madensen

46

1/3/11

10:24 AM

Page 46

PREVENTING CROWD VIOLENCE

1. For an overview, see Reicher (2001a). 2. The ESIM is derived from the now dominant theoretical models of group psychology and intergroup relations within European social psychology (see Turner et al. 1987). 3. For a detailed theoretical and technical analysis of how these dynamics of power are understood to operate in relationship to the psychology of social influence, see Turner (2005) and Drury and Reicher (2005). 4. There are other important and influential theories of public disorder, most notable of which is the Flashpoints model of David Waddington. Although consistent with ESIM, these do not focus upon the social psychological dynamics of the events themselves, but rather the broader contextual determinants (see D. Waddington 2007a). 5. Observations were conducted at two Initial Public Order Commander (IPOC) courses, one Major Sporting Events (MSE) course, one Tactical Advisors course, and three public order Common Minimum Standard (CMS) courses. 6. Coordinating and Planning Committee for Euro2004 (PSP) (2003). Planning directive No/03. Level of police intervention to be adopted for Euro2004. PSP Coordinating and Planning Committee for Euro2004. National Directorate, Ministry of Internal Administration. 7. Study funded by Economic and Social Research Council (RES-000-23-0617) and the UK Home Office. 8. Police Cooperation Working Party of the Council of the European Union (2006), C32. 9. Although, it is important to recognize that influence processes are dependent upon circumstances. 10. A Police Support Unit (PSU) is a tactical formation developed specifically for dealing with public order situations, composed of three vehicles and twenty-five police officers with protective equipment.

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 47

4 Initiation and Escalation of Collective Violence: An Observational Study Otto M. J. Adang

EVERY SOCIETY HAS SPORADIC AND SOMETIMES VEHEMENT OUTbursts of collective violence, be it in the form of urban riots, soccer hooliganism, escalated protest events, or festivities gone out of hand. Over the years, analyses of what are considered to be excesses in “crowd behavior” have generated controversy. In 1972, the US sociologist Richard Berk noted that the bythen extensive social-scientific literature about the behavior of crowds and the behavior of people in crowds was mostly based on restricted information and unsubstantiated interpretations (Berk 1972). Over twenty years later, US sociologist and collective behavior researcher Clark McPhail (1997, 35) indicated that: “For more than a century the study of crowds was strangled by the methodological stereotype that ‘systematic research can’t be done.’” Similarly, in providing an excellent recent overview of the literature on football hooliganism, Steven Frosdick and Peter Marsh (2005) note how little of this literature was and is based on direct observations, let alone systematic observations, of football violence. As Frosdick and Marsh aptly note: “This lack of objective facts in theory and research on football hooliganism has bedevilled the debate since the 1960s” (2005, 31). And it’s not just the lack of objective facts. Little attention is being paid to the fact that collective violence occurs in an intergroup context. Echoing several other authors (e.g., P. A. J. Waddington 1987; McPhail, Schweingruber, and McCarthy 1998; della Porta and Reiter 1998b), Clifford Stott and Stephen Reicher (1998a, 510) indicate that: “When it comes to psychological explanations of crowd behaviour and—more especially—violence in crowd events, the focus tends to be almost exclusively upon one party: the crowd itself. Psychological studies of the police or army in such encounters are notable almost exclusively for their absence.” Stott and Reicher (1998a, 1998b) therefore urge the study of the intergroup dynamics of crowd-events and the inclusion of the role and perspective of the police. 47

04_Madensen_Ch04.qxd:Madensen

48

1/3/11

10:37 AM

Page 48

PREVENTING CROWD VIOLENCE

From a theoretical perspective, explanations for collective violence have traditionally focused on its supposed pathological, irrational, and apparently chaotic nature. According to various sociopsychological theories, people in a crowd lose their “awareness of self,” and the usual moral inhibitions and limitations are dropped. Following Scipio Sighele (1892) and Gustave LeBon (1895), it was—and still is—often assumed that there is no longer any kind of rational behavior in a crowd or mass of people, but on the contrary that there is a (primitive) tendency to do what others are doing. Suppressed desires surface in the behavior. A transformation is supposed to occur where people change and display other behavior purely as a consequence of the fact that they are part of a crowd. Philip Zimbardo’s so-called deindividuation theory (Zimbardo 1969) is often cited in this context. According to deindividuation theory, the excitement and the anonymity of being in a group lead to uninhibited behavior, and the normal limitations that people force upon themselves are lost. Due to the supposed psychological processes, individual identities are lost and individuals become exceptionally susceptible to suggestions and incitement by “leaders.” Characteristic of this view is that every crowd (a collection of people) can become a “mob” as a result of the actions of leader figures, the appearance of a hated person or hated object, acts of violence, or police action or the lack thereof. In many countries, these ideas have in the course of time become (and often still are) the basis for the education and training of police units deployed in (potential) riot situations (e.g., Schweingruber 2000). This was the situation when my research on the initiation and escalation of collective violence started in the 1980s. The main interest was not so much violent fans or protesters themselves or the root causes of collective violence, but rather to gain insight into the way in which violence in collective settings, such as football or protest events, starts and escalates. To study the initiation and escalation of collective violence, interactions around protest events and football matches were observed in a structured and systematic manner. The word interaction is important here, as the starting point for the research was that violence always involves at least an actor and a target and that the behavior cannot be understood without paying attention to the interaction between these. Apart from the focus on interaction, the study differed in a number of other respects from earlier studies because of its quantitative and comparative nature. Importantly, the study was not restricted to escalated riot-situations, but it included a number of comparable events that were considered to pose a risk to public order and where violence was seen as a real possibility, some of which led to disorder or turned into a riot whereas others did not. In total, 700 observation hours were recorded on audiotape (this is exclusive of time spent in pre-observation reconnaissance and information gathering and postevent data gathering). The first goal of the study was to provide a structured and contextualized description of violent interactions around protest and football events in the

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 49

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

49

Netherlands. The second aim was to analyze factors in the interaction and its immediate context that seemed to be associated with the initiation or escalation of collective violence, realizing that the factors associated with initiation of violence are not necessarily the same as the ones associated with escalation. So far, most of the results of this study have been published in Dutch only (Adang 1988, 1991, 1998). Here, I will be introducing the main results first, with data on protest and football being presented in a comparative way. As a next step, the initiation and escalation of collective violence is discussed, especially in light of how the results of the study fit in with existing theories of collective violence. Finally, the impact of police style and tactics on the initiation and escalation of collective violence is discussed, as well as the potential implications of these results for the management of public order.

Material and Methodology: Observing Collective Violence The investigation period lay between May 1986 and September 1989. Data gathered around 60 football events and 77 protest events in the Netherlands— some of which went off peacefully whereas others were characterized by extensive disorder—are presented. Additional efforts were made to observe events where there seemed to be an increased likelihood for violence to occur. The background of the methodology as well as definitions of the behavioral categories used are dealt with elsewhere (Adang, in preparation). In all cases, observations on locations that were in principle accessible to anyone (i.e., on public roads and in stadiums) were made by walking around with a photo bag over the shoulder and a portable cassette recorder in the jacket pocket, while listening into police channels via a scanner. The observations were recorded directly onto tape. Observational positions were chosen to obtain an overview of events (thus usually not in the middle of a group but at the edge or at a short distance). Systematic observations were conducted specifically during and around protest events where special police units (riot police, arrest units, horses, or dogs) were deployed and violence seemed likely based on past experiences or explicit calls for violence by protesters. As it turned out, 80 percent of all violent protest events reported about in the media (on national television or at least one of two national newspapers) during the research period were included in the observations. Additionally, protest events with at least 1,000 expected participants were included in the observations, irrespective of the likelihood of violence. As it turned out, 50 percent of all protest actions with at least 1,000 participants reported in the media (national television or at least one of two national newspapers) during the research period were, in fact, observed.

04_Madensen_Ch04.qxd:Madensen

50

1/3/11

10:37 AM

Page 50

PREVENTING CROWD VIOLENCE

There were large variations between the numbers of protesters per observation. The smallest protest event included a few dozen participants, whereas the largest demonstration had an estimated turnout of around 150,000 participants. The median number of protesters over all observations was 150 (it is meaningless to provide a mean, given the wide range with a few very large demonstrations). The observations typically started at the meeting point designated by the participants some time before the activity was planned to start. The observations lasted until after the event had ended and the dispersal of the participants had clearly started. Out of the 77 protest observations, mobile units (riot police) were deployed in 36, civilian arrest units were deployed in 21, mounted police were deployed in 29, and dog handlers were used in 8 cases. The observed football events were selected based on whether one or more of the five clubs perceived by police and media as being accompanied by socalled risk-fan groups were involved. Meetings between two risk clubs playing in the Dutch premier league (Eredivisie) occurred in 28 of the 60 football observations conducted over the course of three football seasons. The observations started when a group of away fans arrived in or nearby the city where the game was to be played (usually by special train) and ended when they left. On average, three times as many away fans (average 765, median 700) were present during meetings between two risk clubs than meetings between a risk club and a non-risk club (average 277, median 200). The away fans generally formed 4 to 5 percent of the total number of spectators. Matches between two risk clubs were attended, on average, by around 15,000 spectators (median 10,000, minimum 2,500, maximum 45,500), with other games averaging around 6,800 spectators (median 6,000, minimum 2,000, and maximum 13,500). Of the 60 football observations, there were 41 cases where mobile units were actually deployed, 27 cases where arresting units were present, 35 cases with mounted police, and 49 cases with dog handlers.

Results Violent Interactions

Violence by protesters was observed in 53 percent (n = 77) of protest events, while violence by fans was observed in 92 percent (n = 60) of football matches. In total, 428 violent initiatives by protesters and fans could be observed, 138 by protesters and 290 by fans (66 percent, away fans; 34 percent, home fans). For the protest observations, this amounted to 2 per observation and 0.7 per observation hour; for the football observations, there were

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 51

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

51

almost 5 violent initiatives per observation and 1.4 per observation hour. As far as the police are concerned, 130 violent initiatives by police were observed (i.e., not counting police use of violence in direct response to violence directed at the police), 51 around protest (i.e., 0.3 per observation hour) and 79 around football (i.e., 0.4 per observation hour) events. In addition, 485 instances of (nonviolent) provocative behavior (offending or threatening words/gestures; running charges; blocking a street or an entrance to a building/compound) were observed (protest, 168; football, 317). Both protesters and fans frequently engaged in insulting or threatening behavior. Protesters (but not fans) frequently blocked roads or entrances, fans (but not protesters) occasionally performed running charges. Coercive police measures (i.e., stopping, removing, or arresting) were more frequent: a total of 666 police measures were observed (442 around football, 224 around protest). The escorting of fans frequently involved (temporarily) stopping fans (e.g., after arrival by train), before entering the stadium (where they would be searched before being allowed to enter) or before leaving the stadium. During protest events that involved blockades, police would frequently act to remove protesters and break the blockade. Figure 4.1 represents the estimates of the number of people behaving violently during interactions. The average number of individuals from the same Figure 4.1

Number of Individuals Actively Performing Violent Behavior During Violent Interactions in Protest and Football Events Respectively

80

Percentage of violent interactions

70 Protest (N = 138) 60 Football (N = 262) 50 40 30 20 10 0 1 to 4

5 to 9

10 to 19

Number of violent individuals

20 to 49

04_Madensen_Ch04.qxd:Madensen

52

1/3/11

10:37 AM

Page 52

PREVENTING CROWD VIOLENCE

group behaving violently was estimated at three to four. In more than half of the cases, the maximum number of violent individuals in an interaction was less than five. During both protest events and football matches, instances where fifty or more individuals from a single group were actively violent during the same interaction comprised less than 1 percent of all violent interactions. On no occasion was more than 10 percent of individuals belonging to the same group (e.g., the away fans) actively violent during the same interaction. Relative to the total number of individuals present (e.g., all spectators in a football stadium, all participants in a demonstration) the number of violent individuals was always less than 1 percent of the total group. This does not mean that other group members were not involved in violent altercations. During both protest events and football matches, larger numbers of other group members lent vocal support to the violence (64 percent of protest events incidents; 47 percent of football incidents). The number of individuals performing these behaviors was generally five to ten times as high as the number of actively violent individuals: up to 80 percent of participants could be involved. On average, 20 protesters or 40 fans were involved in insulting or threatening behavior when violence was not occurring. A similar picture arises from participation in blockades by protesters (on average 25 protesters actively participated in blockades, n = 82), and participation in running charges by fans (on average 20 fans participated in running charges, n = 39). Explicit disapproval of violence by individuals forming part of the same group was rare: it happened in 9 percent of violent incidents during protest events and in 3 percent of violent incidents around football matches. Only a very limited number of individuals were involved in acts of disapproval or self-policing. Almost all protesters and fans performing violence were male with an estimated age between 15 and 25 (violent protesters > 90 percent males; fans > 99 percent males). For fans this is not surprising, as groups of fans traveling to away matches are almost exclusively composed of young males, but groups of protesters were generally more evenly composed of males and females and of higher average age (estimated average age around 30 to 35 years). Types of Violence and Targets

The most common form of violence observed was object-throwing, especially at other individuals and groups (Figure 4.2). Physical violence was less common, as was vandalizing of objects. In using violence, protesters sometimes used objects (such as paint-bombs or sticks) that they had brought to the scene themselves. Fans, who would know that they were to be searched, almost always used objects that were at hand. Figure 4.2 shows the percentage of violent interactions with either objectthrowing or physical violence against individuals or objects (when several forms were displayed during the same interaction, only the more severe form

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 53

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

Figure 4.2

53

Form of Violence Used During Violent Interactions in Protest and Football Events Respectively

80

Percentage of violent interactions

70 60

Protesters (N = 138)

50

Fans (N = 290)

40 30 20 10 0 Throwing objects at individuals

Throwing objects at objects

Physical violence

Demolition of objects

was counted, with physical violence being considered more severe than throwing, and violence directed at individuals considered more severe than violence directed at objects). The main targets (60 percent) for protesters were the police or policerelated objects such as police vehicles or police buildings (Figure 4.3). Other objects that were targeted frequently had some connection to the specific issue of the protest (e.g., government buildings or the embassy of the country to which the protest was directed were targeted in 20 percent of the cases) or to protests in general (e.g., banks and offices of multinationals were a target in 7 percent of the cases). On five occasions (4 percent), protesters directed violence toward third parties (e.g., passing car drivers or counterprotesters). On seven occasions (5 percent), protesters themselves were the targets of violent initiatives by third parties. On thirteen occasions (9 percent), violence was used against a target that was unrelated in any way to the protest (e.g., a traffic light). For football fans, “others” (i.e., rival fan groups) were the main target (45 percent). Police were also a frequent target (25 percent), as was the playing field (players, referee, linesmen: 20 percent) or objects (fences, buses, trains, etc.: 10 percent).

04_Madensen_Ch04.qxd:Madensen

54

1/3/11

10:37 AM

Page 54

PREVENTING CROWD VIOLENCE

Figure 4.3

Targets of Violence During Violent Interactions in Protest and Football Events Respectively

70 Protesters (N = 138)

60 Percentage of violent interactions

Fans (N = 290) 50

40

30

20

10

0 Police

Playing field

Others

Objects

Figure 4.3 shows the percentage of violent interactions with violence aimed at police, other groups or individuals, objects, or the playing field (during football events). Contexts in Which Violence Occurred

In almost all cases (protest events, 96 percent; football, 95 percent), the events immediately preceding violent initiatives were observed as well. In around half of the cases, violent initiatives followed one of previously defined immediate contexts (i.e., occurring within two minutes of that context; see Figure 4.4). For protesters, the most common immediate context was measures being taken by police (32 percent of all violent initiatives). In 10 percent of the cases, a violent initiative immediately followed a previously occurring instance of violence, for example, when protesters first attacked the police and then a building (agonistic context). Competition with or provocations by others (e.g., individuals trying to pass a blockade, drivers trying to drive through a demonstration) were less common contexts for violence. For football fans, 14 percent of violent initiatives were preceded by events on the playing field (e.g., a goal

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 55

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

Figure 4.4

55

Immediate Contexts of Violent Interactions

60

Protesters (N = 138) Percentage of violent interactions

50

Fans (N = 290)

40

30

20

10

ts en

ex at M

N

o

ch

co

ev

nt

se ot N

ya no An

t

en

e nc

io tit pe om C

Po

lic

e

m

Ag

ea

on

su

is

re

tic

s

n

0

being scored, a disputed referee decision), 12 percent occurred in an agonistic context, 11 percent were preceded by police measures, and 9 percent followed provocations by other fans. It became clear that the immediate context of violent initiatives was different for different targets: violence by fans in the direction of the pitch was relatively strongly linked to match events, while violent initiatives from fans or protesters against the police appeared relatively often in connection with coercive police measures. During both protest events and football matches, more than half of the violence between fans/protesters and police occurred in the context of coercive measures applied by the police (56 percent of the violence directed at the police and 71 percent of the violence performed by the police). In addition to this, 25 percent of the violence by fans against objects (but only 5 percent of the violence by protesters against objects) occurred in a context of coercive police measures. Both for protesters and fans, however, no clear immediate context could be discerned in about half of the cases in which the situation immediately preceding a violent initiative was observed. In these cases, the violence did not seem to arise in response to a specific triggering event. Of course, it could be argued that this was due to observational constraints and that more triggering events would have been identified had observations been more accurate. However,

04_Madensen_Ch04.qxd:Madensen

56

1/3/11

10:37 AM

Page 56

PREVENTING CROWD VIOLENCE

based on observations with a continuously good view of all participants involved and based on repeated viewing of incidents caught on video, sufficient evidence exists to conclude that violence did not occur in direct response to an identifiable event or action from the target in many instances. Figure 4.4 shows the percentage of violent interactions starting within two minutes after one of several predefined immediate contexts. “Not seen” indicates instances where the period preceding the violent initiative was not observed in its entirety. “No context” indicates instances where the period preceding the violent initiative was observed in its entirety and none of the predefined immediate contexts was deemed to have occurred. For protest events, there was a strong association between the presence of individuals wearing balaclavas (as determined at the start of the observation) and the likelihood and frequency of violence. When individuals wearing balaclavas were present (n = 35), violence occurred in 80 percent of the observations, compared to 33 percent when they were not present (n = 42). The difference with regard to the frequency of violence was even more pronounced: 1.6 vs. 0.45 violent initiatives per observation hour. There seemed to exist a strong antagonistic relationship between groups of individuals wearing balaclavas (often referring to themselves as “autonomen” or “black bloc”) and the police. In relation to football, the observations suggested that the relationship between different fan groups is an important factor and that the mere presence of rival fans was often sufficient to elicit violence. An analysis of official police data gathered by the Dutch National Football Information Point (CIV) from the season 1986–1987 made clear that the likelihood of an incident was increased sixfold (47 percent vs. 8 percent) if a discernible group of away fans was present compared to when this was not the case (the CIV data contain an incident for 87 of 140 matches with a discernible away group and for 14 out of 165 matches without away fans). When groups of rival fans met, the likelihood of incidents was particularly high (the CIV data contain incidents for 89 percent of fifteen encounters around matches between what they designate as “risk teams”). As far as the observations are concerned, away fans were present during all observations except four. In three of these cases, no violence was observed. The frequency of violence differed markedly based on whether or not two risk teams encountered one another (2.5 per hour vs. 1.2 per hour, Mann-Whitney U test, p < .001). The analysis of immediate contexts reveals what potential triggers preceded violence. However, it is important to note that not all potential triggers were followed by violence. The frequency with which potential triggers were followed by a violent initiative is discussed below. Other conflicts as a potential trigger. Of all 290 violent initiatives by fans,

13 percent were followed by another violent initiative (counterviolence not

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 57

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

57

being considered as such). There were different ways in which this could occur: • Fans that used violence chose a second target (12 cases); this second target was mainly the police. • Fans reacted to (and against) other fan groups that used violence against the police, the pitch, or objects (5 cases). • Fans that experienced violence from another fan group reacted with violence against the police or against objects (2 cases). • Fans acted violently against objects first and then against other fans or the police (4 cases). Of all 138 violent initiatives by protesters, 5 percent were followed by another violent initiative (as when protesters were first violent against police and then against another target, or the other way around). Pitch events as a potential trigger. Disputed referee decisions, violence on

the pitch, and goals being scored were included as potential triggers for violence by football fans. The event on the field that was followed by violence most frequently was the score of a goal point by one of the teams. The 32 goal points that formed the trigger for violence constituted 19 percent of all scored goal points (total 168 goals). Goal points where one of the teams reached an advantage over the other side with one or two goal points were most frequently followed by violence; more so than when a lead was being reduced, an equalizer was being scored, or a lead was being raised still further (chi-square test, p < 0.05). Fans from the scoring team and those from the other team were equally likely to initiate violence (44 percent against 56 percent, Chi-square test, not significant). In most cases (20), the violence was directed at the rival fan group; in 9 cases the violence was directed at (individuals on) the pitch. After the game there was, for both away and home fans, absolutely no connection between violence and the result of the match. Both home and away fans were not more violent after their club lost compared to when it had won. Of the observed 72 instances of violence on the pitch, 7 of those (10 percent) were followed by violence in the stands. Of the observed 32 controversial referee rulings, 4 were followed by violence. In these instances, the violence was directed in 55 percent of cases toward the pitch and in the rest of the cases to the other fan group. There were no indications that the initiation of violence was linked to other match events, such as prolonged interruptions (e.g., because of an injury treatment). Provocations as a potential trigger. During the football observations, a total of 217 provocations were observed. Only a small part of these were followed by fan violence (less than 5 percent). Mostly, there was no reaction to provocations.

04_Madensen_Ch04.qxd:Madensen

58

1/3/11

10:37 AM

Page 58

PREVENTING CROWD VIOLENCE

If there was a response, it was usually in a similar vein. During protest events, provocations by third parties were rare (n = 15, 2 of these were followed by violence). Competition as a potential trigger. Competition for space was a frequently occurring potential trigger during protest events (n = 305). This obviously occurred when protesters blocked a road or an entrance, but the mere presence of a demonstration on a public road could create disagreements with car drivers. This only led to violence on occasions in which police had not taken any regulatory measures (in the case of a blockade) or when car drivers tried to force their way through a demonstration (7 occasions). Police measures as a potential trigger. An important finding was that if

away fans were stopped by police on arrival at the station there was a significant increase in fan violence compared to when they were not stopped (46 percent vs. 0 percent, chi-square test, p < 0.001). There was no noticeable effect between stopping or not stopping fans during searches at the entrance of the stadium. After the game, there was a clear difference: there was more often violence when away fans were held back compared to when they were permitted to leave the stands immediately following the game (45 percent vs. 5 percent, chi-square test, p < 0.001). Holding fans in the stands was supposed to prevent violence outside the stadium by keeping away and home fans more effectively separated from each other. However, detaining away fans for a short period did not appear to decrease the frequency of violence during the phases that followed. By far, the most frequent coercive police measure used during protest events was the (re)moving of protesters. During protest events, the use of coercive police measures in response to blockades was often associated with violence by protesters (19 percent, n = 359). Reactions to Violence

Violence by protesters or fans did not always elicit a police response. During protest events, the police responded within two minutes in some way to 45 percent of violence toward the police (n = 83) and 7 percent of the violence directed toward other targets (n = 55). For 22 percent of violence against other targets (n = 55, often inanimate objects such as buildings), a response could be noted from the target or individuals related to the target (e.g., people inside a building). The police responded within two minutes to 49 percent of violence by fans toward the police (n = 65) and 35 percent of the violence directed toward other targets (n = 225). Compared with police, fans responded more often (chi-square test, p < .05) to violence directed at them by other fans: 68 percent of the cases (n =

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 59

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

59

128). Fan violence led to counterviolence in 24 percent of the cases if the police were targeted and 36 percent of the cases when other fans were the target. Fan reaction to violence could also be limited to nonviolent behaviors, such as vocalizations (abusive calling, threats, laughter). It is worth noting that flight and evasive reactions were rare, as was the absence of positive (conciliatory) reactions. Other reactions that amounted to “ignoring” the violence (e.g., “just looking”) frequently occurred. The manner in which fans or the police reacted to violence appeared to depend upon the number of individuals that committed violence: if there were more violent fans, the likelihood of a violent reaction was greater (chi-square test, p < 0.05). Less than half of all violent initiatives taken by protesters or fans were followed—within two minutes—by other violent acts from the same group (protest events, 43 percent; football, 41 percent). Fan violence tended to last longer if the target reacted with violence. When violence continued, interactions lasted an average of three minutes (protest events, 161 seconds, n = 59; football, 187 seconds, n = 119). Interestingly, this average was the same irrespective of the reaction of the target or the police response (if any). Police interventions seemed to have more of an effect on the likelihood of subsequent initiatives to occur, as analysis of between-interaction intervals made clear (see below).

Police Measures Police were visibly present during all observations. In most cases, the police were a very short distance away from both protesters and fans (within talking/hitting distance). Even if the police presence was less conspicuous (as would often be the case inside the stadium), they were available (and known to be so) to respond to any situation that might arise. Nevertheless, there was a clear trend showing that the greater the distance between police and away fans, the greater the probability of fan violence. This difference was especially marked at the station after arrival of the train carrying away fans (chi-square test, p < 0.03) and inside the stadium, both before the game and during the halftime break (chi-square tests, p < 0.01). As indicated earlier, frequently during protest events, the targeted objects had some connection to the specific issue of the protest (so-called risk objects). These objects were sometimes “guarded” by a police presence. When they were not guarded, they were more likely to be targeted (54 percent of twenty-six risk objects identified beforehand that were not guarded were attacked, as opposed to 15 percent of sixty-eight that were; chi-square test, p < .001). The same trend was visible on occasions with potential for confrontation with “third parties”: if police were not close by when traffic tried to pass a demonstration or people passed a picket line, violence by protesters was more likely (chi-square test, p < .05).

04_Madensen_Ch04.qxd:Madensen

60

1/3/11

10:37 AM

Page 60

PREVENTING CROWD VIOLENCE

Both protesters and fans were frequently subjected to coercive measures by police. Away fans were often stopped by the police (1) after arrival at the railway station, (2) after arrival at the stadium, and (3) before departure from the stadium. Stopping of fans occurred frequently in an effort to keep the group together or hold it effectively under control (e.g., while entering or leaving buses). Fans were searched and their tickets were checked before entering the stadium. In a number of cases, fan groups were stopped to keep them separate from other fans. Removing fans often served another aim, namely to put an end to violence (or clearly impending violence). In these cases, the measures were a response to events. On other occasions, fans were removed without any preceding violence in an attempt to keep different fan groups separate or to regulate the behavior of the fans (e.g, while exiting or boarding the train). Making an arrest (sixty-four arrests were observed around football matches) was, in principle, a reactive form of police action, but only 44 percent of fan arrests directly followed the actual offense. In the rest of the cases, the arrest was delayed and made at a moment considered more suitable. In all cases, the aim of an arrest was to serve judicial purposes: upholding of the law and confrontation of individuals with the judicial consequences of their actions. It is not known if this actually happened in all observed cases. In addition, the arrests immediately following the offense could also be aimed at stopping undesirable behaviors. (Re)moving protesters served to get them away from risk objects or to break blockades. Moving was the most frequently observed police measure, especially when protesters blocked streets or entrances to buildings. The monthly blockades held at the military base in Woensdrecht, Netherlands (where US cruise missiles with nuclear warheads were planned to be stationed) led to frequent (temporary) moving of protesters to allow traffic to pass. On a number of occasions (groups of) protesters were arrested as well. The stopping of protesters was used to keep them away from risk objects or blockades and was most often used proactively (in 87 percent of the cases), whereas removing protesters was usually a response to behavior considered undesirable by police (83 percent of the cases). On a few occasions, protesters were moved as a means to stop violence. During protests, all but two observed arrests (a total of sixty-three arrests with 335 arrestees were observed during protests) directly followed the alleged offense. Thirteen of these arrests were in connection with violent behavior by protesters, while all others were made in connection with less serious offenses (usually related to blocking a public highway). Again, the aim of arrests was to serve judicial purposes: upholding of the law and confrontation of individuals with the judicial consequences of their actions. Arrests immediately following the offense could also be aimed at stopping undesirable behaviors, and the groupwide arrests made during the monthly blockades in Woensdrecht mainly served that purpose (personal communication by commanding officer).

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 61

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

61

Riot police in full gear were responsible for most of the violence used by the police, both against fans and protesters.1 Of course, these units were deployed specifically in situations considered to constitute increased risk. This was less true of dog handlers, who also used violence relatively often, whereas mounted police in comparable situations used less violence (chi-square test, p < 0.001). However, these differences in police violence could not be explained by the severity of the violence or the amount of troublesome behavior that was directed at the police by fans (chi-square tests, not significant). The police reacted in some form (within two minutes) to 38 percent of all violent initiatives by fans and 50 percent of violent initiatives by protesters. This police interference did not systematically lead to discontinuing of violent initiatives: the percentage of violent initiatives that were continued and the duration of the violence did not vary in relation to whether or not the police had acted. To determine if police interference had a long-term effect, the intervals between violent fan interactions were examined using survival analysis.2 It appeared that police interference significantly decreased the chance that violence would start anew: both for protest and football events, it took, on average, longer before a new violent initiative occurred if the police reacted (no matter in what manner) compared to situations where there was no police reaction. In total, seventy-nine violent police initiatives were observed (n.b., these do not include police reactions to fan violence directed at the police itself). Police initiatives to use violence or force appeared most often as part of the use of coercive measures (71 percent of the cases). In almost all other cases, violent police initiatives were a reaction to troublesome or violent behavior of fans (not directed at the police themselves). In thirteen cases, the police were seen to use violence or force in reaction to fan violence directed at the police (and it was thus not a violent initiative by police). In six cases, the use of violence by the police seemed incomprehensible in the sense that it was unclear what aim was served by the violence.

Discussion: The Initiation and Escalation of Collective Violence The first conclusion to be drawn from the study is that it is possible to conduct meaningful structured observations of episodes of collective violence. A limitation of the methodology is, of course, that structural causes or the psychology of collective violence are not being addressed. The current methodology concentrates on the short-term processes that are associated with violent behavior in collective situations. Even though protest events and football matches provide very different situations for collective behavior, the results clearly indicate some important similarities between them. Both in relation to football and protest events, even in highly escalated incidents of collective

04_Madensen_Ch04.qxd:Madensen

62

1/3/11

10:37 AM

Page 62

PREVENTING CROWD VIOLENCE

violence, the relative number of people actually committing acts of violence is low, targets of violence do not seem to be randomly chosen, and in approximately half of violent incidents there was no recognizable potential “trigger” for the initiation of violence in the immediate context. The observations lead to the following conclusions with regard to the behavior of people in collective settings: • The agency in collective violence is clearly at the individual level: people make individual choices and do not behave more or less uniformly (the “illusion of unanimity” mentioned by McPhail [1991]). This does not detract from the fact that individuals are fundamentally social beings and that their decisions are influenced by their social environment. Both for protest events and for football matches, the relevance of the immediate context and the interactive action–reaction nature of these processes is manifest. • There is no reason to assume that, within collective settings, the mere presence in a crowd makes people more likely to use violence; the vast majority of the people in the observed instances of collective violence (which included some severe riots that made headlines in the media) were not violent at all. The literature indicates that there are reasons to believe that many of those actively seeking out violence in collective settings are more likely to use violence in other circumstances as well (e.g., van den Brug 1986; van de Valk and Linckens 1988), and of course, in most collective situations there is no violence at all (this study focused specifically on situations where violence was considered to be more likely). • There is no reason to assume that, within collective settings, people show a higher tendency to perform so-called emotional or irrational behavior. On the contrary, even in violent collective settings, people clearly made choices and behaved in ways that were meaningful to themselves. The choices individuals made seemed to be consistent with a model of (bounded) rationality (cf. McPhail 1991): they prepared themselves, pursued goals, and clearly took risks into account. There is every reason to believe that the behavior of people in collective settings is influenced by the same factors that influence their “normal” everyday behavior. Rather than behaving without norms, the data indicate that even in violent situations, certain norms seem to be taken into account, as indicated by restrained behavior (all-out violence is very rare), the choice of “appropriate” targets, and examples of self-policing. In addition, both violent protesters and violent fans adhered to certain rules, the most important being to stick together and support one another during confrontations. In this respect, the findings are consistent with Turner and Killian (1987) and Reicher (1984, 1996) who convincingly show

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 63

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

63

that (perceived) norms become more rather than less important in collective situations, although the nature of the norms may vary. Specifically with regard to the initiation and escalation of collective violence, the results of this study indicate that as far as the initiation of collective violence is concerned, a distinction should be made between two types of violence: 1. Violence that is linked to a clearly identifiable trigger. This type of violence is reactive—it is a response to specific elements or frictions in a situation, be it provocations by other fans or third parties, events on the pitch (in the case of football), measures taken by police, or some other identifiable trigger. Theoretically, this type of violence is easily linked to familiar aggression theories (e.g., aggression out of frustration), competition for limited resources, or as a response to threats. As with other forms of aggression, males are more likely to react aggressively than females, and adolescents or young adults (individuals between the ages of 15 and 25) are more likely to react aggressively than individuals from other age groups. Having said that, males from other age groups and females may react aggressively to certain triggers on occasion. The targets of the violence may vary, but they are usually linked with the trigger preceding the behavior (except in cases of redirected aggression). 2. Violence that is not linked to a clearly identifiable trigger. This type of violence is not reactive, but seems to arise more spontaneously. It is performed almost exclusively by groups of male adolescents or young adult males and is directed specifically at similar, rival groups of young males. The individuals and groups concerned seem to actively seek out opportunities to confront rival groups. Theoretically, this type of violence can be seen as another expression of the so-called young male syndrome (Wilson and Daly 1985), the tendency of young males to take risks and be violent because they discount the future in favor of short-term gains. The evidence suggests that this tendency is primarily a masculine attribute and is socially facilitated by the presence of peers in pursuit of the same goals. Violent male-male disputes are really concerned with “face,” dominance status, and what Erving Goffman (1959) calls “presentation of self in a highly competitive social milieu.” The involvement of especially young males in episodes of collective violence is well documented in the literature. The distinction between the two types of violence is not absolute and an obvious overlap is created by the fact that the young male syndrome may also be expressed in response to triggers that may seem trivial to outsiders.

04_Madensen_Ch04.qxd:Madensen

64

1/3/11

10:37 AM

Page 64

PREVENTING CROWD VIOLENCE

Different mechanisms are responsible for the escalation of violence however: 1. On the one hand, there is the (perceived) risk of retaliation. From the observational data, it is quite clear that only a small minority of a group engages in the most risky types of behavior, while the majority of participants opt for less risky alternatives (shouting, gesturing, running) or do not become involved at all. Even for those being violent, there is a lot more missile throwing than physical fighting, and redirected aggression at inanimate objects (fences, buses, trains) rather than at individuals that can fight back. There is something contradictory here, as the young male syndrome is characterized by risk-taking behavior and engaging in violence involves just that. However, within that framework, individuals seem intent on avoiding “unnecessary” risks and on reducing risks. The fact that the young males, when violent, operate in groups is a form of risk-reduction in itself, as is the fact that they avoid or flee from confrontations that they seem unable to win, as was observed on several occasions. In several respects, the data show that violence became more likely when there were no police present at risk locations. In addition, police are more often avoided than confronted and—especially for fans—most confrontations with police occurred only after police had taken some kind of coercive measure. Violent fans and protesters regularly took measures to hide their faces to make recognition more difficult and avoid identification and arrest. In the last part of the study period, closed-circuit television cameras were introduced into football stadiums. It was clear that violent fans disliked these cameras intensely and preferred to remain anonymous to authorities that could punish them for their behavior. These risk-reducing attempts to maintain “anonymity” (to authorities, not to their fellow fans!) is to be distinguished from the so-called deindividuation effect of anonymity, for which there is no support (Postmes and Spears 1998).3 The evidence for bounded rationality, in combination with the relevance of opportunities to be violent with limited risk for retaliation, provides a link between collective violence and principles of situational crime prevention (Clarke 1995). 2. The second important escalation mechanism is the existence of an “us versus them” antagonism. The more antagonistic the relationships between different groups, the higher the frequency of observed violence. This was clearly the case for the relations between rival fan groups and for the relationship between certain groups of protesters (autonomen/black bloc) and police. Stott and Reicher (1998a, 510) claim other studies show that “crowd conflict characteristically arises

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 65

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

65

when official agencies such as the police or the army intervene against unofficial mass action.” Although the results of this study indicate that collective violence was not characteristically triggered by police action, collectively applied coercive police measures were often associated with an escalation of violence. The collective nature of these measures contributes to or exacerbates (or may even create) an “us versus them” perspective that can lead to more explicit in-group/ out-group behavior and to more individuals behaving violently. Theoretically, the Elaborated Social Identity Model, which states that collective “disorder” is made possible through the shared psychological salience of a common social identity among crowd participants is relevant here (Reicher 1984, 1996). The defining dimensions of this identity serve to explain the normative limits of collective action (what people do) and the extent of participation (who does and does not join in) during a crowd event. This social identity analysis argues that the dynamics of intergroup interaction are integral to the psychology of widespread disorder. Stott and Reicher (e.g., 1998a) indicate that when an initially heterogeneous crowd has come to be treated as a homogeneous whole by the police, this has led crowd members to reconceptualize themselves as members of a common category, thus setting up a cycle of tension and escalating conflict. There is a theoretical debate going on between different explanations for collective violence (Reicher 1996; Stott and Reicher 1998b), with one side arguing for the importance of predispositions. This perspective suggests that collective violence is an outcome of the convergence of individuals who are predisposed toward creating disorder (e.g., hooligans). This approach does not explain how and why collective violence erupts in specific circumstances but not in others. The other side argues the need for a contextualized, groupdynamic understanding of collective violence. The results of this study with regard to the relevance of context, intergroup interaction, and intergroup relationships support this last approach. At the same time, it is clear that there are variations in the willingness of individuals to become involved in violence, with some actively seeking opportunities to be violent, without the need for external triggers (other than the presence of a rival group). The initiation/escalation model of collective violence presented in this chapter could be seen as a first step toward combining these different theoretical approaches.

Public Order Management The findings of the study have clear implications for the management of public order. To prevent collective violence, wherever possible and feasible, frictions

04_Madensen_Ch04.qxd:Madensen

66

1/3/11

10:37 AM

Page 66

PREVENTING CROWD VIOLENCE

should be avoided. This requires facilitating the legitimate activities and intentions of participants as far as possible and identifying potential areas of friction. In addition, monitoring and observing events in real time serves to note instances where frictions start to arise. Measures should be taken to prevent frictions or to deal with their consequences from an early stage onward. This involves communicating with participants and informing them to learn what is affecting them, to avoid misunderstandings about measures being taken, and to gain compliance. With regard to groups of young males looking for confrontations, it will be relevant to identify and get to know them, to be able to influence them from an early stage onward, making clear to them what limits will be set to their behavior. This also involves communicating with them directly. For their risk perception, it should be clear to them (and others) what effective measures will be taken if they transgress these limits. By getting to know them, their anonymity to authorities will be reduced. The management of public order should avoid as much as possible taking measures that create or emphasize an us-versus-them situation and make in-group/out-group behavior more likely: interacting and communicating are important tactical tools in this respect. When violence does occur, it is vitally important to act in a timely fashion rather than waiting for situations to escalate and get out of hand and to do this in a focused and targeted way, aimed specifically at those individuals transgressing limits. These strategic and tactical principles fit very well with the four principles for crowd policing as formulated by Reicher et al. (2004, 2007). So far, the results of the study have inspired several practical applications. On the occasion of the Euro summit organized during the Dutch European Union presidency in 1997, the Amsterdam police formed what they called a “peace unit”—a unit composed of some forty uniformed officers who were specifically trained to engage protesters by communicating with them. This unit, which still exists, is frequently used in all types of situations of tension where it makes an important contribution to preventing disorder. Although in line with the findings of the study, it cannot be claimed that the formation of this “dialogue unit” is a direct result of the study, especially as it was in line with previous experiences and experiments in the Netherlands (e.g., see IJzerman 1982). A more direct line exists between the results of the study and police preparations for the European football championships Euro2000 held in the Netherlands and Belgium. On the basis of the findings of the study and the practical expertise of the officers involved in the binational police project preparing for Euro2000, a “Police Behavioural Profile Euro2000” was drawn up and adopted by authorities to bring about an equal influence on the behavior of visitors in the different host cities. At the core of this profile was the idea that a friendly but firm low-profile approach would be most suitable. The results of the research into the implementation of the behavioral profile in all host cities demonstrated the effectiveness of the friendly and firm low-profile approach

04_Madensen_Ch04.qxd:Madensen

1/3/11

10:37 AM

Page 67

INITIATION AND ESCALATION OF COLLECTIVE VIOLENCE

67

(Adang 2001; Adang and Cuvelier 2001a, 2001b). Characteristic of the lowprofile approach were a substantial, but limited number of police officers in daily uniform, patrolling in pairs or small groups, interacting with fans. These interactions were friendly, but transgressions by fans were responded to quickly. In the low-profile approach, police deployment was based more on intelligence and on information provided by spotters’ teams. For the contrasting “high-profile” approach, more than three times as many officers were visible in the streets. These officers were more often dressed in riot gear and accompanied by their riot vehicles; they formed larger groups, which made it less easy to approach them. It is important to note that the distinction made here is much more subtle than the distinction between paramilitary and nonparamilitary styles of public order maintenance (Jefferson 1987, 1990; P. A. J. Waddington 1987, 1993b). In both high-profile and low-profile approaches, a clear command and control structure was in place and use could be made of intervention units with riot gear. To varying degrees, these experiences were later used during Euro2004 in Portugal (Stott et al. 2007, 2008), the World Cup 2006 held in Germany (Schreiber and Adang 2010) and Euro2008 held in Austria and Switzerland (Adang and Brown 2008). The study was relevant in other practical applications as well. The city of Arnhem in the Netherlands was one of the Euro2000 host cities. For the tournament, police in Arnhem had paid specific attention to the quality of information gathering and to information analysis. Following the tournament, starting in 2002, police in Arnhem took initiatives to improve their information management and intelligence in relation to football violence. Using knowledge about the young male syndrome and the fact that violent football offenders are more likely to be violent on other occasions as well, they gathered and analyzed targeted information to be able to better identify individuals and groups involved in instigating violent incidents. However, the approach was not just a traditional “hooligan” approach aimed at repression. From the start, the aim was to make possible a more tailor-made approach to the policing of fans, both at an individual and at a group level. In this way, police in Arnhem were less often surprised by sudden incidents, were able to deploy fewer police (and especially fewer riot police) around football matches, and were able to make more “better” arrests (in the sense that arrests more often led to successful prosecution). At the individual level, in cooperation with club and youth workers, specific approaches were developed to encourage individuals to refrain from involvement in violence (Ferwerda and Adang 2005). Applied in pilot form in three other forces, the methodology has subsequently been implemented in all Dutch police forces. Another offshoot of the study was the development of peer review evaluation teams (Adang 2006; Adang and Brown 2008), which are currently being deployed on request in several countries across Europe. The idea of a peer review evaluation team is to involve, under scientific guidance, experienced

04_Madensen_Ch04.qxd:Madensen

68

1/3/11

10:37 AM

Page 68

PREVENTING CROWD VIOLENCE

police officers from several countries to observe police operations in real time and provide feedback. Not to blame, but to exchange and to learn, creating a win-win situation in which both requesting police forces and participating reviewers gain something, and more general lessons for public order management can be drawn as well. An explicit goal of the peer review evaluation teams is to increase the application of theory into practice and the continued development of theory based on practical experiences and field work.

Notes The research was carried out with a grant from the Dutch Interior and Justice Ministries while the author was a postdoctoral researcher at the University of Utrecht. 1. Use of force by police is labeled as “police violence,” as it conforms to the definition of violence as used for the behavior of protesters and fans. 2. With this analysis, the fact that not all the cases of violence were followed by new violence (the so-called censors) was taken into account. 3. According to deindividuation theories, anonymity causes antinormative and disinhibited behavior.

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 69

5 Police Use of Active and Passive Mitigation Strategies at Crowd Events Ingrid Hylander and Kjell Granström

The attitude we got—this experience of “us” and “them”—automatically appeared when the police attacked all the people there, all of us as one group. Well, we united in a way, even if we actually didn’t have so much to agree on. —Young protester in Gothenburg, 2001 If you took part in one of the large demonstrations, you belonged to “the good protesters” but if you took part in a smaller event, like “Reclaim the Streets,” you suddenly belonged to the hooligans. —Young protester in Gothenburg, 2001 I’m not one for violence, but I figure if I were about 20 I’d probably have gone along with it. I don’t think those people were hired hooligans egging each other on. I just think this was a normal, healthy reaction from the kids. . . . I’d have thrown my handbag because people would’ve thrown anything, because people are entitled to defend themselves when they’re attacked without doing anything. . . . I don’t know how you pry up paving stones, so I wouldn’t have thrown them. —Middle-aged demonstrator interviewed after taking part in a demonstration in Gothenburg, 2001

THESE THREE EXAMPLES FROM RIOTS IN GOTHENBURG IN CONnection with the European Union summit meeting in 2001 show how protesters were affected by the course of events. The first two examples are from a young protester who took part in a reclaiming party, where paving stones were thrown at the police and the police answered by attacking the entire group of participants. In the third example, the police used vehicles, dogs, and horses to cut off the part of the demonstration—which they categorized as the “black bloc”—that had not been announced to the authorities in advance. This 69

05_Madensen_Ch05.qxd:Madensen

70

1/3/11

10:38 AM

Page 70

PREVENTING CROWD VIOLENCE

resulted in heavy stone-throwing from the blocked-off protesters. Participants, who before the demonstration never would have thought of hurting the police, afterward defended the stone throwers. After these incidents in Gothenburg, the Swedish public was shocked. The police proclaimed, “This must never happen again!” A major change in Swedish police tactics in crowds was gradually enforced. This new tactic is described by Stefan Holgersson and Johannes Knutsson in Chapter 10 of this book but will be analyzed here in light of a model elucidating aggravating and mitigating processes in crowd events (the AM model) (Granström 2008). The new Swedish Special Police Tactic for crowd management is characterized by mobile units, delegation of decisions to adequate operative levels, and a special type of “dialogue police” who communicate with demonstrators before, during, and after the event. This approach to crowd control is similar to what has been labeled “low-profile” policing (Stott et al. 2007) and also resembles the approach David Waddington described as “non-confrontational control” (Waddington 2007a). A central part of the new concept is the dialogue. The police officers patrol in pairs, without helmets and shields, talking with and even mingling with the participants. To state it simply, before this, police and demonstrators had two different goals—to carry out the demonstration and to maintain law and order. With the dialogue approach, the police and demonstrators have a common goal, that is, to carry out the demonstration in a decent and peaceful way. Starting with the Gothenburg riots in 2001, field research on crowd events has been conducted at Linköping University by a group of social psychologists. The field studies have been conducted as the Swedish police have implemented the new police tactics. This means that we have studied events where this tactic has been used as well as events where it has not. Besides the Gothenburg riots, two different types of political protests with annual demonstrations in Sweden have been studied—reclaiming parties on the first of May and nationalist marches with counterdemonstrations in a suburb of Stockholm in December (Granström and Rosander 2008; Guvå 2008). In addition, sports events have also been the target of research by this Swedish group. These studies have led us to believe that inter- and intragroup processes are vital to understanding why some events turn violent and others do not (Hylander and Guvå 2008; Guvå and Hylander 2008; Granström 2008; Granström and Rosander 2008; Guvå 2008; Granström et al. 2009). The following perspectives can sum up the focus of this chapter: • The intergroup perspective, in which intergroup processes guide the course of events in crowds. • The subgroup perspective, in which interaction between subgroups within the crowd affects the intragroup processes and vice versa. • The police perspective, in which the police have a crucial role in whether crowd events turn violent or stay peaceful.

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 71

POLICE USE OF MITIGATION STRATEGIES

71

• The mitigation perspective, in which it is necessary to plan for mitigating strategies and not just avoid aggravating strategies. • The organization perspective.

The Different Perspectives The Intergroup Perspective

In the course of an event, a large group may form a common social identity, a feeling of “us” or “we, the protesters” as opposed to “them, the establishment” or “them, the fascists” or “them, the EU proponents.” A broad social identity is constructed that many demonstrators can adhere to and hence the social identity becomes more salient than the individual identity. When confrontations between the police and protesters turn violent, the common social identity may change. A crowd may then unite and share a social identity of “us, the legitimate protester” against “them, the offending police.” The opposite is true for the police, where a social identity of “us, the legitimate police” against “them, the violent crowd” evolves. Based on the social identity perspective (Tajfel and Turner 1979), Stephen Reicher and his colleagues have developed the Elaborated Social Identity Model for crowd events (ESIM) explaining the interactions between protesters and police in terms of changes of social identities (Reicher 1996, 1997; Drury and Reicher 1999, 2000; Hopkins and Reicher 1997; Stott and Reicher 1998b; Stott and Drury 2000). Our research supports the ESIM approach to crowd violence. According to the ESIM, the out-group (the police) may understand the identity and actions of crowd members in ways that are different from the crowd members themselves (Drury, Stott, and Farsides 2003), for example, when the police define a protest, seen as legitimate by the protesters, as a threat to public order. The police thereby regard all protesters as a group as a potential threat and disturbance. The protesters, on the other hand, regard themselves as peaceful demonstrators in legitimate protest being stopped by a police action. When there is such an asymmetry of categorical representations between the in-group and the out-group and the out-group (the police) has the power to use interventions such as crowd dispersion, a social identity change may occur for the crowd members, who will find themselves being in opposition to the police and even condoning violent behavior that they would not have condoned previously (Drury and Reicher 2000). The proponents of the ESIM have convincingly shown how the interactive categorization of police and crowd members may account for changes in social identities. When a change transcends from a peaceful social identity to a social identity of victimized demonstrators, the demonstrators find that they have the right to defend themselves against the police and then the escalation of violence is a fact. The Swedish research has supported the ESIM and noted

05_Madensen_Ch05.qxd:Madensen

72

1/3/11

10:38 AM

Page 72

PREVENTING CROWD VIOLENCE

the same type of stereotyping categorization of the crowd by the police, resulting in a common social identity of protesters in opposition to the police. As a young boy in Gothenburg said, “it was like ‘we’ and ‘them’ even if we didn’t have so much in common.” The Subgroup Perspective

Even when a common social identity develops within the large group there are several subgroups of demonstrators with different pursuits and different social identities. There are different approaches to describing subgroups in large crowds. Renate Möller and colleagues (2009) explored young globalization activists’ motive structures and discerned eight different categories of protesters with different motives. These motive structures could be sorted into the following four categories: (1) politically idealistic and disagreeing with violence, by far the largest category (75 percent); (2) politically idealistic and agreeing with violence; (3) politically uninterested and disagreeing with violence; and (4) politically uninterested and agreeing with violence. This is an interesting result from a large survey with approximately 3,500 informants. However, this kind of survey does not give any information about how the protesters actually grouped themselves in the protest, their interplay with other participants, and the type of organizing that they were part of. The focus of our research has not been on a single individual motive but on how different subgroups organize their events and, in turn, how the police categorize and thus also treat the protesters. The Organizing Perspective

Besides the important process of mutual categorization between the police and protesters, our research has also elaborated on the organization of the crowd and the mutual treatment of the police and protesters as a result of that organization. In particular, we have focused on subgroups of protesters, their different types of organization, and the effects of this organization on the police treatment of that group. We have explored the relation between categorization, organization, and mutual treatment in the interaction between the police and different groups of protesters. Specifically, we focus on how the type of protest organization influences police categorization of the event and treatment of the participants. The Police Perspective

Police strategies in crowd events have been studied within sociology and political science, but the focus of research from a psychological perspective has typically been on crowd members. In recent years, however, police strategies in connection with political protests and sporting events have also been a focus for research within psychology (Drury, Stott, and Farsides 2003; Stott and Reicher

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 73

POLICE USE OF MITIGATION STRATEGIES

73

1998b; Cronin and Reicher 2006). Few, however, have studied and collected information from the police and the crowd in the same study, as in this study. The Swedish group has explored protesters’ perspectives as well as police perspectives, but in this chapter we will focus on the police perspectives. The Mitigating Perspective

Most crowd research has studied the occurrence and escalation of violence, while only a few studies have focused on the prevention of violence (e.g., see Stott et al. 2007) or on peacemakers instead of troublemakers (Russel and Arms 1998; Russel, Arms, and Mustonen 1999). This book is thus a substantial new approach to the issue of prevention, combining different disciplines in describing preventive police strategies in crowd events. Our chapter focuses on the presence and impact of peacemaking processes, particularly police mitigation strategies. One result of the research, which will be presented, is the AM model of aggravating and mitigating intergroup processes. Those opposite types of intergroup processes have simultaneously been observed between the police and the demonstrators within the same type of crowd events. Part of the interaction seems to be peaceful and mitigating, while other interactions are aggravating and lead to violence and disturbances. Our focus has successively turned to mitigation and strategies for de-escalation and violence prevention, which also will be the focus of this chapter. A main conclusion from the Swedish research is that it is not enough to plan for strategies to avoid aggravation at an event; it is absolutely crucial to plan for strategies that promote peaceful events. Two types of mitigation processes will be described: active mitigation strategies that promote peace and passive mitigation strategies that avoid aggravation. This chapter first describes the aggravation and mitigation model (AM model), which delineates the processes of mutual treatment, organizing, and categorizing. Thereafter it describes different categories of protesters according to their protests’ organization and discusses the outcome of crowd events in terms of how the police categorize different types of organizations and how that categorization governs how the protesters are treated. Then the AM model is exemplified by one event ruled by aggravation (Example 1, “the Hvitfeldtska event”), one peaceful public assembly (Example 2, “Reclaiming the street at Karlaplan”), and one example of how the police successfully used passive and active mitigating strategies at a public event (Example 3, “Anti-Nazi protest at Södermalm”). Finally, we will discuss the active and passive mitigating strategies used by police in these events. Thus, we will focus on the interaction between the police and the crowd, and the police categorization and subsequent treatment of the protesters. The ESIM tradition has soundly demonstrated how police categorization of a crowd brings about specific treatment that may escalate a

05_Madensen_Ch05.qxd:Madensen

74

1/3/11

10:38 AM

Page 74

PREVENTING CROWD VIOLENCE

conflict. What we add through our research is a notion of how the organizing of the protest affects the way the police categorize the event.

Aggravation and Mitigation: The AM Model This model (Table 5.1) is a result of a grounded theory study on the events in Gothenburg in 2001 (Guvå and Hylander 2008). It has been elaborated based on several new crowd events and has proved to be a useful tool for collecting and analyzing data from crowd events, political demonstrations, as well as sports events (Hylander and Granström 2009). Categorizing, organizing, and mutual treatment have proved to be important processes to explain the occurrence—or not—of violence and disorder and the escalation and de-escalation of violence. These processes may take the shape of mitigation or aggravation and lead to trust or distrust between the police and the protesters. When the police trust the protesters, they believe in their peaceful intentions. The concepts of aggravation and mitigation and of categorizing, organizing, and treatment can easily be translated into police tactics and strategies. Mutual treatment refers to actions and positions taken by protesters toward police and by police toward protesters. Mitigating treatment refers to disarming actions, and aggravating treatment refers to provocative actions. Examples of police disarming actions are police officers in plain clothes without helmets and gear, amiably communicating with protesters. Organizing refers to actions, activities, information, and utilities arranged by the protesters or the police to fulfill their goals. Mitigation refers to an organization, which is clear to the counterpart and where the aim of the organizing is peaceful toward the counterpart, while aggravation means creating chaos. Peaceful organizing occurs when all officers adhere to the same facilitating principles and tactics to assist crowds to achieve their legitimate goals, for example, small groups of visible officers deployed to signal peaceful intentions. Categorizing refers to the way the police perceive and label members of the protest and how protesters perceive and label the police. Mitigation takes the shape of differentiation and positive stereotyping while aggravation

Table 5.1

Aggravation and Mitigation (AM)

Relational Conditions

Aggravation

Mitigation

Mutual treatment Organizing Categorizing

Provocative Creating chaos Negative stereotyping

Disarming Peaceful organizing Differentiation and positive stereotyping

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 75

POLICE USE OF MITIGATION STRATEGIES

75

appears as negative stereotyping. When the police can differentiate in the crowd, they simultaneously protect the demonstration and deal with individuals who violate the law. Initial trust between the groups will promote mitigation, while initial distrust between the groups will promote aggravation. Also, the interaction between the subprocesses of mitigation (differentiation, peaceful orderly organization, and disarming) will enhance trust, while the interaction of the subprocesses of aggravation (negative stereotyping, creating chaos, and provoking) enhance distrust between the police and the protesters. Aggravation from one party may instigate a vicious cycle of negative interplay between the police and the crowd. Mitigation from one party may inspire mitigation processes from the other party and instigate helpful cycles. In our experience, one aggravating action from one party seldom instigates violence or escalation of violence. Typically, there needs to be a combination of negative stereotyping, chaos, and provocations from one party and aggravating responses to these actions from the other party to instigate violent disturbances. There are two types of mitigation processes that we have discerned as crucial strategies for the police: active mitigation and passive mitigation. Although aggravation and mitigation are interactive processes and the outcome of an event depends on demonstrators’ as well as police mitigating strategies, it is the police who have the power to legally intervene, and thus the strategies used by the police to secure mitigation are absolutely crucial for a peaceful outcome. Active mitigation strategies are carefully planned by the police and constitute an effort aimed at preserving and supporting the peaceful development of a mass event. The police use active mitigation when they disarm by dialoguing, by intentionally providing for a peaceful organization, and by differentiating and only intervening with those breaking the law. Passive mitigating strategies mean avoiding aggravation. This is the case when the police avoid actions that may be perceived as provocative, chaotic, or stereotypical by the protesters. Thus, passive mitigation means taking into account what protesters can be expected to perceive as aggravation. Hence the police have to be informed about their perspective. Passive mitigation also means that the police make a special effort not to be drawn into aggravation processes started by the protesters. They avoid being provoked even when challenged, avoid being negatively stereotyped by not wearing visors, and make every effort not to get involved in chaotic situations. Passive mitigation means taking responsibility not only to ensure that the police behave in a peace-making fashion, but that they also behave such that critical incidents are not escalated, irrespective of who started it. In some types of events, mitigation strategies come very naturally for the police when they trust the protesters to be peaceful. The challenge is to plan for mitigating strategies when they do not trust the protesters to be peaceful.

05_Madensen_Ch05.qxd:Madensen

76

1/3/11

10:38 AM

Page 76

PREVENTING CROWD VIOLENCE

According to our research, this trust is linked to the organization of the protest, so we place that discussion before an in-depth discussion of police mitigation strategies.

Different Ways of Organizing Protests The demonstrators, when interviewed, seldom view themselves as a unified group, but distinguish between subgroups with different objectives. How the police categorize groups is to a great extent linked to how the different groups organize themselves. This in turn seems to have a major impact on how the group is treated by the police. We have identified four different types of organizations of protests. In addition there are single, nonorganized individuals. A crucial point is how the police categorize the event, as that dictates their treatment of participants. This categorization, as a peaceful public assembly or a crowd that should be dispersed, is typically based on information about the organization of the demonstrators, according to our research (Granström 2008; Hylander and Guvå 2008). While demonstrations with an organizer, planned and announced to the authorities ahead of time, have typically been regarded as public assemblies to be protected, certain spontaneous events have been regarded as threats to the public order and therefore events to be dispersed (Guvå and Hylander 2008). When the assembly is categorized as a crowd that should be dispersed, the treatment of the crowd by the police may be experienced as provocative and chaotic. Thus the type of organization of the demonstrators is a factor that needs to be further explored. For the police, these different ways of organizing protests constitute very different contexts. On the basis of interviews with demonstrators and the police, as well as our own observations, four different categories of organizing are distinguished in terms of their leadership, decisionmaking, and intention (Figure 5.1). These categories are not groups with clear boundaries. One individual may join one type of protest in the morning and be categorized as a violent protester, and another type of protest in the afternoon and be regarded as a peaceful civilian citizen carrying out her/his democratic rights to demonstrate his/her views. Also, the boundaries are constructed in a dynamic interaction of mutual categorization between the police and protesters. In addition to the different types of categories of organizing protests, two categories of nonorganized individuals have been discerned. Although the categories—as we have described them—are dynamic and membership varies, they seem to exist before, during, and after the demonstrations. On the other hand, the extent to which specific individuals will identify with or be identified with a certain group is dependent on the course of events and the mutual categorization between demonstrators and the police. The demonstrators interviewed agree that these categories exist, but they have

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 77

POLICE USE OF MITIGATION STRATEGIES

Figure 5.1

77

Ways of Organizing Protests

Protest Organization

Organized

Leader governed

Rule governed

Goal governed

Nonorganized

Idea governed

Impulse governed

widely varying views on the sizes of the groups. The police, on the other hand, mainly categorize into two categories: those with peaceful intentions, the real demonstrators, and those with violent intentions, the troublemakers. Leader-Governed Protests

It has been possible to distinguish a specific type of organization in connection with nationalist demonstrations in a suburb of Stockholm (Granström et al. 2005b; Granström et al. 2009). From the observation protocols from three consecutive nationalist demonstrations (2004–2006), it was documented that order was maintained within the group with the aid of guards with armbands. One leader took it upon himself to submit an application for the demonstration, but it is unclear what mandate he had to do so. It is not clear, for example, which organizations were part of the umbrella organization arranging the event. It is questionable whether there were any functioning delegation or organized distribution of roles, or whether there was control from above by a self-selected leader. If so, this type of event may be recognized as organized according to premodernistic organization principles, not based on democratic principles. This theory is borne out by the fact that this is a gathering of people around an emotionally charged message, and that the ceremony was characterized by a military-style discipline. However, the protesters did not behave in a provocative manner toward the police. They complied with the requests of the police and had their own bans on alcohol and smoking for the demonstration (Granström et al. 2009). The police stated that it is easier to cooperate with nationalists than with antifascists. The nationalists have a formal leader with whom it is possible to negotiate, and the participants will obey him as their leader even if he is not formally authorized to negotiate on behalf of the organization. By cooperating

05_Madensen_Ch05.qxd:Madensen

78

1/3/11

10:38 AM

Page 78

PREVENTING CROWD VIOLENCE

with the leaders, the police also say that they have control over the members of the organization. Also, this organization bears a major resemblance to the police officers’ own hierarchical leadership organization. Few conflicts arise with the police as long as there is a leader who makes decisions and whom followers obey. Problems arise if the leader disappears or if individuals or groups fail to comply with the discipline. Attitudes toward counterdemonstrators, however, are expressed in extremely aggravating terms. The challenge for the police in this context is to avoid confrontations between the nationalist groups and the antifascist groups without being provocative toward one group or the other. Rule-Governed Protests

These are traditional protests arranged by groups of demonstrators organized in accordance with the modernist project, that is, with rules, leadership, division of roles, and democratic decisionmaking processes. Methods of order creation for demonstrations are similar to those of the police: (1) permission to demonstrate, (2) a prearranged route for the demonstration, (3) a clear arrangement and check of slogans, (4) demonstration guards, and (5) a ban on masks. These criteria were applicable to the three major demonstrations in Gothenburg that went off peacefully. In rule-governed protests, the participants are very explicit about their peaceful agenda. The protesters in these demonstrations often belong to established groups such as political parties or other nongovernmental organizations with plenty of experience in arranging demonstrations and public assemblies. Their arrangements are characterized by mitigation (differentiating, peaceful orderly organizing, and disarming). The rule-governed protest is a type of organization that the police recognize and condone and that also has a certain resemblance to the police organization. As the police perceive these arrangements to be lawful and not threatening, they will treat them with mitigating strategies since trust is high. Goal-Governed Protests

Goal-governed groups may operate on their own in small groups or may be part of idea-governed arrangements. These groupings range from more established to nascent groups, which arise, for example, as festival committees for one event. They are project organizations: the group exists as long as the project. The project group typically has a clear goal, for example, to perform a sitin demonstration or a “train-stopping.” The ultimate goal is usually to attract media attention to one political issue, such as Greenpeace actions to stop the extinction of species or the pollution of a specific area. Their intentions are typically not violent but governed by civil disobedience. As one of their means is to surprise authorities, these actions are not mediated to the police ahead of

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 79

POLICE USE OF MITIGATION STRATEGIES

79

time. The actions are well planned by small groups and may not even be announced to supporters. How the interaction with the police turns out depends on the target of the protest and how the police categorize the demonstration. The police can overlook offenses, like an unannounced sit-in demonstration, but not the destruction of tools or buildings. Protesters’ interaction with the police may range from the police peacefully guarding their actions to violent confrontations. We have seen examples of peaceful interplay between the police and protesters, where both groups know the rules of the game. The protesters expect the police to do their job and move the protesters or block them from entering a forbidden area, and they also expect to take their punishment. The police may trust the protesters to be nonviolent and therefore behave politely and negotiate before intervening. In other cases, however, the protest is categorized as unlawful just because there was no notification, and the police don’t trust the protesters to behave peacefully; the result is violent confrontations between the protesters and the police. Idea-Governed Protests

This category consists of individuals and groups who organize themselves into networks with no formal leadership. Their type of organizing is self-organizing, which means that things are organized on the spot at the time of the event. The organization may include music, vehicles, drums, fire eating, dancing, and food stands. It may also include collective actions such as marching with banners, and shouting and yelling. Individuals may be involved in many different groups. As individuals, they also join in with arrangements organized by umbrella networks. Participants in idea-governed protests have a more or less anarchist ideology. They agree on a common set of principles but each and every individual speaks for him/herself and not on behalf of any group; they do not appoint a spokesperson. Neither do they accept that they need to request permission for public events. Paradoxically, this makes them in accordance with the Swedish law, which says that a public assembly is to be guarded by the police, and it is never against the law to participate in a public assembly, but it may be an offense to arrange a public meeting without notifying the police. Thus not having leaders is a way to make the protests legal; however, it is not always recognized as such by the police. Even if most idea-governed protesters would wish to demonstrate peacefully, they do not take responsibility for what other protesters do. Attitudes toward violence are characterized by leaving it to each individual to decide what violence is necessary. This means that there is a wide variation in views: (1) protesters who never would use violence, (2) protesters who believe that they would never start using violence but have the right to defend themselves, and (3) protesters who believe that, in some cases, violence can be justified in

05_Madensen_Ch05.qxd:Madensen

80

1/3/11

10:38 AM

Page 80

PREVENTING CROWD VIOLENCE

political struggles. Since they invite everyone to take part in their events without restrictions, that means that even individuals acting as provocateurs without political motives may take part. For the police, the idea-governed event constitutes a challenge. It is crucial whether the police categorize these events as public assemblies in need of protection or as crowds causing disturbances. The organizing is not so easily recognized as peaceful, however peaceful the goal of the organization is. Even if music and dancing are strong attractions for young people and may prevent other activities like “rushes” that could cause chaos, such arrangements may still appear chaotic to the police when there is no formal organizer. Hence, the police do not know who to negotiate with and have no clear view of the event, as they have not been notified. Thus, self-organized events like “Reclaim the Street” and other spontaneous protests are perceived by the police as being much more risky than the traditional authorized and rule-governed demonstrations. Nonorganized Individual Protesters Provocateurs. Many protesters agree with the police that there is a small group of provocateurs who have objectives other than demonstrating their views. However, opinions are divided as to how large this group is and who belongs to it. This category is least defined and most dependent on how others categorize demonstrators. It is dubious that any of them actually would categorize themselves as provocateurs. During the Gothenburg incidents, the police believed that the group of provocateurs was fairly large, while views from the demonstrators varied from believing that it was fairly small to believing that it barely existed (Granström 2008). The police seem to expect this category to exist and to be present at major events. When things have gone calmly and peacefully, the police may draw the conclusion that “the hardcore weren’t there” (Granström et al. 2005a). This implies that the police believe that when there is violence it has been instigated by this specific group of provocateurs. The police recognize the difficulty in distinguishing the real troublemakers from others who have allowed themselves to be dragged into things. In contrast, protesters imply that their provocative actions may be the result of the police regarding them as provocateurs and trying to disperse the crowd. Among demonstrators, opinions are divided as to whether some people attend demonstrations with the intention of using violence. However, we have not interviewed one demonstrator who does not agree that there is a small group who can be thought of as agents provocateurs and that demonstrators see them as disruptive to their demonstrations. Thus, when the police intervene in a differentiating way with individuals whom other protesters regard as agents provocateurs, the protesters will find it a legitimate action and will not protest or question the police.

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 81

POLICE USE OF MITIGATION STRATEGIES

81

Thrill seekers. At several crowd events, we have noticed a specific category of bystanders who are very young (mostly) boys and who arrive in small groups of two to four. They are easy to talk to and happy to be interviewed; they are not part of any established groups and do not participate in the activities but float around on the outskirts of the event area. When asked what they are going to do, a typical answer is “see what is going to happen.” It doesn’t seem like they start rushes or violence, but as soon as a movement takes place, they join the rush and may participate in harm. As this category seems very susceptible to contact from adults, the challenge for the police would be to identify them early on and make contact.

Different Contexts

To summarize, it is clear that different categories of protests—regarded with more or less trust—make up very different contexts for the police. Trust between the police as a group and the protesters as a group is crucial for the chance of mitigating. Thus, there is very little danger that a protest arranged in a rule-governed way will get out of hand as the police will trust the participants to be peaceful and recognize that it is their duty to facilitate the protest. The risk for turmoil is much higher at a goal-governed or idea-governed event. If the police do not trust the protesters to use mitigation but fear aggravation processes and categorize the event as an illegitimate protest, they will act in a way that provokes the protesters. Since subgroups in the idea-governed category may use violence, the police may think they are right in distrusting the whole protest and use force to disperse the crowd. Then the subgroups, believing in their right to use violence in their own defense, will do so, and the police will appear to be supported in their argument that it was an illegal protest with violent intentions. On the other hand, if the police use mitigating strategies—that is, categorize the event as legitimate, differentiate, only use force on individuals using violence, and treat everyone else politely and with respect for their right to protest—the chance for a peaceful outcome is much higher. This argument will be further explored by examples from three different events: (1) the Hvitfeldtska event, (2) the reclaiming party at Karlaplan, and (3) an anti-Nazi protest.

A Crowd Event Ruled by Aggravation In order to contrast an event ruled by aggravation with events ruled by mitigation, the AM model is applied to one of the events in Gothenburg that turned into violence between police and demonstrators. It will be referred to as “the Hvitfeldtska event.” The report of the event is based on focus group interviews and individual interviews with police (n = 13) and demonstrators (n = 17) who

05_Madensen_Ch05.qxd:Madensen

82

1/3/11

10:38 AM

Page 82

PREVENTING CROWD VIOLENCE

took part in the event. Their narratives are checked against the Gothenburg report (SOU 2002). The event reported here is one of several events in the study of the Gothenburg riots and thus took place before the use of the new Swedish police tactics. However, police had started to dialogue (cf. Holgersson and Knutsson, Chapter 10) with a mandate to negotiate with the different groups of demonstrators prior to the event. The police did not at this time, however, have a clear mandate to negotiate during the events. Event Report

On the morning of June 14, 2001, the police in Gothenburg got indications that demonstrators carried weapons into a school that was being used to provide overnight accommodation for demonstrators. This was the same day that President George W. Bush was expected to arrive in Gothenburg, which meant that the police were under heavy stress. The police surrounded the school (Hvitfeldtska Upper Secondary School) with heavy containers and vehicles. The peaceful demonstrators involved in this incident thought that the police saw them, wrongly, to be violent. Opinions differ as to the rules of the game. According to the police, everyone who wished to leave the premises could do so. But a great many of the demonstrators believed that they were locked in. Others protested against being surrounded by the police and invited resistance by failing to leave the school voluntarily. There was probably a lot of confusion as to whether people were allowed to leave and how they should do so, but the police treated everyone who was left behind as if they had opted to stay and hence had also chosen sides. The police automatically categorized the ones who stayed at the school or in the schoolyard as demonstrators with evil intentions who could therefore be attacked. The police perceived the demonstrators who stayed as admitting to collaboration with the weapons bearers (who proved not to exist). At this point some protesters started throwing stones toward the police, thus confirming that they were violent demonstrators. There was no formal organization of the demonstrators, but a small group negotiated with the dialogue police. While they were still negotiating, the police stormed the building and took hundreds of protesters into custody. Demonstrators who trusted the police to protect their right to demonstrate had gathered at Hvitfeldtska Upper Secondary School, as the local authorities had opened up the school for their accommodation, but now had lost their trust in the police and in the organizers. Interpretation

In this example, the aggravation processes were allowed to dominate. The police started with distrust of the demonstrators based on high stress because

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 83

POLICE USE OF MITIGATION STRATEGIES

83

of President Bush’s visit and because of unchecked rumors creating chaotic information. Then they negatively stereotyped all demonstrators staying in the school as violent protesters. They provoked the protesters by surrounding the school. The police organization created chaos for the police and for the protesters. The communication between the dialogue police and the commander in charge was missing; negotiation was carried on at the same time as the commander in chief ordered the storming. The large gathering of police officers outside the school and the heavy vehicles were unexpected and not comprehensible to the demonstrators inside. Also, the demonstrators had a chaotic organization not clear to the police. The provocative encircling fed counterprovocations from some protesters who threw stones toward the police as a group. This situation probably could have been resolved differently if mitigation strategies had been given a chance to prevail. However, the actual event can be described by use of the AM model as follows. AM Model Description

As can be seen from Table 5.2, the event was characterized by the presence of aggravating relational conditions between police and demonstrators. The lack of mitigation efforts is noticeable. Thus, according to the AM model, it is not surprising that confrontations, oppression, and violence were the outcomes of this event.

Table 5.2

A Crowd Event Ruled by Aggravation

Relational Conditions Mutual treatment

Organizing

Categorizing

Aggravation

Lack of Mitigation

Police officers surround demonstrators. Demonstrators defend themselves with stones. Rumor. Deficient internal communication among police troops. Confusion among demonstrators. The police considered all demonstrators as hooligans. Demonstrators saw the police as a united force of aggressors.

Interrupted dialogue.

Lack of comprehensible organization.

Lack of differentiating.

05_Madensen_Ch05.qxd:Madensen

84

1/3/11

10:38 AM

Page 84

PREVENTING CROWD VIOLENCE

A Peaceful Public Assembly The example from Hvitfeldtska in Gothenburg is contrasted with an example from an event where the police intentionally used mitigation strategies. Two researchers observed the meeting from 6 P.M. to 10 P.M., one observer focusing on the participants and one observer focusing on the police. The data also include participant observations at meetings when the police discussed the event before and after the reclaiming party (Granström et al. 2005a). The example can be seen as a prototypical mass event resulting in a peaceful solution. Event Report

For several years, reclaiming activities on the first of May had gotten out of hand with confrontations between the police and young people in Stockholm. In 2004, the police, in accordance with their new tactics, had been encouraged to patrol in pairs in ordinary uniforms and caps and to mingle with the participants. First of all, they decided to categorize the event as a public assembly instead of a crowd event even if there had been no notification of authorities. Furthermore, they decided to make themselves very visible (wearing yellow jackets), but not to wear helmets, to mingle with the participants, and to keep vehicles, dogs, and horses at a distance from the main crowd. For some weeks prior to the reclaiming party, stickers had been put up all over the city bearing the words “Pirate Party, Karla Square, 1 May, 18:37.” Very little information was provided by the activists on the Internet, and nothing was stated other than the fact that the planned activity was a joint party. This message led to a certain sense of unease among shop owners in the district. When the young people arrived at the meeting place, those carrying rucksacks were searched by the police. This was seen as done in a friendly way, by single police officers or police officers in pairs, giving information about the reason for the intervention. Representatives of voluntary organizations (“Parents in the Streets,” etc.) were seen in the crowds. In other parts of the city, the police were monitoring groups that could have jeopardized the “pirate party”: a gang of militant football supporters and an extreme right-wing group. These groups were not permitted to go to Karla Square. Thus, the police protected the right of the young people to carry out their public assembly. The party got off to a rather tentative start; the youths wandered round and chatted, a few firecrackers went off. Bongos were played, young people in costumes danced. The music gradually got going; fire-eaters appeared. The police wandered round and searched the young people respectfully while chatting with them. Younger boys around thirteen to fifteen years old were seen to hang around the police officers, who showed them their equipment and talked to them. One fight broke out between participants. The police quickly moved on the young people involved. Various activities took place, like “islands” in the crowds of young people. A number

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 85

POLICE USE OF MITIGATION STRATEGIES

85

of “rushes” took place over the course of the evening. Suddenly, a small group of people started running, and a large number of young people joined in. The police reacted to this as a threat: they put on their helmets and formed small groups. However, all these rushes but one came to nothing. The rushes were observed to decline when several people in the rush seemed to realize that it was just movement and that there was nothing to rush for. Also, police officers said afterward that they had negotiated and talked with the young people rushing and persuaded them to return to Karla Square. As soon as the rushes had faded away, the police removed their helmets and put them in their belts, thus signaling that they were there to guard a peaceful event. Also, the police did not permit the rushes to leave the area. Vehicles, horses, and more police officers were placed in nearby streets, and this helped to restrict the street party to the occupied area. Thus there was control of those coming and leaving the area. There were no interventions against the crowd, which totaled around 1,000 participants, although individual youths with dangerous tools were dealt with discreetly. Two hours after the party started, groups of young people were seen leaving the party. In the observation protocol it is noted that those leaving seemed to be somewhat younger than the rest of the participants. Soon after that, a rush took place. It started outside the party place. Some people at the party place also started running but were stopped by the police and they returned to the party. At this rush, window crashing took place in one street. It was, however, a minor and isolated incident compared to a year earlier, and it did not spread to the rest of the assembly. One person was arrested and authorized action was begun. Young people interviewed at the party place were totally unaware of the damage. At about 10 P.M., the music stopped playing and the young people drifted away, while some stayed to clear the party site of beer cans and rubbish. Interpretation

The interpretation of these strategies is that the police managed to let mitigation rule the event (Table 5.3). Those few who used aggravation processes did not have an influence on the party and the police managed to protect the party as a public assembly and intervene with those individuals violating the law. The conclusion is that it was a totally different outcome compared to earlier years when violent confrontations between the police and the young people had taken place as the police had tried to disperse the crowd. The police used mitigation strategies by categorizing the event as a public assembly that should be respected and protected and avoided negative stereotyping by walking in pairs wearing plain clothes. The police differentiated treatment of the participants and only intervened with those who caused trouble. Using the AM model, the outcome of this mass event can be described as below.

05_Madensen_Ch05.qxd:Madensen

86

1/3/11

10:38 AM

Page 86

PREVENTING CROWD VIOLENCE

Table 5.3

A Peaceful Public Assembly

Relational Conditions

Lack of Aggravation

Mitigation

Mutual treatment

A couple of discreet apprehendings.

Organizing

Rushes faded away, did not instigate chaos.

Categorizing

No signs of negative stereotyping.

Police and demonstrators were chatting. Police officers treated the participants nicely. Information was given simultaneously with demonstrators being searched and dangerous objects taken away. Heavy vehicles placed out of sight. Police negotiated with demonstrator volunteers. Police in other part of the city to protect the event from “hooligans.” Police categorized the mass event as a public assembly to be protected and the participants as party guests: positive stereotyping.

AM Model Description

In the example above it is obvious that mitigation was dominant. All signs of aggravation were prevented or prevented from spreading in the crowd. As can be seen from these examples, the police conduct was quite different in the two cases. In Gothenburg, the police forces focused on preventing riots with an aggravation strategy, which resulted in violence, oppression, and confrontation. In the second example, the police officers used a quite different approach. First and foremost, the party was categorized as a public assembly to be protected, even if no prior notification to the authorities had been made. The police adjusted their organization to meet the protesters and to negotiate with volunteers even though no formal spokesperson was appointed. According to this categorization, the police treated the participants as peaceful and harmless party guests. They even marginally took part in the party. A crucial factor in this context seems to be the police officers’ ability to differentiate between various types of activists. In this second example, the police were successful in identifying a group of troublemakers who intended to join in and to cause violent encounters. They prevented them from entering the sealed-off

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 87

POLICE USE OF MITIGATION STRATEGIES

87

area, and thereby probably secured a peaceful situation. In contrast, the police officers at the Hvitfeldtska event proved to be unable to differentiate between peaceful teenagers and what they believed were violent protesters, thereby provoking peaceful youngsters to behave defensively and violently.

An Example with Active and Passive Mitigation Strategies An example of both passive and active mitigation from the police toward an idea-governed protest is given below. Mitigation strategies were based on police knowledge of the demonstrators’ intentions and their acknowledgment of the protesters’ right to demonstrate. Event Report

In a counterdemonstration against a neo-Nazi march in Stockholm in 2004 (Granström et al. 2005b, 2009)—an idea-governed protest—around 400 people first gathered at Södermalm Marketplace. The authorities were not notified. The police did not know if they were to walk somewhere else. According to our observations and interviews, the participants did not know if there was a marching route. The police negotiated with protesters who offered to talk with them and they agreed on a marching route. People started to walk and ended up at Maria Marketplace, which they had agreed on. However, the protesters did not stay there but continued to a third marketplace, Medborgarplatsen, where protests had been held earlier in the day. The police did nothing to prevent this new destination. However, they firmly blocked the streets leading to the central station where the neo-Nazis were expected to arrive by train. They also later blocked the entrance to the subway, telling the protesters that it was dangerous for them to go there because of the neo-Nazis and it would be much safer for them to go on the bus. Thus the police did not get provoked by the demonstrators not following the agreement but focused on mitigation by preventing any confrontation between the two countergroups. Interpretation

The police avoided becoming provoked and organized their activities in such a way as to keep their heavy vehicles at a distance, but still blocking the road so the opposing groups would not be in contact with each other. By letting the demonstrators march and shout and sing and choose their own meeting place, the police signaled that they recognized their legitimate right to protest and that they wouldn’t stop nonviolent activities.

05_Madensen_Ch05.qxd:Madensen

88

1/3/11

10:38 AM

Page 88

PREVENTING CROWD VIOLENCE

AM Model Description

To summarize our observations and learning, we will use the AM model once again, now focusing on passive and active mitigation strategies (Table 5.4).

Successful Active and Passive Mitigation Strategies Used with Idea-Governed Protests Based on this research, we conclude that active and passive mitigation are the pillars of police crowd management. Other researchers have put forth similar conclusions on crowd dynamics in connection with both demonstrations and football matches (Stott et al. 2007; Stott and Drury 2000; Adang and Cuvelier 2001b). It has been suggested that promoting a change from frustrating to facilitating a demonstration—that is, turning from crowd management to facilitation—is a key issue for success (Reicher et al. 2004). Passive mitigation

Table 5.4

Examples of Active and Passive Mitigation Strategies Proved to Be Effective

Relational Condition

Passive Mitigation (avoiding aggravation)

Mutual treatment

Avoiding being provoked Keeping calm. Avoiding provocations Appearing without heavy equipment.

Organizing

Avoiding the creation of chaos Avoiding sudden movements. Keeping horses, dogs, and large vehicles at a distance from demonstrations.

Categorizing

Avoiding being negatively stereotyped Not appearing in troops.

Active Mitigation Disarming Friendly dialoguing. Mingling with demonstrators. Giving information and answering questions. Supporting demonstrators’ festival arrangements. Peace-organizing Adjusting to an unofficial organization of the demonstrators. Ad hoc negotiation. Keeping agreements. Planning ahead and being informed. Positive stereotyping Categorizing a mass event as a public assembly to be protected. Differentiation Intervening only with individuals who violate the law.

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 89

POLICE USE OF MITIGATION STRATEGIES

89

may be a strategy for crowd management, while active peacemaking may be a strategy to facilitate and secure the right of the demonstrators to carry out their legitimate pursuits. Both passive and active mitigation seem necessary. Lately, passive strategies have been criticized in Swedish media, although when used, the police have reported that they result in less violence. However, the passive mitigation strategies most likely must be combined with active mitigation to appear meaningful to the police. The two strategies go together. Specific to the AM model is the emphasis on the type of event organizing and how this influences the way the police categorize and thus treat the demonstrators. For the police, the idea-governed protests constituted the greatest challenge and ran a risk of being categorized as illegitimate crowd events that should be dispersed. This category is also the most vulnerable to negative stereotyping by police; that is, if the police categorize this event as a crowd to be dispersed, they expect opposition and resistance that they would not expect in a rule-governed event. Also, the idea-governed events were those that indeed profited from the police mitigating strategies. The police have to understand and recognize the mitigation processes of the demonstrators to achieve successful passive mitigation. This is not always easy, as the behavior of the ideagoverned demonstrators may seem chaotic, provocative, and stereotypical. Through the use of dialogue, through communication before, during, and after events, and especially through the adjustment of the police organization to the informal organization of the protesters, the police may learn more about the idea-governed protests and what “gestalt” their peaceful intentions may take as well as be able to differentiate those individuals who instigate violence. Furthermore, by using active mitigation—disarming through friendly dialogue and information, peaceful organizing, and the categorizing of events as legitimate—the police signal their peaceful intentions, which may inspire trust and more peaceful behavior from the protesters. The following is a summary and discussion of the successful active and passive mitigating strategies as they emerged in the three examples given in this chapter. Mitigating Treatment

Active and passive strategies were used for mitigation in the treatment of groups and individuals. Disarming as active mitigation. The police turned up in ordinary uniforms and caps instead of helmets, chatted with the demonstrators, and even cracked a joke or two instead of behaving formally. This gave signals to the demonstrators that the police would treat them in a peaceful way and respect their assembly. This echoes the studies of Russell, Arms, and Mustonen (1999) that humor and good interaction between police and citizens seem to have defused potential conflicts in crowd events.

05_Madensen_Ch05.qxd:Madensen

90

1/3/11

10:38 AM

Page 90

PREVENTING CROWD VIOLENCE

From the observation protocols, there are an abundance of examples of police officers entering into communication with the demonstrators so as to form the basis for a disarming and peaceful progression. The people who asked questions of the police at the reclaiming party at Karlaplan received answers, in contrast to what happened at the spontaneous protests and reclaiming parties in Gothenburg, where the demonstrators perceived that the police avoided answering or did not know what to answer (Granström et al. 2009). Maintaining a dialogue and negotiating was a particular challenge to the police with the idea-governed protesters, who advocated individual responsibility and did not allow themselves to be represented by anyone. However, the police were seen to negotiate with those who would negotiate with them. This was facilitated by the presence of the dialogue police—the specially trained police officers whose special task is to communicate with demonstrators before, during, and after an event. The dialogue police were the ones that the protesters turned to first, when they judged that the police had violated their rights. The dialogue approach also proved to be successful for football tournaments (Hau 2006). By protecting the area where the reclaiming party at Karlaplan took place, the police showed that they attempted disarming treatment. Searches could lead to actual disarming if they took place in a friendly and informal manner and also proved to be an opportunity for contact and communication. This was possible as the police officers appeared without any provocative equipment, such as helmets, visors, shields, horses, or military formations. Combining a provocative act such as searching with clear information and friendly dialogue neutralized the act, and such treatment did not become the spark that sets off violence. Disarming also took place when the police and participants openly showed their weapons. For example, young boys at the reclaiming party at Karlaplan stood around and chatted with the police, who showed them their helmets and shields, which the boys could touch and ask about. Disarming as active mitigation is similar to what has been labeled lowprofile policing, which has proved to be effective in connection with high-risk football tournaments (Stott et al. 2007). Also, the respectful treatment of the police may lead to the participants recognizing the police differentiating actions as just differentiating and thus legitimate (Reicher et al. 2004). Disarming as passive mitigation. Disarming as passive mitigation involves

avoiding provoking and being provoked. This involved the police treating demonstrators in a way not perceived as provocative, and keeping calm even when protesters used provocative language or did not follow the rules. We have reported several examples where the police have kept calm when violation of rules or agreements, such as minor changes of route, have been accepted because they have not been seen as causing severe disturbances. If a

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 91

POLICE USE OF MITIGATION STRATEGIES

91

single policeman behaves in a provocative manner and protesters react to that as a group, de-escalation is more likely if the other police officers stay calm and do not react as a group. Mitigating Organizing

There were many examples of active and passive mitigation through the organizing of police officers. Organizing as active mitigation. An organization that signals trust and com-

prehensibility promotes mitigation. For the police it is important to be clear and informed of their tactics so as to avoid internal chaos, in contrast to what happened in Gothenburg, where the police described how the organization they had built up and planned rapidly descended into disorganization (Granström et al. 2009). Using situation-specific tactics, part of the new police tactics, is one way of attempting to bring about appropriate “here-and-now” organization that may be better adjusted to meet the idea-governed organized protests. Our observations from the reclaiming party at Karlaplan gave examples of how the police were organized to mingle with the protesters to inform and clearly answer questions instead of being organized as a troop. In this way the police modified their own organization and adjusted it to the type of organization that the demonstrators were using. This adjustment of the police organization was seen in their negotiating with volunteer demonstrators at the antiNazi protest at Södermalm, when they had no official spokesperson with whom to carry on the dialogue. Organizing as passive mitigation—avoiding chaos. Police officers avoided chaos when they prevented rumors from spreading, or avoided sudden unexplained or inexplicable actions. But even very well organized actions may create chaos for the demonstrators if they are perceived as frightening. Thus, the tactic of having heavy vehicles and horses at a distance from the demonstration was a way to avoid chaos. In order to make this adjustment, the police had to learn about how the demonstrators experience the police organization. Even if the aim of a police intervention is to create order (to counteract or prevent a breach of order), the consequence may be the opposite, that is, creation of disorder and chaos. If equipment and joint movements from the police cause uncertainty and fear, this may lead to rumors gaining ground and spreading, which in turn results in further uncertainty and chaos. To counteract this, the police must understand how their interventions are perceived. There is not much in the research literature about the direct impact of chaos on the course of events at demonstrations. Rather, chaos is seen as the result of a riot. The ESIM proposes that a change in social identity from a

05_Madensen_Ch05.qxd:Madensen

92

1/3/11

10:38 AM

Page 92

PREVENTING CROWD VIOLENCE

peaceful identity to being in opposition with the police is enhanced by an ambiguous situation (Reicher 1997). We have seen several examples of cases where chaotic organization of the police as well as of the demonstrators actually seems to be crucial to a negative stereotyping of the other group and thus leads to provocation (Granström et al. 2002; Granström et al. 2009). Mitigating Categorizing

Police categorizing of groups may facilitate passive or active mitigation. Positive stereotyping as active mitigation. “All demonstrators have the right to demonstrate and should be protected.” This is an example of positive stereotyping, which was carried out by the police in connection to the reclaiming party at Karlaplan. The event was categorized as a public assembly that should be protected even though no authorities had been notified prior to the event. The participants as a group were regarded as peaceful party participants, and most participants responded with mitigating processes, like obeying the rules of the police to stay at the party place. Also, the anti-Nazi protest at Södermalm was considered an event to be facilitated by the police although no notification had been given. This is in sharp contrast to what happened at the Hvitfeldtska event in Gothenburg, where the police considered all demonstrators staying in the school to be violent protesters opposing the police. As a matter of fact, all idea-governed protests studied in our material—not announced to the police but categorized as legitimate assemblies—turned out peacefully, while those categorized as illegitimate turned into violent confrontations (Guvå and Hylander 2008). Differentiation as active mitigation. The police at the reclaiming party at

Karlaplan differentiated by intervening only with those individuals who violated laws. The result was that participants also differentiated and did not oppose the police when the police handled participants involved in fights. Avoiding negative stereotyping as passive mitigation. The police used the

dialogue to get to know protesters and thus also gain the understanding that all young people dressed in black or with hooded jackets are not potentially violent protesters. Avoiding being negatively stereotyped as passive mitigation. When the police showed their faces and talked to protesters, they also avoided being negatively stereotyped. The protesters realized that they were human beings and individuals, which is in sharp contrast to testimonies from Gothenburg, where demonstrators perceived the police as a troop of “stone-faces.”

05_Madensen_Ch05.qxd:Madensen

1/3/11

10:38 AM

Page 93

POLICE USE OF MITIGATION STRATEGIES

93

Conclusion From the five perspectives we outlined in the beginning of this chapter—the intergroup perspective, subgroup perspective, police perspective, mitigation perspective, and organizing perspective—we can now conclude that the challenge for the police in Swedish crowd events is treating idea-governed and goal-governed protests. The outcome is peaceful if the police manage to perceive the organization of those protests as orderly and reflecting peaceful intentions, and hence categorize the protests as legitimate events to be facilitated. The new Swedish Special Police Tactic for crowd management seems to have given a basis for a type of organization that can match idea-governed and goal-governed protests. Furthermore, the new tactics seem to include strategies for positive categorization and mitigating treatment of idea-governed and goal-governed protests. Lessons learned from our research are summarized below. If applied, these strategies will probably increase the likelihood of successful and peaceful crowd events. We have learned that the following types of situations must be understood in depth in order to promote peaceful events and avoid strategies that are likely to escalate a critical situation. 1. Promote positive categorizing. Positive categorizing is promoted when the police regard events as legitimate public assemblies and enhance and support all mitigating strategies of the demonstrators. When the police officers succeed in supporting the organization of an event and at the same time differentiate between peaceful participants and those causing trouble, they are successful in maintaining peace. 2. Avoid negative stereotyping. When the police avoid activities directed to the whole crowd, they also avoid giving peaceful participants the feeling that their legal rights have been violated, which could give them arguments to violently defend themselves. We have noticed that when police officers appear in pairs without heavy equipment, the participants stay relaxed, as they don’t apprehend a threat from the police. Also, participants will not generalize to all policemen if there is a controversy between some participants and some policemen. 3. Promote peaceful organizing. The police promote peaceful organizing when they adjust their own organization to the organization of the event, to make informal negotiations and agreements and to support the organization of the event. This is particularly important when dealing with idea-governed and goal-governed protests. 4. Avoid chaos. In chaotic situations, the risk for misunderstandings that lead to violence is probably much increased. Checking rumors before acting is one such strategy. Good planning, clear information, and

05_Madensen_Ch05.qxd:Madensen

94

1/3/11

10:38 AM

Page 94

PREVENTING CROWD VIOLENCE

decisionmaking close to the actual events promote peaceful events but also an understanding of what types of actions are perceived as creating chaos by the protesters. 5. Promote disarming. It is clear that friendly communication and dialogue seems to be a very effective tool to maintain peace and order. Also, actions such as searching for dangerous tools or objects are neutralized when friendly information is given at the same time. 6. Avoid provocative behavior by the police. Participants may react negatively to equipment like dogs, horses, police buses, and riot fences and feel that the police are expecting trouble. Once again dialogue has proved to be a most effective disarming strategy. From several observations we noticed that chatting and mingling with participants are more appropriate conduct than, for instance, showing up with heavy equipment. The police have the responsibility for and the power to secure order at crowd events. It is therefore important for the police to have access to explanatory models that can guide them to select appropriate strategies, but also to understand the arguments behind the strategies. We think that the AM model may be useful in this connection.

Note The research presented here was supported by the Swedish Emergency Management Agency.

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 95

6 Policing the British G8 Protests: A Contextualized Analysis David P. Waddington

CONSIDERABLE ACADEMIC RESEARCH AND THEORIZING HAS recently been focused on the police management of the transnational protest accompanying international summit meetings, such as the World Trade Organization and G8 (Group of Eight) summits. Invariably these studies have sought to understand the nature and choice of particular public order tactics and strategies and their possible contribution to any accompanying violence (e.g., della Porta, Peterson, and Reiter 2006; Fernandez 2008; Noakes, Klocke, and Gillham 2005). The most recent UK example of a major transnational protest involved the policing of a series of G8 ministerial meetings in Derby, Sheffield, and Edinburgh in March, June, and July 2005, respectively. A case study was undertaken by the present author and a colleague of the police management of the Sheffield protest (D. Waddington 2007a; Waddington and King 2007). Corresponding research was carried out by Hugo Gorringe and Michael Rosie (Gorringe and Rosie 2006, 2008a, 2008b; Rosie and Gorringe 2009) on the policing of events in Edinburgh. This chapter draws together the main findings and conclusions of these complementary studies and employs the Flashpoints Model of Public Disorder (e.g., D. Waddington 1992; D. Waddington et al. 1987, 1989) to highlight the benefit of adopting a contextualized approach to aid our understanding of the nature and impact of police public order tactics and strategy. Policing operations implemented both in Sheffield and Edinburgh and at the preceding meeting in Derby had several characteristics in common. However, there were fundamental differences, too, concerning not only the management of each summit, but also the handling of separate protest activities occurring within each city. For the Derby meeting of G8 Environmental Ministers in March, the police followed recent precedents set at earlier G8 meetings (D. Waddington 2007a, chapter 5) by erecting a metal fortification perimeter (a “ring of steel”) around the golf course location where the summit 95

06_Madensen_Ch06.qxd:Madensen

96

1/3/11

10:40 AM

Page 96

PREVENTING CROWD VIOLENCE

meeting was held. The Public Order Act 1986 was simultaneously invoked to confine 3,000 protesters to a small marketplace in the city center, miles from the summit venue, where no arrests were made. The series of three main protest events accompanying the meetings of the G8 Justice and Home Affairs Ministers in Sheffield were jointly organized by members of Sheffield Dissent, a network of resistance to the G8, and the Sheffield Stop the War Coalition. The first of these occurred on the Saturday prior to the scheduled meetings (June 11), when 1,000 demonstrators engaged in a trouble-free city-center march. This was followed by a pair of static demonstrations, staged to coincide with two civic receptions in honor of the ministerial entourages at the downtown locations of the Winter Gardens and Cutlers’ Hall on Wednesday, June 15, and Thursday, June 16, respectively. At each of these gatherings, South Yorkshire Police (SYP) also invoked the Public Order Act 1986 to restrict the protesters to designated protest, limit the numbers of those present to a maximum of 150, and forbid any marches or processions. No such numerical or spatial restrictions were placed on a secondary protest activity also occurring on Thursday night, on the outlying Devonshire Green, situated a mile or so from the Cutlers’ Hall. Here, at a “Rice for Dinner” meeting organized by Make Poverty History, some 200–300 people present ate small portions of rice to draw attention to the inadequate diets of the world’s most impoverished nations. The small instances of disorder accompanying the midweek events occurred well away from the actual protests. Police made three arrests on the Wednesday evening as a small group of anarchists tried to evade a line of officers protecting a side road on the periphery of the protest. A further seven arrests were made the following night when a section of people attending the Rice for Dinner meal suddenly engaged in an impromptu march toward the city center. Police quickly intercepted them and corralled them into a side street, where they were kept waiting for well over two hours. During this time, police snatch squads entered the crowd and made occasional arrests. Three weeks after the Sheffield demonstrations, the focus shifted onto Edinburgh, where a summit meeting of the G8 heads of state was staged from July 6–8 inclusive at the Gleneagles Hotel, Auchterarder, just outside of Edinburgh. Prior to the actual summit on Saturday, July 2, the city played host to its largest-ever demonstration, a “Make Poverty History” march comprising 225,000 people from families, faith groups, and charitable organizations, UKwide. The march proceeded within a carnivalesque atmosphere, with the majority of participants wearing white clothing with a view to encircling Edinburgh in a “white band” that was symbolic of the antipoverty movement. The only notable instance of disorder occurred in the midafternoon when police in riot gear corralled a group of 60–70 “black bloc” protesters into the confines of a side street and detained them for an hour. One person was arrested in the process. On the following day, a smaller than expected gathering of 4,000 peo-

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 97

POLICING THE BRITISH G8 PROTESTS

97

ple attended a city-center “Stop the War” rally. Police had been expecting a much larger and potentially unruly crowd of 25,000, but the event passed off without incident. The hitherto peaceful outcomes of the weekend protests were in stark contrast to the so-called Carnival for Full Enjoyment of Tuesday, July 5. Here, police reacted to perceived threats to major city-center financial institutions by using corralling-in tactics and powers available under Section 60 of the Criminal Justice and Public Order Act 1994 to both contain and detain demonstrators for some three hours. At one stage, protesters reacted by overturning park benches and employing them as offensive (or, arguably, defensive) weapons against the police. Further violence ensued on the first day of the meetings of the heads of states (July 6). Early that morning, police accused residents of temporary “eco camps” in Stirling and Bannockburn of having vandalized nearby cars and business premises. Confusing messages then followed, suggesting that the main day’s protest—a march through Auchterarder, bypassing the summit venue—was being called off by the police. Amidst the mounting confusion, a police decision to cancel several buses scheduled to go to Auchterarder provoked a spontaneous protest march, which culminated in city-center confrontations between protesters and the police. At Auchterarder itself, police and demonstrators confronted each other on either side of protective fencing separating off the public from the Gleneagles Hotel. The temporary breaching of this fortification resulted in police being flown in by Chinook helicopters and the intervention of mounted police and dog handlers who quickly regained control of the crowd. In the course of studying the Sheffield protests, the present author interviewed a total of thirty-two respondents, including two senior police officers, city council officials, media personnel, and an assortment of protesters (D. Waddington 2007a, 142). The Edinburgh research likewise included formal interviews with police officers (in this case, three senior and three of lower rank) and a range of informal interviews with “dozens of protestors and police officers” (Gorringe and Rosie 2008a, 191). Both studies also employed participant observation as one of their main methods and each drew on local and national media coverage of events. The Sheffield study was primarily geared toward “analyzing the various institutional, cultural, practical, professional and political considerations affecting South Yorkshire Police’s (SYP’s) tactics and strategy for managing the Sheffield meeting” (D. Waddington 2007a, 139). Gorringe and Rosie’s research emphasized the importance of the knowledge, preconceptions, and assumptions (or frames) informing the police operations, and the way they were influenced by such factors as local media precoverage, police culture, and the history of police-protester interaction. This chapter attempts to consolidate these separate explanations into a more unified and coherent analysis

06_Madensen_Ch06.qxd:Madensen

98

1/3/11

10:40 AM

Page 98

PREVENTING CROWD VIOLENCE

based on an application of the flashpoints model. Following a brief outline of its seven levels of analysis, the model is then used as the basis of a contextualized explanation of the police methods employed both in Sheffield and in Edinburgh, and of their consequences in terms of any disorder that may have occurred. The chapter concludes by noting the implications of the analysis for police public order policy and best practice.

The Flashpoints Model of Public Disorder The flashpoints model provides a general framework for explaining the conditions in which disorder breaks out or, alternatively, fails to ignite. It incorporates seven interdependent levels of analysis: structural, political/ideological, cultural, institutional/organizational, contextual, situational, and interactional. The structural level of analysis refers to the various forms of material inequality, inferior life chances, and political impotence underlying collective grievances in society. It further emphasizes, however, that “ideological alienation from the state” (involving moral opposition to governmental policy regarding, say, the environment, world peace, global poverty, or sexual politics) may also serve as a preliminary basis for bringing dissenting sections of society into opposition against the police. The political/ideological level is concerned with the way in which various forms of political and ideological pressure are influential in shaping police policy toward these dissenting groups. Sometimes this pressure is implicit, as where the police feel an almost instinctive professional obligation to “die in a ditch”—resort to firm and, if need be, confrontational measures—in defense of politically symbolic “occasions, locations and personalities,” such as government buildings, national monuments, foreign embassies, and royal ceremonials (P. A. J. Waddington 2003). A classic example of this is the fortification mentality surrounding the protection of the Internationally Protected Persons attending world summit meetings (e.g., King and Waddington 2005). On other occasions, political pressure is more overtly exercised—as in the form of direct governmental intervention, or where public opinion is brought to bear by “democratic audiences” (Jefferson and Grimshaw 1984), such as local politicians or pressure groups. Tony Jefferson and Roger Grimshaw make the point that it is primarily where these audiences “can represent themselves as speaking on behalf of other than sectional interests” that they are able to influence police policy. The mass media are pivotal in this regard: “They are the most powerful mediators of public opinion, in their self-claimed role of ‘public watchdogs’; they also partly construct it, given their unparalleled ability to determine whether interests are reconstructed in negative (‘sectional’ or ‘partisan’) or positive (‘universal’ or ‘public’) terms” (Jefferson and Grimshaw 1984, 86–87).

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 99

POLICING THE BRITISH G8 PROTESTS

99

Those dissident groups finding themselves and/or their aims delegitimized by the media are liable to be subjected to uncompromising police policy. Equally liable to affect police attitudes and conduct toward protesters are factors residing at the cultural level of analysis. Here we are concerned with the ways in which sections of society (including the police) make sense of the social world, based on shared material conditions, experiences, and position in the social structure. This incorporates not only people’s definitions of themselves and others, but also the rules informing their attitudes to using violence and recipes of how to respond in particular situations. In policing terms, senior commanders have often been known to distinguish between “good” and “bad” protesters, the former being those defined as harboring genuine grievances, and the latter, those merely “jumping on the bandwagon” in order to have a go at the police (P. A. J. Waddington 2003, 410–411). Disorder is more likely to occur where there are extreme “clashes” in the cultural values subscribed to by police and protesters—most predictably, perhaps, in situations involving officers from outside forces, not sufficiently accustomed to local sensibilities and relatively unconcerned with the need to preserve healthy long-term police-community relations. There may be problems of communication, coordination, and authority, as well, wherever visiting officers are subjected to local structures of command (della Porta and Fillieulle 2004; della Porta and Reiter 1998b). Another important cultural factor is that senior police may seek to learn from such “lessons of the past” as the outcomes of their own previous dealings with particular groups of protesters, or the general implications for strategy and tactics arising from the policing of major international summits (King and Waddington 2005, 2006). This latter point is emphasized by the work of John Noakes and Patrick Gillham (2004, 2006), who show how, in the wake of the infamous confrontation between police and protesters at the 1999 World Trade Organization summit meeting in Seattle, North American and European police forces have resorted to a broadly similar range of tactical approaches for dealing with “transgressive” (unruly and uncooperative) protesters, for which they have coined the term strategic incapacitation (Noakes and Gillham 2006, 111–114). This involves such tactics as disrupting the preparations of transgressive protesters, making preemptive arrests, establishing large no-protest zones, using corralling techniques, employing surveillance methods before and during protest events, and deploying less-lethal weaponry. All of this is done with the primary goal of disabling the effectiveness of the targeted protesters for the duration of the event. The danger here is that the practical implementation of such measures at the situational level (see later in this chapter) is liable to produce heightened feelings of outrage and indignation among those protesters who believe the right to protest effectively has been violated by the police (Mitchell and Staeheli 2005).

06_Madensen_Ch06.qxd:Madensen

100

1/3/11

10:40 AM

Page 100

PREVENTING CROWD VIOLENCE

There is considerable overlap here between the cultural level of the model, and its more recently introduced institutional/organizational level (D. Waddington 2007b). Based on insights contained in the work of della Porta and her colleagues (della Porta and Fillieulle 2004; della Porta and Reiter 1998a, 1998b), this level is used to refer to a range of formal structures, systems, rules, regulations, and documentation, concerning such variables as: the availability and nature of public order legislation; the degree of police paramilitarism inherent at local and national levels; the extent and quality of relevant training regimes and methods and of police accountability; and the nature and pervasiveness of any policing traditions, philosophies, or ‘missions’ embodied at national or local levels of policing. Recent work attributes the exceptional instances of police violence occurring at the G8 summit meeting in Genoa (Italy) in 2001 to, among other factors, such key institutional variables as the weak Italian structures of accountability and the police’s traditional commitment to a “king’s peace” operational philosophy, predicated on an unquestioning obedience to whatever government happens to be in power (della Porta and Reiter 2006). A complementary local example shows how the heavy-handed and authoritarian treatment of antiwar protesters by the New York Police Department in 2003 was underpinned by the force’s formal adherence to a “zero tolerance” philosophy of public order policing that had recently evolved in response to civilian concern regarding the proliferation of incivility and nuisance throughout the city (Vitale 2005). The contextual level concerns those dynamic communication processes occurring in the build-up to an event that can serve to elevate or suppress any potential for disorder. Examples of this might include a negative history of interaction and/or controversial incidents between the police and a dissenting group, or it might be the arousing effects of rumor, media sensitization, or threatening statements by one or both sides. The possibility of disorder is more likely to be averted due to the presence of advance liaison and negotiation between police and protest organizers. Evidence suggests that the police are well-practiced in the dark arts involved in this process (P. A. J. Waddington 1998, 126–127). Typically, they resort to all manner of “interactional practices and ploys,” combined with “home ground advantage” and “monopoly of expertise,” to gain the compliance of their opposite numbers and ensure that the demonstration proceeds according to police preferences: Once negotiation begins, the police stance is one of proffering help and advice—“How can we help you?”—“help you,” that is, to “do it our way.” Many organizers are inexperienced and so the police “organize it for them.” They recommend routes along which to march, provide the telephone numbers of officials in other organizations that need to be contacted, and suggest how difficulties might be resolved. (P. A. J. Waddington 1998, 121)

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 101

POLICING THE BRITISH G8 PROTESTS

101

When done successfully, this process helps to generate mutual trust, safeguard against the possibility of police overpreparation, and pave the way for a set of mutually understood “ground rules” designed to accommodate both parties’ objectives. The absence of such dialogue generally forces the police into relying on various forms of intelligence gathering and surveillance (which will inevitably vary in terms of accuracy and authenticity) and making contingency plans based on “worst-case scenarios,” with all the associated risk of creating a self-fulfilling prophecy. The situational level focuses on relevant physical, dynamic, and subjective aspects of the spatial location or setting in which the event actually occurs. For example, the presence of targets of derision (e.g., representatives of a rival political faction or detested foreign politician) may enhance the prospect of disorder. Of corresponding importance are the ways in which the relevant space is perceived and managed by those present. Regarding the former, it is often the case that police and protesters will arrive at discrepant interpretations of each other’s behaviors. Thus, while police officers may justify their decision to clear the road of protesters and provide a protective cordon for delegates entering a National Front meeting as “upholding the right to free speech,” onlooking demonstrators might regard such conduct as reprehensively profascist. Disorder is generally more likely to be averted wherever self-stewarding arrangements have been agreed upon, policing is avowedly “low-profile” (with reinforcements in riot gear kept well out of sight of protesters), and there are clear lines of communication and command, not only between police units, but between senior police and demonstrators. This brings us on to the final, interactional level, which is concerned, as its name suggests, with the quality of interaction (whether respectful, cooperative, provocative, or restrained) between the parties involved. This is the level of analysis where “flashpoints” are apt to occur. In the course of a particularly fraught encounter, it may only require a single, precipitating incident, such as the hurling of a brick, an arrest or police charge, to trigger disorder. The significance of such “flashpoints” is that they are invariably interpreted symbolically as indicating a refusal by one or both sides to accommodate the perceived “rights,” interests and objectives of their opponents. (King and Waddington 2005, 259)

There is a strong likelihood that such incidents will be regarded as all the more provocative and “inflammatory” where they are accompanied by “intensifiers” relating to the characteristics of the individual(s) concerned (e.g., a high-ranking police officer, or a vulnerable person, such as a woman or child) or the type of activity perpetrated (e.g., an overzealous or degrading arrest). Oppositely, there is still a chance that the confrontation may be de-escalated with the help of placatory gestures (“pacifiers”) designed to restore the ruptured

06_Madensen_Ch06.qxd:Madensen

102

1/3/11

10:40 AM

Page 102

PREVENTING CROWD VIOLENCE

status quo—for example, apologizing for any miscreant or excessive behavior, or immediately releasing an arrested person.

Understanding Events in Sheffield In order to fully understand the nature of police tactics and strategy during the Sheffield summit, and appreciate the quality of their relations with protesters, it is necessary to explore the nature and interaction of key variables located at each of the above levels of the flashpoints model. It was primarily a sense of “ideological alienation” concerning such issues as Britain’s involvement in the Iraq War and the perceived inadequacy of global measures for dealing with world poverty that provided the structural basis for the G8 protests. For sections of those present, notably the participating anarchist groups, there were also undoubtedly keen feelings of political powerlessness and disaffection. Such grievances are not adequate in themselves, though, to explain the instances of violence occurring during the main midweek protests. A truly comprehensive explanation must first dwell on the political/ideological context framing the events. Various forms of implicit and explicit political pressure impinged on the strategy and tactics eventually adopted by SYP. To begin with, it was politically imperative for them to ensure the safety and security of the visiting G8 dignitaries and their various entourages. The perceived need to “die in a ditch” was undoubtedly enhanced by the fact that the demonstrations were likely to occur in close proximity to expensive city-center property and symbolic landmarks, such as the Winter Gardens and Cutlers’ Hall. Adding to this implicit political pressure was the more explicit influence exerted by Sheffield City Council, who made it plain well beforehand that they were keen to stage the G8 summit in such a way as to “showcase” the city as a tourism and conference center of potential world renown. Preliminary meetings between the chief constable and council executives had dwelled on the possibility of banning all organized protest. Police sources suggest that it was the former who ultimately reasserted the right to protest: The Chief had discussions at Chief Exec level and their strategy was “Let’s make Sheffield look as good as we can.” So the Chief says to [the council’s Chief Executive] “Well, that’s alright but there is this issue of the right to protest, and we will control it.” And they came to the view, “Okay, then. Well, make sure they don’t embarrass us.” (bronze commander, SYP, quoted by Waddington and King 2007, 422)

Senior representatives of Sheffield Dissent and the Sheffield Stop the War Coalition each argued in interviews that the police decision eventually to allow

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 103

POLICING THE BRITISH G8 PROTESTS

103

organized protest was heavily influenced by the intense political lobbying and media publicity they brought to bear in the buildup to the summit. For example, in one regional newspaper report, the chairperson of the Stop the War Coalition stridently objected to rumors that the police were considering the option of allowing a protest to take place a mile away from one of the ministerial meetings. “It would be an abuse of our civil liberties,” she protested, “and a great shame in a city that has a proud record of standing up for peace and justice” (Yorkshire Post, June 7, 2005, quoted by Waddington and King 2007, 422). Such appeals to longstanding political tradition were difficult for senior police officers to resist. Equally relevant in this respect was the coalition’s wide compositional base (incorporating such groups as the Green Party, local trade unions, the Campaign for Nuclear Disarmament, and the Muslim Association of Great Britain) and the fact that protesters were committed to broad humanitarian causes, rather than their own parochial interests. Interviews with senior police suggest that SYP’s strategy and tactics were further underpinned by an important institutional/organizational factor— namely, an officially documented force mission, formally committing officers of all rank to regaining the public trust and confidence, which had been lost in the wake of the force’s perceived heavy-handed policing of the 1984–1985 miners’ strike and its controversial role in the Hillsborough football stadium disaster of 1989, when dozens of supporters were crushed to death inside the stadium (see Mawby [2002] for a more detailed discussion). The chief constable undoubtedly took these responsibilities seriously: This is why, in fact, I don’t believe in having a national public order police. I take the decisions about public order in the G8 in the light of the fact that I’m still going to be here the day after. I’ve been here the year before, so I’ve had time to meet the community, talk to leaders, get to know the MPs a bit, get to know the councillors, appear in front of them and tell them what I’m gonna do, and make my decisions on the basis of very firm local roots. Because we do remember the legends of the NUM [National Union of Mineworkers] dispute and the stories of the Metropolitan Police and we are all very anxious to put those legends to bed. (quoted by D. Waddington 2007a, 149)

Two final factors—one at the contextual level, the other cultural—also helped shape the force’s strategic and tactical orientations. It is evident from interviews with senior police that prior intelligence based on monitoring public websites suggested that members of anarchist groups were intent on dropping in on Sheffield en route to the main G8 protest in Edinburgh. Police stereotyping of one such anarchist group, the Clandestine Insurgent Rebel Clown Army (CIRCA), suggested that they would deliberately set out in the course of the protest to poke fun at onlooking police officers and lampoon the

06_Madensen_Ch06.qxd:Madensen

104

1/3/11

10:40 AM

Page 104

PREVENTING CROWD VIOLENCE

visiting dignitaries. Such unpredictability was anathema to South Yorkshire Police and underpinned the chief constable’s decision to invoke the Public Order Act 1986 and place corresponding limitations on the right to protest: I did take the opportunity to prevent marches at a certain time and place because I couldn’t separate the sheep from goats in terms of those who were intent on demonstrating lawfully. . . . I set out very clearly in the Gold strategy for Sheffield that we would facilitate the lawful business of the summit— they were entitled to be here and entitled to have their meeting in peace and quiet; we would facilitate the lawful protest, with the emphasis on lawful; we would try and locate venues where people could make their feelings known so that they were close enough to let people know they were there and protesting, but not so they could commit any act of violence. (quoted by D. Waddington 2007a, 148–149)

This determined operational stance had a bearing on developments at the contextual level of analysis. For several years prior to the summit, working relations between senior members of the Stop the War Coalition and police liaison officers had been consistently accommodating in nature. However, on this occasion, the coalition’s chairperson was privately informed by police contacts that any formal application to stage protest events comprising over 100 people would be rejected out of hand. She therefore submitted an official application form in order, as she put it, “to keep the protest alive.” Echoing the type of conclusions reached by Mitchell and Staeheli (2005), Stop the War Coalition colleagues foresaw the likelihood of a self-fulfilling prophecy resulting from the confining of protesters to compounds: “It’s not that the demonstrators would have been looking for this sort of conflict but it’s something that naturally occurs when people see their right to protest being unnecessarily obstructed or denied” (quoted by Waddington and King 2007, 423). Other variables—at the contextual and situational levels—may have served to counterbalance any potential toward disorder resulting from cultural discrepancies between police and public. At the protests taking place on the Wednesday and Thursday evenings, local police numbers were buttressed by reinforcements drawn from the West Yorkshire, Greater Manchester, Lancashire, Northumbria, and Durham constabularies. Members of the South Yorkshire Police public order units were briefed by their commanders of the need to be suitably respectful and take a long-term approach in their dealings with the public: I said, We’ve got to talk to these people because we might be going to jobs at their houses because they live at Nether Edge, Ecclesfield, Fulwood and what have you. And if you talk to them, you might well have a laugh because they’re reasonable people like you and me. Poverty was the bottom line of the protest—that’s why they had the rice meal on the Thursday—and noth-

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 105

POLICING THE BRITISH G8 PROTESTS

105

ing could have been worthier than that. (bronze commander, SYP, quoted by Waddington and King 2007, 425)

Visiting rank-and-file officers received no such preliminary instructions. The chief constable nonetheless instigated a system of “cascade briefings” with local and visiting commanding officers, urging them to do their utmost to facilitate and encourage all genuine forms of protest. Ultimately, the police strategy was used of deploying only SYP officers “in engagement” with the protesters, with outside officers being kept back “on standby” on the periphery of the protest venues. It was the variables outlined at each of these preceding levels that helped to determine not only the subjective readings by police and demonstrators of each others’ motives and intentions, but also the eventual quality of interaction between them (interactional level). In the main, relations between the police and public were cordial and relaxed, with officers refusing to be goaded by complaints or accusations that the right to protest was not adequately being upheld. Such positive relations were especially characteristic of the Saturday march and the main midweek venues, none of which was attended by a significant number of anarchists. By contrast, the two notable instances of disorder to occur involved young anarchists and members of visiting police forces on the periphery of the designated protest locations. The first of these took place on the Wednesday evening when a handful of men belonging to a larger group of 15–20 self-professed anarchists, who had been wandering around the city under the watchful eyes of police “spotters,” suddenly decided to try to penetrate a line of Greater Manchester Police blocking off a side road to the north of the Winter Gardens. In keeping with the predictions of Stop the War Coalition members, the anarchists had grown resentful of the police unwillingness to allow more than 150 into the protest compound: “We never went there to confront the police,” one thirty-one-yearold arrestee subsequently explained. “It was simply that the police were stopping us from getting where we wanted to be” (quoted by D. Waddington 2007a, 154). Police commitment to stopping the anarchists in their tracks suggests a more sinister attribution of their motives. Both the thirty-one-year-old and his sixteen-year-old male colleague were roughly overpowered, the latter by four Greater Manchester officers who succeeded in twisting his arm up his back. This second development falls into the category of an intensifier. According to one onlooker, a male students’ union officer, “It was one of those incidents like a spark which produced a really noticeable change of atmosphere” (quoted by D. Waddington 2007a, 154). As a second eyewitness further explained, Everyone was saying, “Get your hands off him! He’s only sixteen! What are you doing to him? What’s he done?” I think people were extra pissed off

06_Madensen_Ch06.qxd:Madensen

106

1/3/11

10:40 AM

Page 106

PREVENTING CROWD VIOLENCE

because he’s only a kid and looks every inch like a kid! They had him pinned against this wall and that’s when the copper pulled out his baton. He just went mad, you could see it in his eyes. It was a complete abuse of power. (anarchist, male, quoted by D. Waddington 2007a, 154)

The situation briefly escalated when the same police officer struck out at “a slightly drunken man” who had approached him in order to remonstrate. However, the sting was taken out of the situation when the errant officer had his baton confiscated by his senior officer and was whipped out of the firing line. Police in riot gear arrived within seconds but strategically stood their ground without attempting to disperse the growing crowd. Following a brief standoff, tempers subsided and the crowd went on its way. The only obvious repercussion of the incident was that a local activist spat at police officers under what he wrongly supposed to be the cover of the crowd. Unbeknown to him, police officers had photographed his face, enabling their colleagues to identify him for the purpose of arrest as he walked home later that night. At the following evening’s Rice for Dinner event, the prevalent mood was carnivalesque, with the 500 or so people present generally lazing around on the grass verges, eating their rice and listening to speeches and organized music, while police officers casually looked on in twos and threes. Midway through the evening, however, a group of twenty to thirty young anarchists, mostly wearing black and some with masks across their faces, rose to their feet and staged what looked like a sit-down protest in the middle of the nearby road. Members of the group later explained in interview how this had been a spur-of-the-moment response to the realization that this form of protest was futile and that they somehow needed to make more impact on the public consciousness. A spontaneous decision was made to set off down West Street toward the Cutlers’ Hall. At this stage, the participants had no particular objective in mind: “Even the other demonstrators assumed that we had had this ‘radical action plan’ for weeks but it was nothing of the sort” (anarchist, male, quoted by Waddington and King 2007, 426). The suggestion that the wearing of face masks symbolized a sinister underlying intention precipitated only amused denials: “‘It’s laughable! I was masked up but that’s basically because I don’t want the police—I don’t want anybody for that matter—taking my pictures. It’s nothing to do with wanting to look mean or menacing’” (anarchist, male, quoted by Waddington and King 2007, 426). The now eighty-strong contingent of anarchists and fellow-protesters initially set off toward West Street, but suddenly decided to give accompanying police officers the slip by swiftly veering down a side road. There they were met by a cordon of police officers already lying in wait as part of a preformulated contingency plan. Police to the rear then closed up on the marchers and securely hemmed them in. Among those who remained trapped for the next two to three hours were eight members of the local media and a number of

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 107

POLICING THE BRITISH G8 PROTESTS

107

unwitting bystanders—including a female history professor and male English lecturer, who both complained of being bundled to the floor in the course of the police operation. Seven arrests occurred as police snatch squads occasionally plucked out individuals from the crowd. Eyewitness interviewees (including a BBC Radio Sheffield reporter) refuted police claims that the prisoners had been doing anything sufficiently provocative or offensive to have deserved being arrested. The SYP bronze commander countered that the arrestees had all been identified as trouble-causing anarchists by Metropolitan Police intelligence officers—hence the decision to remove them: “In went the snatch squads. That killed it dead— you take off the head and the rest just withers. We knew who they were, even with their faces covered” (bronze commander, quoted by Waddington and King 2007, 427). The incident was successfully concluded when an SYP inspector approached members of the Sheffield-based Rhythms of Resistance (a group of drummers who regularly attend local demonstrations) to first apologize for the police’s unnecessary heavy-handedness, and then ask the band members to lead an orderly march of the protesters back onto the original protest site. It is difficult to imagine that this peaceful conclusion would have been so easily achieved in the absence of this pacificatory gesture.

Understanding Events in Edinburgh Variations in the types of strategy and tactics adopted by the Scottish police in relation to separate protest activities in or around Edinburgh, and in the violent conduct associated with them, are also explicable in terms of key variables embedded in the flashpoints model. Here, as in Sheffield, the structurallybased motive for the protests was an ideological alienation related to issues of peace and the plight of the world poor. A constant overarching feature of the Scottish protests was a particularly negative political/ideological climate in which the Scottish media roundly vilified the protesters and their aims and made dire predictions of impending violence and atrocity: “For example, the tabloid Sunday Mail ran an extensive story in January 2005 under the lurid headline: ‘Ready to Riot G-Hate: Anarchists stage secret talks to plot chaos at Gleneagles,’ accompanied by violent images from Genoa” (Rosie and Gorringe 2009, 44). Media prophecies of this nature were reproduced throughout the spring and early summer where, for example, a Daily Record article continued to provoke fears by asserting how “there are plenty hoping to use the [Make Poverty History] jamboree as an excuse to create as much bloody mayhem as they can. The G8 is the anarchists’ World Cup, their dream date” (Gorringe and Rosie 2006, 4). It might well be anticipated that vilification of this nature would automatically induce an uncompromising police response. However, this would be to

06_Madensen_Ch06.qxd:Madensen

108

1/3/11

10:40 AM

Page 108

PREVENTING CROWD VIOLENCE

discount the importance, in this instance, of crucial intervening variables located at the institutional/organizational level. With regard to the former, it is evident that both a particular philosophical approach to public order policing in Scotland, combined with a formal obligation to engage in public consultation, appears to have offset any negative media influence: Indeed, preceding the summit the notion of a distinctive “Scottish approach to policing”—based on negotiation, consensus and interaction—was prominent in media accounts and our interviews. As senior officers in Tayside Police put it: “Police by consent” is the Scottish way of policing: “Nicey, nicey: let’s not react.” . . . Furthermore Scotland has little recent history of major public disorder. This means that Scotland’s police have no officers routinely deployed in (and thus hardened to) “riot” situations. (Gorringe and Rosie 2008a, 192)

Lothian and Border Police (LBP) were also sensitive to the conclusions of the Scottish Executive’s Review of Marches and Parades in Scotland, which advocated that the authorities should canvass a wide range of local views in deciding whether or not to permit large protest events (Gorringe and Rosie 2008a). Thus, in the twelve months prior to the summit, police posted a community liaison officer in Auchterarder who “engaged in ongoing dialogue” in the course of over sixty meetings with local residents with a view to obtaining their views and, where possible, allaying prominent concerns stoked up by the more sensationalist media reporting (Gorringe and Rosie 2008a). It is therefore evident that there were basic continuities from one protest event to another in terms of key structural, political/ideological, and institutional/organizational factors. On balance, these tended to encourage a tolerant approach to the policing of each of the demonstrations involved. In order to explain variations in the amount and intensity of disorder occurring at each event, it is necessary to highlight important distinguishing variables residing at the political/ideological, cultural, contextual, situational, and interactional levels. Gorringe and Rosie explain, for example, how highly discrepant police cultural definitions of participants involved in the two presummit marches occurring over the weekend of July 2 and 3 were fundamental to explaining the highly differentiated tactical approaches they adopted. The overridingly cordial relations existing between police and protesters on the Make Poverty History march were in marked contrast to the hard-faced police approach displayed toward the Stop the War protesters. This discrepancy was undoubtedly due to the differing degrees of legitimacy accorded to the participants in these marches. The police’s culturally mediated definition of the “good” (well-intentioned and “consensual”) Make Poverty History marchers contrasted with their more cynical view of the “bad” Stop the War movement and its organizers. Relevant “lessons of the past” were undoubtedly important here: as we saw previously, in Sheffield, the Stop

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 109

POLICING THE BRITISH G8 PROTESTS

109

the War Coalition had traditionally enjoyed a good working relationship with South Yorkshire Police. In Scotland, however, three days of disruption of traffic by Stop the War protesters in Edinburgh city-center in 2003 had earned them the reputation of “troublemakers” who could not readily be trusted (Gorringe and Rosie 2008a, 192). These views were influential at the situational and interactional levels. Thus, on the Make Poverty History march, the police demeanor was cordial and low-key. Even when called in to deal with a recalcitrant breakaway group of anarchists, the Lothian and Border Police were consciously restrained out of respect for the objectives of the vast majority of the protest participants and the city’s wider political and civic goals: The LBP confirmed that “serious confrontations” on the outskirts of the march could have resulted in arrests but the police did not want to tarnish the general mood of the day. . . . Had the LBP opted to emphasize disorder, clamp down on “misbehaviour” and make arrests, then media coverage and public opinion might have been swayed and Edinburgh could now feature— to some degree—alongside Genoa, Seattle and other bywords for “violent” global protest. (Gorringe and Rosie 2008b, 700)

On the following day, however, “there seemed to be a much more tangible (and stern-faced) police presence and it was only when it became evident that any repetition of 2003 was unlikely that a marked reduction in police numbers occurred” (Gorringe and Rosie 2008b, 700–701). Key variables at the cultural, contextual, and situational levels all had an equally pronounced influence on the violence that broke out during Tuesday’s (July 5) Carnival for Full Enjoyment. The LBP’s stereotypical conception of the core groups involved as “criminal tourists” or “leisure protesters, there for a ruck” (in contrast to “good” or “consensual” campaigners) highlighted their strong anticipation of confrontation (Gorringe and Rosie 2008b, 699). Adding to this wariness was the fact that, here especially, the police had found it impossible to identify and form a liaison with accredited protest leaders or organizers (Gorringe and Rosie 2008a, 196). Gorringe and Rosie’s account of the event suggests how, in the course of the day, a number of potential flashpoints occurred in a variety of situational contexts, each of which required urgent police definitions of the protesters’ underlying motives and the implementation of a corresponding tactical response. Having initially assembled at an agreed location, the 300–600 protesters involved then split off toward separate destinations, apparently lending credence to preformulated police suppositions of a mass intention to vandalize Edinburgh’s main financial and retail sectors. The police’s general reaction was to use a variety of strategic incapacitation techniques (in this case, corralling and/or “penning-in”), including the deployment of “iron horse” metallic barricades, manned by scores of riot officers. In those instances where

06_Madensen_Ch06.qxd:Madensen

110

1/3/11

10:40 AM

Page 110

PREVENTING CROWD VIOLENCE

numerical strength was overwhelmingly on their side, the police were also able to use powers under Section 60 (S60) of the Criminal Justice and Public Order Act 1994 to corral protesters into confined spaces, detain them for periods of up to five hours, and if necessary, require them to submit personal details about themselves. In such circumstances, initial scuffles subsided as the protesters knew that resistance was futile. Elsewhere, though, as in a major incident on the city’s main Princess Street thoroughfare, protest marchers reacted indignantly to the “message sent” by the “precautionary tactic” of using “hundreds of police in protective equipment” to hem them in from both ends: “As a ‘non-verbal signal,’ these precautions effectively delegitimized the protests and resembled the strategy of ‘total control’” (Gorringe and Rosie 2008b, 703). A spiraling process of “mutual antagonism” quickly developed: Mounted police and officers in riot gear were confronted by protestors— many dressed in black. As the standoff became heated, hooded activists chanted “polizia assassini” (by reference to Genoa) and pressed forward. There was aggression but, initially, no violence. Throughout a two-hour faceoff protestors occupied the road abutting Jenners (Edinburgh’s “classiest” store), Marks and Spencer and other retail outlets but made no attempt to attack them. The lack of a clear protest objective seemingly troubled police officers. Uncertain whether to hold the line or retreat, there were periodic displays of aggression. At 2:45 p.m., for instance, as the chanting crowd pressed up against the police line, mounted officers drew batons and thrust forward, scattering protestors into Princes Street Gardens, but the horses were replaced by shield-bearing and helmeted police. (Gorringe and Rosie 2008b, 703–704)

Gorringe and Rosie (2008b, 705–706) make the point that such antagonism was compounded by an absence of police attempts to communicate, not only with protest marchers, but also with passersby, many of them tourists or shoppers, who found themselves unwittingly caught up in the event. In the absence of any clear rationale for scattering the crowd, many civilians responded by uprooting flowers, bins, and even park benches from Princes Street Gardens and hurling them at the police. Problems arising from negative stereotyping and a lack of preliminary liaison were reproduced in relation to the city-center disorder occurring on the first day of the summit meeting (July 6). Adding to the potential for disorder were numerous contextual variables referred to by Gorringe and Rosie (2008a). Primary among these were the fact that, “uncertain about the aims and objectives of those in the camp, and claiming to have specific ‘intelligence’ of coming trouble, police tried to prevent protestors from leaving” (Gorringe and Rosie 2008a, 196). Following the consequential violence that occurred, police mismanaged the situation to the extent that, while some coaches were allowed to depart from Edinburgh to Auchterarder, others were

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 111

POLICING THE BRITISH G8 PROTESTS

111

prevented from doing so. A heightened sense of relative injustice prevailed, stemming from which protesters spontaneously set off on an impromptu protest march. Sensing perhaps that the march was understandable in the circumstances, the police mood changed from one of impatience to a growing tolerance: Faced by the prospect of more protests in Edinburgh, only days after scenes described as “the battle of Princes Street,” police actions encompassed both toleration and repression. Early on, perhaps hoping to nip the protests in the bud, three self-identified “organizers” who tried to negotiate with the police were arrested. . . . The police’s message seemed clear: “we are not prepared to negotiate.” Thereafter, though, police tactics became less interventionist and concentrated on monitoring the protestors as they embarked on an uncharted march around the city. (Gorringe and Rosie 2008a, 198)

This discernible attitude of ambivalence pervading the policing of the impromptu march was equally detectable in police tactics adopted toward the protest around Auchterarder, which skirted the summit venue of the Gleneagles Hotel. There, as we already know, the traditionally tolerant Scottish policing policy toward public protest, allied to the directive contained in the official review, appeared not only to have overridden the influence of an alarmist media opposition to the planned march, but also to have induced a force policy of consulting with and reassuring local residents. Against this, however, there were at least three countervailing factors—based at the political/ideological, cultural, and contextual levels—compelling the police to be less compromising in their orientation. Uppermost, here, was an exceptional political obligation on the police to guarantee the safety and security of the visiting heads of state and their accompanying delegations. A second crucial (cultural) consideration was the police mistrust of the international groups of hardened “criminal tourists” presumed to be descending on Auchterarder. Finally, the police expectation of and preparedness for trouble was undoubtedly reinforced by such contextual considerations as the lack of preliminary liaison with these “transgressive” groups of protesters and the violence occurring in neighboring Scottish locations earlier that day (Gorringe and Rosie 2008a). These and other factors helped to determine key elements of the immediate situation in which the protest eventually occurred, which influenced, in turn, the nature of police-protester interaction. Thus, on the route of the scheduled march through Auchterarder to the periphery of the summit meeting, police were deployed in a low-profile manner, as part of which they chatted in a decidedly friendly and helpful fashion with passing demonstrators. This contrasted with the extremely forbidding spectacle awaiting the head of the march as it eventually emerged from the village, in the form of a six-foot-high, wiremesh perimeter fence, patrolled by rows of riot officers, standing five-deep,

06_Madensen_Ch06.qxd:Madensen

112

1/3/11

10:40 AM

Page 112

PREVENTING CROWD VIOLENCE

with a further row of mounted police positioned to their rear (D. Waddington 2007a, 1). This symbolic “target of derision” became the immediate focal point for confrontation between the police and the activist core of the march—probably just as the police had planned it. While this disorder was taking place, other protesters were able, much to their surprise, to bear down, unopposed, on an unguarded stretch of fencing to the right of the ensuing battle, which they set about dismantling. It did not take long, however, before a pair of Chinook helicopters and an Army Lynx helicopter brought police reinforcements to the scene. Then, after a brief period of confrontation, police equipped with batons and round, Perspex riot shields succeeded in forming a restrictive arc around the protesters. Mounted police were then deployed to nudge the protesters back onto the official route of the march. A few stones, placards and sods of grass were defiantly thrown in the police’s direction, though not sufficiently to prevent them from re-establishing complete control. (D. Waddington 2007a, 2)

One possibility is that the entire police response to the attempted breach of the wall represented “a hastily improvised measure.” There is a greater likelihood, however, that this constituted a classic piece of strategic incapacitation, “a policy of entrapment, designed to lure protesters away from local housing and into an open field where they could be even more effectively controlled” (D. Waddington 2007a, 160).

The Lessons for Public Order Policing Our two case studies of the British G8 protests emphasize how the theoretical value of the flashpoints model lies in its providing a framework for understanding the ways in which a variety of contextual, interpretive, and interactional variables may possibly combine to produce an enhanced or, alternatively, attenuated potential for disorder. Indeed, by looking at the interrelatedness of, and interplay between, these variables in relation to events in Sheffield and Edinburgh, it has been possible to account for patterns of ambivalence and contradiction in the nature of the police tactics and strategies employed throughout. The model is arguably less useful in attempting to explain the processes by which dissenting members of a crowd become unified and mobilize in opposition to the police once the flashpoint has “ignited” (Otten, Boin, and van der Torre 2001). That is why it is so necessary to consult complementary social psychological approaches, such as the Elaborated Social Identity Model (e.g., Stott and Reicher 1998b), in order to achieve a more comprehensive understanding.

06_Madensen_Ch06.qxd:Madensen

1/3/11

10:40 AM

Page 113

POLICING THE BRITISH G8 PROTESTS

113

The flashpoint model’s practical utility resides in the recommendations it raises with regard to public order policing. While it is obvious that the police are incapable of exercising much discretion or control over factors relating to the structural level of analysis, the preceding case studies have a number of tactical implications residing at the remaining six levels of analysis. As a general principle, it is imperative that the police do their utmost to avoid being perceived by protesters as a “repressive agency of the state.” They are more likely to achieve this objective by striving, at the political/ideological level, to advocate and encourage the public’s “democratic right” to protest. Senior police should therefore endeavor to question, and, where necessary, resist calls by democratic audiences to restrict the freedom to assemble. They should habitually adopt a healthy skepticism in response to media coverage attributing diabolical intentions to participating groups. A sensible balance must be achieved between guaranteeing the safety and security of symbolically significant individuals, monuments, buildings, and occasions, and ensuring that the protesters remain able to achieve their goals. At the institutional level of analysis, senior officers should automatically question whether their force missions or philosophies encourage sufficient flexibility and interpersonal engagement in their relations with the public, or tend to reduce the police capacity to compromise. In the two cases reviewed, such outlooks tended to reduce the potential for disorder. Especially where police mutual aid provision has been invoked, operations should be primarily managed and coordinated by local commanding officers with the requisite knowledge at their disposal. There is a clear overlap, here, with the cultural level of analysis, where steps should be taken to involve local officers at the point of contact with protesters. Both these and visiting officers should receive advance briefings designed to offset misleading stereotypes and encourage greater understanding and tolerance of the protesters’ values and objectives. Police planning and preparation for major political demonstrations should equally be devoted to ensuring that they occur within a favorable communication context that is oriented, as far as circumstances will allow, toward facilitating protest objectives. This should include a commitment to being wary of overgeneralizing or overreacting on the basis of preliminary intelligence gathering. Senior police officers should resist the temptation to warn groups of protesters to behave themselves in case of provocation or misinterpretation. Rather, there should be an accent on liaison and cooperation, such that, even where the police deem it necessary to veto certain protest activities, they are careful to recommend alternative arrangements that are in keeping with the participants’ objectives. Police tactics being implemented at the situational level should be geared toward creating an overriding impression of police impartiality, downplaying their involvement and allaying any possible suspicion or mistrust among the

06_Madensen_Ch06.qxd:Madensen

114

1/3/11

10:40 AM

Page 114

PREVENTING CROWD VIOLENCE

public. This is more liable to be achieved by encouraging the self-stewarding of the crowd, keeping targets of derision out of view, ensuring effective communication and clear lines of command between separate police units, and avoiding the high-profile and provocative deployment of junior ranks. In addition, commanding officers should ensure that open lines of communication are maintained between police and protesters in order to facilitate adequate instruction and explanation. Video monitoring of the crowd should be exercised as covertly as possible, with a view to avoiding any “indiscriminate,” illchosen, and potentially inflammatory police interventions. All of this brings us onto a series of recommendations for police tactics at the crucial interactional level of analysis. Authors (e.g., Reicher et al. 2004) have repeatedly emphasized how police tactics perceived as unfair or discriminatory by the crowd will have the effect of generating unified and determined opposition. For this reason, police interventions must be as accurate and judiciously implemented as possible. It is therefore imperative for the crowd to receive clearly audible and repeated notification of impending police activities. The police should likewise ensure that their actions are not undertaken too suddenly or abruptly, or without due concern for the safety of those present. Consideration should also be given to the possibility of delaying the making of arrests until later in the day in order to avoid creating a wider confrontation. Where unavoidable, police interventions should be as swift and decisive as possible, the aim always being to avoid trapping the crowd in a confined area, without dispersing them so widely that the participants become virtually unmanageable. As a general rule of thumb, the police should strive at all times to accommodate the protesters’ perceived rights and objectives, and ensure strict adherence to any prior formal or informal basis of agreement.

Note I wish to acknowledge the assistance of my colleague, Kerry McSeveny, who helped with the formatting and editing of this chapter.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:37 AM

Page 115

7 Crowd-Related Crime: An Environmental Criminological Perspective Tamara D. Madensen and John E. Eck

ENVIRONMENTAL CRIMINOLOGY IS THE STUDY OF CRIME EVENTS, their patterns, and how to prevent them. To date the vast majority of environmental criminological studies have focused on common “street” crimes—such as homicide, robbery, burglary, and similar offenses. There is another growing literature focusing on other forms of crime, such as terrorism (Clarke and Newman 2006), child sexual assault (Wortley and Smallbone 2006), and Internet fraud (Newman and Clarke 2003). A characteristic common to all these crimes is that the offenders are acting alone or in small groups. When the crimes occur in large groups, the crowd is largely a passive context of the broader, immediate environment. Passengers waiting for trains in a congested commuter-rail station may become victims of pickpockets, for example (see Felson 1996). In these circumstances, the crowd provides a source of crime targets and a context within which the offender can remain hidden. Studies of crimes committed by relatively large numbers of individuals who use their numbers to commit crimes are far less common. Björ, Knutsson, and Kühlhorn (1992) provide a notable counterexample with their study of disturbances created by a Midsummer Night’s Eve celebratory gathering and how these disturbances were prevented. Examples of such crowd-related problems include political protests where members of the crowd attempt to breach security cordons around government buildings; fights between fans of rival sports teams before, during, or after matches; conflicts between proponents of rival political positions; student drinking party riots; clashes with police following police actions; and industrial strikers attempting to enter a factory from which they have been barred by management. The complexities of crowd-related crimes present a unique challenge to environmental criminologists. Yet, if environmental criminology has the universal scope its proponents assert—capable of explaining any form of crime 115

07_Madensen_Ch07.qxd:Madensen

116

1/3/11

10:37 AM

Page 116

PREVENTING CROWD VIOLENCE

and useful for suggesting a range of effective prevention measures for all crimes—then environmental criminologists need to consider how their theories can be applied to these problems. The objective of this chapter is to show how standard assumptions of environmental criminology would apply to the study of crowd-related crimes and the development of solution methods. We will begin by addressing what forms of crime we are concerned with. We will pay particular attention to the concept of crowd and behaviors of crowd members that cause harm to others. In the second section, we will use standard criminology as a foil for showing how environmental criminologists think about crime. We will then show how basic principles of environmental criminology can be applied to crowds. In the third section, we will look at basic theoretical perspectives that comprise environmental criminology and show how each describes different aspects of crowds and their members. These perspectives are rational choice, routine activity theory, crime pattern theory, and situational crime prevention. The fourth and final section draws implications for further study of crowd-related crimes and the development of prevention methods.

Is It the Crowd, or Is It the Behavior? To understand how environmental criminologists approach crowd-related crime, we need to clarify what we mean by “crowd-related crime.” There are two parts to this concept: the crowd and the unwanted behaviors that may be committed by members of the crowd. Should environmental criminologists be concerned about preventing crowds, or preventing the behaviors? The short answer is that behaviors are the focus of concern. Simply put, crowds are not the problem. The problem arises from behaviors of some participants in the crowd that cause harm (either to other crowd members or to people outside the crowd). In this sense, crowd-related crimes are treated like other crimes, such as the pickpocketing example used above. Our concern is not that commuters are gathered together in large numbers waiting for a train. Our concern is that some individuals in this regular gathering are stealing from members of the crowd. Similarly, our concern is not the crowd gathered to protest a police shooting, but the throwing of objects at police, or the breaking of shop windows, or the theft of goods from these shops. That is the simple answer. But all simple answers to complex questions are misleading. Crowds make some illegal behaviors easier to commit than they would be without the crowd. Crowds can lower the risk of some offenses. Crowds can encourage possible offenders. The presence of crowds makes some crimes more rewarding than they would be if no one else were present. And members of crowds can provide excuses for misbehavior that give potential offenders “permission” to act in harmful ways. So in some circumstances,

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:37 AM

Page 117

CROWD-RELATED CRIME

117

crowds are instrumental in facilitating misbehavior. And in these circumstances, crowds create an opportunity for crime that would not occur otherwise. Environmental criminologists focus on crime opportunities. We will say more about this later. But for now it is sufficient to say that environmental criminologists will be interested in ways of preventing crowd formations in order to prevent crimes they may help facilitate. We can use a theft reduction example to show how this might work. While investigating street robbery patterns and prevention strategies, Barry Poyner found that some robberies were of older men boarding buses at crowded bus stops. The thieves would enter the crowd, disrupt the boarding process and thus create a distraction, and in the confusion steal the wallets of older male patrons (Poyner 1986). Poyner’s solution was to create simple barricades at selected bus stops. Patrons were then forced to line up behind the marshaling barrier in an orderly queue. This made it easier for patrons to detect attempts to disrupt boarding. And the barricade helped keep offenders out. In this example, the crowd was part of the problem in that it facilitated the offending. Poyner’s solution was to change the organization of the crowd—in essence change it from a mass of people to an orderly line. In a related example, police in Baltimore County (Maryland, United States) addressing fights between youths awaiting buses to rival high schools had the bus company change the pick-up points so the rival groups would not meet (John Eck, personal observation). Again, the nature of the gathering was changed. Johannes Knutsson and colleagues describe how police altered the conditions for an annual gathering in order to reduce disturbances associated with the Midsummer’s Night Eve gathering by deflecting high-risk offenders (Björ et al. 1992). After two years of drunken disturbances from a block party, police and university officials in Cincinnati (Ohio, United States) prevented the party from forming (Madensen, Fisher, and Eck 2004). While harmful behaviors were the concern in all these examples, the prevention efforts were directed at the crowd. How police and others alter crowds to prevent violence and other problems is a strategic choice. In the Cincinnati example, students had no constitutional rights to gather to drink in public. Indeed, many of the participants were not old enough to drink legally in Ohio. Preventing the crowd to prevent disturbances was an option. If the students routinely gathered to protest (say university restrictions on drinking), then banning the gathering would not be an option. In some circumstances, attempting to prevent a crowd from forming is likely to make things worse: the crowd is likely to form anyway but under less police control. Therefore, other actions directed at manipulating members of the crowd may be more successful at preventing problems.1 In sum, our concern is with harmful behaviors associated with crowds. Environmental criminologists will seek to understand how these behaviors arise, and what can be done to prevent them. One set of options is to focus on the crowd itself or on particular ways the crowd behaves. Crowds are not

07_Madensen_Ch07.qxd:Madensen

118

1/3/11

10:37 AM

Page 118

PREVENTING CROWD VIOLENCE

inherently problematic. Most crowds cause little or no trouble (McPhail 1991). At the extreme, preventing crowd formation is one possible way to prevent crime. However, this approach is often an undesirable, infeasible, or impractical method of preventing most crowd-related crime. Professional tennis matches, concerts, and parades are just a few events that require large groups of people to gather together. Crowds are a natural outcome and necessary feature of many common activities. The right of people to gather and voice objections to government or business policies is cherished in democracies. Even in circumstances where a crowd serves no legitimate political, economic, or social function and its presence is potentially harmful, preventing its formation may not be possible. As such, environmental criminologists will more often focus on identifying specific characteristics of particular crowds that can be manipulated to reduce the likelihood of trouble.

How Would Environmental Criminologists Examine Crowds? Environmental criminologists have a distinct perspective on crime and its prevention. This perspective has important implications for how environmental criminologists would study crowd-related crime and develop prevention strategies. We have briefly introduced some elements of this perspective in the preceding sections. Here we will more explicitly examine seven elements by contrasting standard criminological views with those of environmental criminologists (Clarke 2004). We will be painting with a broad brush here— within standard and environmental criminology there is a diversity of interest, approaches, and opinions, and there are hybrid perspectives that take from both. Nevertheless, this contrast is useful for it highlights core ideas within environmental criminology that have important implications for crowd-related crime. Goals

The first thing to consider is the explanatory goal of each perspective. Standard criminology is interested, primarily, in explaining why some people are engaged in crime and others are not. Theories of criminality look at a variety of possible causes, ranging from biological, to psychological, to social, economic, and political. The theories may be at the micro level—addressing causes of individual criminal involvement—or macro level—addressing why some areas have more offenders than others. In contrast, environmental criminology has far less interest in explaining people’s disposition toward crime. Its primary goals are to explain how crimes are committed, how they form patterns, and to develop and test methods to prevent crime and patterns. It is a perspective centered on understanding events.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:37 AM

Page 119

CROWD-RELATED CRIME

119

Consequently, environmental criminologists will not seek answers to crowd-related violence by focusing on the disposition of crowd members. Participants in crowds almost certainly come to the situation with different interests and dispositions, but these are insufficient explanatory factors for environmental criminologists attempting to address crowd-related problems. Instead, they will look at the immediate contexts of individuals within crowds and how these influence individual decisions. Since decisions of one person may change the crowd dynamics and influence decisions of others nearby, the behavior of a crowd can be an emergent outcome of decisions in the sense that no one need orchestrate the crowd and the outcome can be something that none of the participants anticipated or sought. Since individual decisions will be influenced by both social and physical situational factors, environmental criminologists will attempt to manipulate these characteristics to reduce motivation to engage in harmful behaviors. Motives

Standard criminologists and environmental criminologists use the term “motive” in two different ways. For the standard criminologist, “motive” often means some deep-seated propensity, drive, or orientation. Motives are at least semistable, in that, if they change, they change slowly. Further, the motives are formed long before any criminal act is carried out. Environmental criminologists discount the influence of longstanding propensities and instead focus on more mundane motives that drive everyone to varying extents: acquisition of resources, self-preservation, power, status, and so forth. The immediate situation heavily influences these motives by providing cues that a particular action is likely to be productive and within tolerable limits for risk and effort. Some situations may even provoke or otherwise encourage particular actions. Environmental criminologists will take the same approach to crowdrelated crimes. The motives that matter are defined in the immediate situation rather than in deep-seated propensities originating long before people joined a crowd. Since the crowd itself is a large part of the context within which these people will act, members of a crowd are influenced by the social and physical features of the gathering that offer opportunities for action. Opportunities

Environmental criminologists differ considerably with standard criminologists over the role of opportunities for crime. Traditionally, standard criminologists ignored opportunities for committing crime, usually on the grounds that opportunities were so common that they could not explain variations in criminality.

07_Madensen_Ch07.qxd:Madensen

120

1/3/11

10:37 AM

Page 120

PREVENTING CROWD VIOLENCE

In short, opportunities were a constant and so could be ignored (Gottfredson and Hirschi 1990). As suggested earlier, opportunities are so important to environmental criminologists that their theories are often called “opportunity theories.” As we will see in the next section, most of the basic theories within environmental criminology explain how opportunities arise and vary, and how offenders recognize opportunities and exploit them. Theories of prevention explain how opportunities for crime can be altered. As Marcus Felson and Ronald Clarke (1998) state, “opportunity makes the thief.” Opportunities are to environmental criminology what propensities are for standard criminology. Environmental criminologists would be interested in how crowds provide opportunities for some people to commit crimes. More crime will occur when opportunities are greater and more widespread in a crowd. The fewer and more limited the opportunities in a crowd, the less crime. So, environmental criminologists will be interested in studying the “opportunity structures” of crowds (Clarke 1995). These should vary from crowd to crowd, within a single crowd, and over time within a single crowd. And as implied previously, the crowd itself is part of that opportunity structure. An individual will likely make different decisions in a crowd than outside of it, and crowds with different features will likely lead to different decisions by the same person. In a small gathering, all participants may have the same perception of the opportunity structure partially created by the crowd. In very large crowds, individuals in different parts of the gathering may perceive very different opportunity structures. In a political demonstration, for example, those participants on the edge can see the police, whereas those in the center may not be able to see the police, though they may be aware of their presence. These differences in perception will influence how participants in different parts of the crowd will behave. Though participants’ perceptions and knowledge are local, they can be augmented by communications. Before the advent of inexpensive electronic communications, information (rumors) was spread by word of mouth. Loud hailers and speakers expanded the distance information could travel in a crowd before it had to be repeated by a participant. Now cell phones providing voice, text, or image can spread information faster, further, and with greater fidelity within a crowd. Nevertheless, face-to-face communication and other direct visual cues probably have greater power over individuals’ perceptions than indirect communications. So, local situations will necessarily influence individuals’ perceptions of opportunities. Choice

According to standard criminology, choice is largely irrelevant for understanding criminality. If drives, propensities, and similar factors predict who will and who will not commit crime, then choice is simply an outcome of these forces and can be safely ignored.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:37 AM

Page 121

CROWD-RELATED CRIME

121

Environmental criminologists believe choice is important. This follows directly from their assumptions about opportunities. Environmental criminologists generally believe offenders are rational, though not in the extreme form of classical economics. Rather offenders are no more or less rational than other people; they use standing decision rules and other shortcuts to make fast and frugal decisions in an uncertain world (Cornish and Clarke 1986; Gigerenzer and Todd 1999). But at the core of the decision are attempts to improve the offender’s position. That is, people are more likely to commit crimes if they perceive they will get something beneficial from the effort, that the risks and other costs are not extreme and are manageable, and that relevant others will not reject them. When people perceive that the rewards are low, risk and other costs are high, and that others disapprove of the actions, people will commit fewer crimes. Within crowds, individuals will not lose this basic form of rationality. They will make decisions based on how they perceive the immediate situation. The opportunity structure of the immediate situation will suggest what sorts of actions are more or less rewarding, risky, costly, and approved. Consistent with modern research on crowds (see Reicher, Chapter 2, and Stott, Chapter 3, in this volume), environmental criminologists will not assume that crowds turn normally rational people into irrational people. Though individuals maintain their rationality, it does not make sense to describe the crowd itself as rational. As with many social phenomena, the collective outcome of many individual decisions is not necessarily what the individuals intended or desired (Schelling 1978). It may be just the opposite. Consider a large number of people in an enclosed space when individuals perceive a threat. Each person may seek to leave through the nearest exit. However, if the exits are few and small, these individual rational decisions may result in a jam at the exits that prevents people from quickly leaving. Seen as a whole, the event appears senseless; yet each individual, operating with the knowledge they had available, made a sensible decision. The larger the crowd, the greater the differences between the various locally based perceptions within it (assuming no one is making a great effort to communicate the same message throughout the crowd). And this means that there is likely to be greater variation in the choices made by people in the crowd. This will make it even more likely that rational decisions based on local knowledge will not lead to an overall (global) action by the crowd that seems rational to an outside observer (including the police). In short, because the global actions of a crowd may bear little resemblance to the choices (or desires) of the individual participants, we cannot judge individual rationality (based on local knowledge) by the global behavior of the crowd. Actors

Standard criminology, as we stated earlier, focuses almost exclusively on one actor: the offender. Others, to the extent that they are important, contribute to

07_Madensen_Ch07.qxd:Madensen

122

1/3/11

10:37 AM

Page 122

PREVENTING CROWD VIOLENCE

or detract from the offender’s drive to committing misdeeds. Parents, teachers, and neighbors, for example, influence the offender’s propensity to engage in crime, but within standard criminological theories are seldom present when the crime takes place. There are multiple interacting actors within environmental criminology. The offender may be the lead actor in that it is her choices that ultimately result in a crime event, but like an improvisational performance, these choices are heavily influenced by other actors on the stage. And in turn, the offender influences these other factors. We will discuss this in greater detail when we discuss routine activity theory later in this chapter. Within crowds, environmental criminologists would look at how offenders make choices based on the actions of the people around them, including the police. In fact, prior to the commission of a crime, there may be little to differentiate nonoffenders from offenders in a crowd. Specificity

Environmental criminologists differ from standard criminologists in their requirement that crimes are differentiated in some detail. Standard criminologists tend to be more interested in classifications of offenders, though there is some interest in the question of whether offenders are generalists or specialists. Environmental criminologists tend to emphasize small differences among crimes because these differences often indicate that they require different opportunity structures. For example, though theft of vehicles and theft from vehicles both involve theft and vehicles, the process necessary to steal a car is very different from the process of taking things from the same vehicle (Poyner 1991). Retail sales of drugs in an open market are different from retail sales of drugs in a closed market (Eck 1994). Further, if two crimes have different opportunity structures then the measures necessary to prevent each crime will be different. Environmental criminologists, therefore, will be interested in different types of crowd-related crimes (e.g., throwing missiles at the police, breaking shop windows, or attacking members of an opposing crowd). Further, differentiation among crowds and gathering settings is likely to be important. The structure of leadership, participant objectives, and the characteristics of the gathering location will influence the behavior of crowd members. A formally organized crowd of laborers protesting the closing of a factory or mine along a busy highway presents a different set of opportunity structures than an informal meeting of college students gathered to drink and party in a dormitory, for example. Similarly, these crowds are likely to have different opportunity structures than a fractious crowd consisting of various distinct groups gathered to protest a conference of international economic leaders.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:37 AM

Page 123

CROWD-RELATED CRIME

123

Process

Standard criminologists are more interested in “why” crimes occur, whereas environmental criminologists are more interested in “how” crimes occur. Standard criminologists have limited interest in the processes that offenders (or other actors) need to go through to successfully carry out a crime. “Why” is answered by the offender’s disposition. Environmental criminologists, not being all that interested in dispositions of people, are very interested in the steps offenders (and others) take leading up to, during, and after a crime. Each step in a crime process requires a set of opportunities and it is possible to find interventions for multiple steps designed to interrupt a crime process, thus preventing it. Consequently, environmental criminologists examining crowd-related crimes will be interested in how offenders in crowds commit their crimes. They will want to know about actions and decisions that precede the crimes, occur during the crimes, and follow the crimes. This will reveal possible prevention actions that could keep offenders from engaging in criminal behavior. For example, for demonstrators in a crowd of political protesters to take over a building adjacent to the protest march, what steps need to be taken beforehand? Do they need to deceive the police as to their intentions? Do they need any special tools? And so forth. Once inside, what actions will the demonstrators need to take so they are not immediately removed? Are these easy to accomplish or do they require special preparation or training? These and many similar questions give some insight as to how to prevent entry to the building and allow the march to proceed in an orderly way. These differences between standard criminology and environmental criminology are summarized in Table 7.1, along with the implications of the environmental criminological approach for examining crowd-related crimes and their prevention. Environmental criminology is more than a simple set of assumptions about crime; it is a set of interrelated theories, both descriptive and applied. These theories provide more detailed frameworks for examining crowd-related crime.

How Do Environmental Criminological Theories Inform Our Understanding of Crowd Problems and Prevention? Five theories form the core of environmental criminology: routine activity theory, crime pattern theory, rational choice perspective, situational crime prevention, and problem-oriented policing. The first three seek to describe how crimes are committed and how crime patterns form. The last two describe how crime can be prevented. These theories expand on the seven basic principles

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:37 AM

Page 124

124 Table 7.1

Crowd-Relevant Differences Between Standard and Environmental Criminology Perspectives of

Topic

Standard Criminology

Environmental Criminology

Environmental Approach to Crowd-Related Crime

Goals

Explain why some people commit crimes and others do not, why some areas have more offenders or crimes compared to other areas, and what can be done to reduce criminality

Explain how crimes are committed, under what circumstances, how they form patterns, and what can be done to prevent crimes

More interested in the circumstances created by the crowd and surrounding it that lead to violence than in long-term propensities of crowd members to engage in violence

Motives

Motives are longstanding propensities

Motives are immediate goals and objectives heavily influenced by proximate social and physical circumstances

Focus on the immediate objectives of crowd members and discount any deepseated longstanding propensities

Opportunities

Opportunities to commit crime are assumed to be ubiquitous and therefore of little importance relative to propensities

Opportunities to commit crime are highly variable and are influential on choices people make to engage in crime

Focus on the opportunities for crowds to form and for members of crowds to engage in misbehavior

Choice

Offenders are assumed to be driven by either outside forces or by biological or psychological propensities to be offenders; choice is not considered

People choose to commit crimes because they see some relative advantage in doing so compared to alternatives; they employ bounded rationality

Assume crowd members make individual decisions whether to engage in violence and are not driven to violence by outside factors

Actors

The offender is the primary concern; other actors are of limited importance, generally

The offender is one of several lead actors and the interactions of these actors are important

Interested in the relationships among crowd members, and crowd members and people outside the crowd

Specificity

Little specificity of crime types but often highly specific about offender types

Much specificity of crime types but usually little specificity about offender types

Differentiate among different types of crowds, subgroups within crowds, and different forms of violence

Process

The sequence of actions necessary to successfully complete a crime is not relevant

Process matters and varies across different forms of crowd misbehavior

Concerned about activities before, during, and after crowd violence

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 125

CROWD-RELATED CRIME

125

described in the previous section. In this section, we will provide a brief overview of each theory and show how it can be adapted to understanding or preventing crowd-related crime. Routine Activity Theory

Routine activity theory defines the conditions necessary for a crime, the conditions under which a crime is likely, and how these situations form. A crime is likely when a ready offender comes together with a desirable target at the same place and time, but in the absence of an effective controller (Cohen and Felson 1979; Felson 1986; Eck 1994). Offender, target, and place form the inner triangle of Figure 7.1. Each of these elements is necessary, so the absence of any one is sufficient to prevent crime. The outer triangle describes controllers. Each controller protects an inner element. Handlers attempt to keep potential offenders out of trouble. Examples include parents, siblings, friends, and others whose opinion the potential offender cares about. In a crowd, the most obvious handlers are other members of the crowd, though some members are stronger handlers than others— friends of would-be offenders in the crowd probably matter more than nearby strangers, and nearby strangers are probably more capable handlers than distant strangers. Crowd marshals and other members of the crowd who volunteer to keep order are acting as handlers.

em m d ow

rs be s an m em di ar w d G u c ro tc . nd ts , e g e in g s ea r lic T a u ild b le ,

op

Cr

Po Pe

rs H a n ,m d a r le r s Ge sh ne a ls ra O ,a lly ffe nd fe w n d vo c ro e r s lu n wd te er me s mb er s

The Routine Activities Elements of Crowd-Related Crimes

be

Figure 7.1

Violence or other crowd-related crimes Places Location of crowd

Managers Local government and property owners

07_Madensen_Ch07.qxd:Madensen

126

1/3/11

10:38 AM

Page 126

PREVENTING CROWD VIOLENCE

Guardians protect targets. Targets can be people, animals, buildings, or things. Guardians are always people, though they can use technology or animals to aid them. In most circumstances, guardianship is provided by citizens as part of their normal routines. Security guards are also a common form of guardianship. Police also provide guardianship, but typically they are too few and widely spaced to provide a substantial level of guardianship in most cases. However, with crowds, police are often the most prominent form of guardianship: keeping possible offenders separated from potential targets. Members of the crowd probably provide some form of guardianship. This is most obvious when two opposing groups come together without sufficient police. If members of one group feel threatened by the other group—or in some cases, the police—they will act to protect each other. This may produce counterguardianship. Under some conditions, these actions escalate into physical conflict. Finally, we get to places. A crowd must occupy space, and this space is controlled by some entity, often a government. The owner of the space is the manager, as are those the owner employs to see to the functioning of the place. Place managers attempt to control who can enter their space, the behaviors that are permitted, and the physical layout of the location. Management can overlap with guardianship in that managers often make provisions for guardianship. But management includes far more than protection. With regard to crowds, managers can alter the physical setting by adding barriers or removing obstructions and possible weapons (such as decorations that can be thrown). They can provide services that help members of the crowd have a good time or services that excite members of the crowd to bad behavior. The sale or allowance of alcohol can have both functions, and how it is dispensed makes a difference (see Chapter 9 by Joel Plant and Michael Scott). The movement of five of the elements (all but places) brings them together and separates them. People’s movements are seldom random. Instead, they follow routines. It is the routines of everyday life that bring offenders into contact with targets at particular places. The scheduling of controller routines is often designed to ensure controllers are present at locations where offenders and targets are likely to meet. Some crowds are created routinely: crowds around mass transit sites, for example, or crowds of students on university campuses, or crowds of young adults on Friday and Saturday nights in an entertainment district. Other crowds are scheduled or annual events: planned protest marches, and crowds around sporting events and in stadiums, for example (see Chapter 8 by William Sousa and Tamara Madensen related to crowd management on New Year’s Eve in Las Vegas). Some crowds are impromptu, such as the gathering of people around the sight of a police crime scene or shooting. How the crowd gathers and moves will rearrange the six elements described above to create opportunities for offenders or to block opportunities. These dynamic arrangements will structure the pattern of crowd-related crime.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 127

CROWD-RELATED CRIME

127

Crime Pattern Theory

Crime pattern theory describes how offenders search for targets (Brantingham and Brantingham 1981). Like routine activity theory, crime pattern theory suggests crimes occur when offenders and targets are colocated. The basic framework is that offenders follow routines, traveling from various locations to other locations using standard routes, much like nonoffenders move from home to work to shopping and so forth. The network of routes and destinations forms a structure for searching for targets by defining the areas offenders are aware of. When the awareness space of offenders encompasses targets, crimes can occur. When targets are outside the awareness space, crimes cannot occur. Often targets themselves are clustered, so crime concentrations occur when a target area overlaps the awareness space of offenders. Areas with little or no crime—the majority of any region—either have few or no offenders, or few or no targets, or both. This takes place on a backcloth of street patterns, topography, property ownership patterns, zoning, and a host of other factors that give structure to the offenders’ choices of routes and locations. Crowd formation and movement is also likely to be structured by the environmental backcloth that concentrates or diffuses people. The distribution of authority symbols—monuments, government buildings, corporate headquarters, and so forth—can orient crowd members. This will be important for crowds formed to promote social, economic, and political causes. The location of sports facilities vis-à-vis parking areas, drinking establishments, and road networks structures crowds focused on team sports. Consequently, some small areas of cities will repeatedly have crowds and most others will not. The size of streets will influence the movement of crowds, just as they do vehicle traffic. And these backcloth factors will influence when and where offenders in a crowd come into contact with their targets. Left untended, if crowds repeatedly form, the location of crowd-related crime is likely to be predictable. Rational Choice Perspective

As emphasized earlier, environmental criminologists assume a modestly rational offender—that is, an offender who generally is no less rational than nonoffenders. Clarke (2008) has suggested that there are five basic things that influence whether someone will offend: risk, effort, reward, excuses, and provocations. Risk refers to the ability of the offender to get away with the act. Less risk encourages offending and more risk decreases it. Effort refers to the energy needed to carry out the action. Offenses that are easy will be committed more often than those that are difficult. Reward refers to the gain to the offender: money, power, status, sex, and so forth. More-rewarding crimes will be committed more often than less-rewarding crimes. If an offense is easily excusable,

07_Madensen_Ch07.qxd:Madensen

128

1/3/11

10:38 AM

Page 128

PREVENTING CROWD VIOLENCE

either to the offender or to people whose opinion the offender values, then he is more likely to commit it than if it is not excusable. Finally, provocations influence choices. People are more likely to commit crimes if provoked to do so than if not. Situational crime prevention makes use of these influences, as we will see below. There is considerable agreement by those who study crowds that people in crowds behave rationally: crowds do not cause people to lose their reasoning ability (McPhail 1991). Consequently, environmental criminological approaches to reducing crowd-related crime are likely to be relevant. The rational choice perspective makes a distinction between the general decision to be willing to engage in crime and decisions about committing a specific crime. Both choices are rationally undertaken, but the criteria may be very different. A person may decide to join a crowd and use this as an opportunity to engage in criminal behavior. But there will be another decision as to whether to commit a crime at a specific moment in a specific situation. Finally, the rational choice perspective suggests that offenders often use “scripts” to guide their actions (Cornish 1994). A script is a predefined set of actions. Most crimes require a set of choices to be made. Once an offender knows how to commit a crime (either by experience or by being taught), they do not have to make a series of choices. Instead they follow a “canned” pattern of actions. Scripts require less information to carry out a potential crime so offenders can operate faster than if each step needed to be investigated before deciding to proceed. If an offender has to coordinate his activities with others, then shared scripts reduce coordination costs and uncertainty as to how others will act. Offenders in crowds probably use scripts, particularly if they repeat their actions (either in the same crowd or in several crowds). Moreorganized offenders may explicitly create scripts in advance. Less-organized offenders may modify scripts developed from other situations. At a more macro level, scripts are processes. C. McPhail (1991) suggested a three-stage process for understanding crowds: the gathering process, the completed gathering, and the dispersal process. In an earlier work (Madensen and Eck 2006), we suggested that for some crowds, particularly those planned long in advance, there may be earlier stages that need to be considered: the initial planning stage and the preassembly preparation stage. These stages are presented in Figure 7.2. Decisions made at each stage of the crowd process will create opportunity structures for particular types of behaviors. Table 7.2 lists the five stages along with examples of individual decisions that could increase the likelihood of crowd-related crime. An example of an “event manager” decision (e.g., external organizer, place manager, police officer) and an example of a “crowd participant” decision are provided for each stage to demonstrate how both groups affect crowd dynamics.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 129

129 Figure 7.2

The McPhail-Madensen Model of the Life Cycle of Gatherings

Initial Planning

This includes decisions to create a gathering, location, and other organization efforts.

Madensen’s Additions Preassembly Preparation

Assembly Assembly Process Process

This includes activities at and near the assembly site and the dissemination of information just prior to the arrival of participants.

At this stage, participants are coming to the assembly site. It begins with the arrival of the first participants and concludes when most participants have assembled.

Assembled Assembled Gathering Gathering

Most of the participants have arrived at this stage and if there are planned culminating activities, they occur here.

Dispersal Dispersal Process Process

Participants are leaving the assembly location. This stage begins sometime after the gathering peaks and concludes after most participants leave.

Table 7.2

McPhail’s Original Model

Decisions That Increase the Likelihood of Crime in Five-Stage Crowd Process Decision

Crowd Process Initial planning

Event Manager Select a location without necessary facilities for the event

Crowd Participant Identify and correspond with violent groups planning to attend the event

Preassembly Oversell tickets to the event preparation

Hide a dangerous weapon in a personal backpack

Assembly process

Fail to provide adequate parking

Attempt to circumvent established procedures for entry

Assembled gathering

Promote overintoxication by failing to introduce alcohol sale restrictions

Throw missiles at performers, other crowd members, or security

Dispersal process

Fail to establish multiple exit points

Drive aggressively or while intoxicated through exiting traffic

07_Madensen_Ch07.qxd:Madensen

130

1/3/11

10:38 AM

Page 130

PREVENTING CROWD VIOLENCE

Situational Crime Prevention

Situational crime prevention is the application of the rational choice perspective to prevent crime. At its most basic level, it seeks to do one or more things: increase risk, increase effort, decrease reward, decrease excuses, or decrease provocations. There are multiple ways of doing each of these. Derek Cornish and Ronald Clarke (2003) created a typology of twenty-five different situational prevention categories to organize and highlight the variety of factors influencing offender choices. To select the appropriate techniques, careful analysis of a particular crime problem must be conducted to understand how the crime of interest is committed and under what circumstances. Based on this knowledge, one or more situational approaches that appear to fit the problem are identified and implemented. Situational crime prevention generally resists broad application of the same approach to vaguely similar crime problems. Small details in a crime problem can make any particular technique highly effective or ineffective. Prevention needs to be tailored to the specific problem, based on the analysis and what has been learned from evaluations of earlier applications. This approach has been used to develop guides to assist police in handling two types of crowds: college students drinking outside in large groups (Madensen and Eck 2006) and stadium violence (Madensen and Eck 2008). These guides combine situational crime prevention with a five-stage process model to show how situational crime prevention can be applied at all stages of a crowd, from the earliest planning stage through dispersal. Ideas from this guide were used by the Madison Police Department in its attempts to control student drinking-related problems at an annual Halloween gathering (see Chapter 9). Table 7.3 illustrates the application of the situational crime prevention framework in tandem with the crowd process stages. Examples of interventions event planners may use to reduce the potential for crowd-related crime at a specific event are placed within the resulting matrix. These interventions are categorized based on the crowd process stage in which they are likely to occur. They are also categorized based on the dimension of opportunity (risk, effort, reward, excuses, or provocations) most likely to be affected by the interventions. Problem-Oriented Policing

Herman Goldstein (1979) created problem-oriented policing to improve policing. He did not have in mind the theories we have discussed. Goldstein noted that police tend to focus on the means of policing and tend to forget about the goals they are supposed to achieve. Law enforcement, he claimed, is one of many possible methods for assisting the public, but it is often seen as the ends. Goldstein sought to reverse this.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 131

CROWD-RELATED CRIME

Table 7.3

131

The Application of Situational Crime Prevention to Crowd-Related Crime Situational Approaches to Crime Prevention

Crowd Process

Increase Risk

Initial planning Preassembly preparation

Increase Effort

Dispersal process

Decrease Excuses

Require gathering permits Provide adequate facilities Deny entry to intoxicated persons Videotape gathering

Decrease Provocations Set capacity limitations

Hire and train security staff

Assembly process

Assembled gathering

Decrease Reward

Set behavioral expectations; post rules

Install barriers to control movement Require early dispersal for crowd agitators

Allow multiple exit points to avoid trampling

He suggested that the appropriate unit of police work is not specific incidents or events, but rather clusters of events, which he called “problems.” Rather than rely on one way to solve problems, like enforcing the law, Goldstein encouraged police to explore many alternative approaches. To do this, police were to identify problems, conduct an analysis of each, develop solutions, and then determine if those solutions worked (Goldstein 1990). In this regard, it bears some resemblance to the approach advocated by situational crime prevention. Though this was a comprehensive theory of what the police should do, it lacked a theory of problems that could assist police in carrying it out. Environmental criminology, on the other hand, provided a set of theories about problems, but did not have a theory of the institution that could put these theories to use. These ideas were coupled in the early 1990s. Today, when police are taught how to handle problems, they are taught the principles of environmental criminology. Problem-oriented policing is typically aligned with community policing, which calls for greater contact between the police and the community to build trust, exchange information, and prioritize action. Often community policing and problem-oriented policing are used interchangeably. There is a subtle differ-

07_Madensen_Ch07.qxd:Madensen

132

1/3/11

10:38 AM

Page 132

PREVENTING CROWD VIOLENCE

ence in that community policing often treats community engagement as an end in itself, while problem-oriented policing treats community engagement as a means to assist in reducing problems. Problem-oriented policing and community policing are particularly useful for addressing problematic aspects of crowds because they require careful examination of the nature of the problem being addressed, the development of highly tailored solutions, and engagement with members of the crowd, often long before the crowd has formed. These perspectives advocate alternative and creative approaches to harm reduction. Traditional crowd-control tactics (e.g., an aggressive show of force to gain compliance) are shunned in favor of interventions that increase cooperation and trust between police and crowd members.

What Lessons Can We Draw from This? The above analysis suggests that environmental criminology principles and theories are highly applicable to crowd behavior. The standard assumptions of the perspective offer direction for empirical research and implications for the development of effective strategies to reduce crowd-related crime. We offer a set of five principles to guide police-based crowd management strategies. These principles combine the basic tenets of the environmental criminology perspective outlined in this chapter with findings from European crowd studies and social psychological theory (as presented in the first half of this volume). Principle 1: Early and continuous police intervention is necessary to establish a positive and cooperative context for an event. Crowd management

strategies are more likely to reduce the potential for harm if police interventions are planned and implemented throughout all five crowd process stages. Early and consistent interaction between police and others associated with the event will encourage more proactive than reactive responses to crowd-related crimes. Police involvement in the initial planning stage is critically important for at least two reasons: (1) by understanding the event objectives, police can help set peaceful event expectations, and (2) partnerships to facilitate legitimate goals of the event can be established. Once police understand the legitimate objectives of event organizers, they can focus their efforts on identifying and removing potential obstacles to help them safely achieve these goals. When approached as a community policing effort, this exchange between event planners and police will likely increase trust and cooperation between parties. Event planners will then be more likely to respond positively to police requests for assistance or modifications to initial plans, particularly if they view police as helpful facilitators rather than unwelcome obstructers to their event.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 133

CROWD-RELATED CRIME

133

Early police involvement also permits the development of partnerships with others who can share public safety responsibilities related to crowd management. Partnerships can increase the number and effectiveness of the three crime controllers identified by routine activity theory: target guardians, offender handlers, and place managers. A proposed solution to a safety hazard may be better implemented by private property owners, industry representatives, neighborhood organizations, event planners or attendees, or other government entities (Scott and Goldstein 2005). For example, a parks department may provide additional shelter during severe weather conditions, or a neighborhood church may organize volunteers to help direct pedestrian traffic. Forming direct partnerships with crowd attendees can be especially beneficial. As noted by Stephen Reicher (Chapter 2), the empowerment of crowd members will likely increase self-policing activities within the crowd and thus reduce the overall need for police intervention. Put in environmental criminology terms, empowering crowd members not only avoids provocations and excuses for misbehavior, it also improves the chances members of the crowd will act as handlers who control potential offenders. From an environmental criminological perspective, the police, following Reicher’s advice, are acting as “super controllers”—agents that help make handlers in crowds more effective by setting clear and fair behavioral expectations and assisting crowd members (i.e., handlers) to influence others who may act inappropriately (Sampson, Eck, and Dunham 2010). Principle 2: Detailed analyses of crowd characteristics and processes should direct the selection of police interventions. As stated previously, crowds are

not inherently problematic, but specific event characteristics can increase the likelihood of crowd-related crime. These problems can be identified and addressed by analyzing the characteristics of planned activities, event locations, and crowd participant goals. Harmful conditions caused by the dynamic interplay of these elements at each stage of the crowd process can be addressed using the crime reduction techniques advocated by situational crime prevention. It is difficult to predict every potentially problematic crowd dynamic, but the interaction between some social and physical situational crowd features will invariably provide opportunities for misbehavior (e.g., groups of highly intoxicated rival fans waiting in long lines for buses, near trash bins overflowing with glass beer bottles, following a hotly contested game). The Flashpoints Model of Public Disorder draws attention to the importance of the interplay between these significant cultural-, contextual-, situational-, and interactionallevel crowd variables (see Chapter 6). Familiarity with the goals and activities of the participants, along with the physical features of the locations that constitute the crowd’s awareness and action space, will help police identify specific crime opportunity structures.

07_Madensen_Ch07.qxd:Madensen

134

1/3/11

10:38 AM

Page 134

PREVENTING CROWD VIOLENCE

Once uncovered, these opportunity structures can be altered or eliminated at each stage of the crowd process to reduce motivation for misbehavior. Harms can be minimized by systematically manipulating or blocking crime opportunities. To do this effectively, police must understand how opportunities arise and change throughout the crowd process stages. The information gathered during the analysis phase of the problem-solving effort will help piece together crime scripts that describe this process. These scripts will explain how crowd members may perceive and use crime opportunity structures, identify methods they may use to exploit these opportunities, and suggest potential points for intervention. Principle 3: Crime opportunities can be reduced by managing crowd participants’ perspectives. Police can use situational crime prevention techniques at

various points in the crowd process to alter participants’ perceptions of opportunities or remove elements necessary to complete an offense. Interventions grounded in this framework will make decisions to engage in specific crowdrelated crimes appear less attractive and thus encourage alternative behaviors. In general, crowd management strategies are more likely to be successful if: 1. multiple dimensions of crime opportunity structures are altered; 2. multiple situational crime prevention techniques are used to alter specific opportunity dimensions; 3. interventions are implemented in each of the crowd process stages. Situational crime prevention suggests that people are influenced by opportunities on five dimensions (i.e., effort, risk, reward, provocation, excuses). Increasing the effort necessary to complete an offense will likely have a deterrent effect. However, a stronger effect can be achieved by increasing both effort and risk. Increasing effort and risk and eliminating excuses for misbehavior will produce an even stronger deterrent effect, and so forth (Sousa and Madensen, Chapter 8). Two factors will likely impact the success of altering a particular opportunity dimension: (1) how well police execute selected interventions, and (2) how many techniques are used. For example, provocations to throw missiles at police can be reduced by designing interventions to reduce emotional arousal of crowd participants (e.g., use low-visibility policing tactics). Using more than one technique to reduce provocations (e.g., also avoiding disputes between participants and police by engaging in dialogue policing; Holgersson and Knutsson, Chapter 10 in this volume) is more likely to effectively reduce opportunities for missile throwing. Crime script analyses will uncover the crowd process stages in which interventions are needed. The implementation of interventions at multiple stages will increase the likelihood of crime “failure” (i.e., the inability to carry

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 135

CROWD-RELATED CRIME

135

out a particular offense). Using interventions only during the assembled gathering will fail to discourage crowd-related crimes at early and later stages of the crowd process. It is important to consider that crowds are not homogeneous groups. Participants do not come to the crowd with the same viewpoints, but their placement within the crowd places them in different situations. Consequently, managing crowds by managing perspectives will necessarily involve managing multiple perspectives. Principle 4: Crowd participants will be less likely to engage in harmful behavior if offered attractive legitimate opportunities to accomplish their objectives. It is not sufficient to simply block opportunities for misbehavior.

Providing alternative opportunities for crowd behavior serves two important purposes. First, safe alternatives to achieve crowd objectives will make “blocked” crime opportunities appear even less attractive. For example, police may increase the risks associated with tagging local landmarks during a graffiti artists’ convention. However, this intervention would be more effective if police also designate a prominent wall for artists to display their work during the convention. Second, crowd members are more likely to view police interventions as legitimate if they believe the police are helping them achieve their legitimate objectives. Offering to pour alcohol into plastic cups instead of simply confiscating glass bottles at street parties will reinforce positive perceptions of police interventions. Reicher (Chapter 2) argues that police should not view crowds as dangerous mobs that must be stopped, but rather understand their aims and help to achieve their objectives. The graffiti and plastic cup examples demonstrate this principle; police can create opportunity structures for less harmful behaviors. Ingrid Hylander and Kjell Granström (Chapter 5) also assert that it is not enough to try to avoid crowd aggravation (i.e., reduce provocations). Instead, police must also engage in activities that promote peace between authorities and crowd participants. Providing attractive and safe opportunities to achieve crowd objectives is one way to encourage positive interactions. Principle 5: Reactive responses to harmful behavior must be highly targeted. Well-planned police strategies can reduce the potential for crowd-

related crime. Proactive prevention efforts are generally more effective than reactive police responses. Still, harmful behaviors may occur when unforeseen circumstances arise. Appropriate police responses are critical in these situations since they may serve to escalate or de-escalate crowd violence. Environmental criminology encourages us to reject the view that crowds are inherently irrational and uniformly dangerous. Crowds are collections of individual rational actors. As such, responses to misbehavior by individuals or small groups of individuals should never be directed toward the crowd as a whole.

07_Madensen_Ch07.qxd:Madensen

136

1/3/11

10:38 AM

Page 136

PREVENTING CROWD VIOLENCE

Effective responses to crowd violence require accurate differentiation between problematic and nonproblematic crowd members (i.e., those who are engaging in violence and those who are not). Hylander and Granström (Chapter 5) assert that early positive interaction with crowd members will result in positive stereotyping and reduce the likelihood that police will use indiscriminate force against a crowd. Indiscriminate use of force will serve to unite crowd members against police and escalate violence, as explicated by the Elaborated Social Identity Model of crowd behavior (Stott, Chapter 3, or Reicher, Chapter 2). Therefore, police must use force only as a last resort to maintain order and only against specific individuals who are engaging in violent behavior.

A Final Note Environmental criminology offers a promising approach to problems associated with crowds. It is not only consistent with recent research on crowds and police-crowd interactions, it provides guidance to police and policymakers that is consistent with promoting democratic principles while reducing crime. That is not to say that the application of environmental criminological approaches guarantees democratic principles: situational crime prevention techniques, for example, can also be used to suppress decent and prevent legitimate gatherings. European crowd research provides both empirical findings that assist in creating effective crowd management practices and a strong normative perspective that helps to promote fair crowd management practices. Assisting the legitimate goals of crowd members is a description of a police practice that appears to work, but also a prescription for promoting democracy. Though simple to state, in many (though not all) circumstances this prescription will be hard to carry out. Any large crowd is diverse, so there may be multiple, possibly competing goals. It is not uncommon for two crowds to form precisely to oppose the other. The perception of what is a legitimate goal for the crowd may differ, not only within or between crowds, but also between members of the crowd and the police. And finally, police are not simply neutral actors but, as every political protester knows, they are agents of the state and take orders from elected policymakers. Ultimately, how well the police uphold democratic principles when working with crowds will depend on those who hold political power.

Notes We very much appreciate the contributions of the Stavern conference attendees who helped to shape our work through discussions of their research. We thank Darwin Morgan and Johannes Knutsson for their useful comments.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 137

CROWD-RELATED CRIME

137

1. In 2008, 150 people directly participated in the annual Nude Pumpkin Run in Boulder, Colorado (United States), by running along city streets wearing only shoes and a hollowed-out pumpkin on their heads. The next year, police publicized indecentexposure statutes that could require naked participants to register as sex offenders if nude runners were caught and convicted. The run took place in 2009, but all participants wore clothing, thus preventing incidents of indecent exposure (BBC News 2009). Though successful in getting crowd members to wear clothing, one has to be concerned that the potential harm to the nude pumpkin racers from the threatened enforcement far exceeded the harm caused by nude pumpkin racers.

07_Madensen_Ch07.qxd:Madensen

1/3/11

10:38 AM

Page 138

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 139

8 The Police and Major Event Planning: A Case Study in Las Vegas, Nevada William H. Sousa and Tamara D. Madensen

THE TOURISM INDUSTRY IS OF CRITICAL IMPORTANCE TO LAS Vegas, Nevada, in the United States. Revenues generated from gaming, entertainment, and conventions contribute significantly to the economies of the city and the state. In 2007, the city welcomed over 39 million visitors—the overall economic impact of tourists was estimated at over $41.5 billion, and revenues from gaming were estimated at nearly $10.9 billion (Center for Business and Economic Research 2009). To increase tourism, the city actively promotes its reputation as a place for amusement, leisure, and distraction. Nicknames that include “The Entertainment Capital of the World” and “Sin City,” and mottos such as “What Happens Here, Stays Here,” epitomize the often flamboyant (and sometimes risqué) status that the city enjoys. The center of activity in Las Vegas is “the Strip”—a four-mile-long corridor along Las Vegas Boulevard where over thirty casino-resorts are located (see Figure 8.1). Although widely known as establishments for gambling, the casino-resorts offer numerous services and forms of entertainment to visitors, including restaurants, shopping malls, convenience stores, convention centers, museums, nightclubs, and theaters. While all of the casinos employ private security personnel, the Las Vegas Metropolitan Police Department (LVMPD) is the primary agency responsible for police services on the Strip. This agency provides police services to the majority of Clark County (Nevada), with a jurisdiction that extends over 7,560 square miles and contains a population of more than 1.5 million residents. In 2008, LVMPD employed 2,815 police officers, 1,426 civilian personnel, and 802 corrections officers (LVMPD 2008). Although its jurisdiction is large, particular emphasis is placed on maintaining safety within Las Vegas’s core revenue-generating district. The department specifically maintains a Convention Center Area Command and a Tourist Safety Unit to address crime and safety 139

08_Madensen_Ch08.qxd:Madensen

140

1/3/11

10:40 AM

Page 140

PREVENTING CROWD VIOLENCE

Figure 8.1

The Las Vegas Strip and Significant Landmarks

Stratosphere S tra to s p h e re Sahara S a h a ra A Avenue venue

Sahara S a h a ra A Avenue venue S a h a ra Sahara Fontainebleau F o n t a in e b le a u R Resort e s o rt Riviera R iv ie r a Convention C onvention Center Center Drive D riv e Encore E n c o re Wynn W ynn L Las as Vegas Vegas

Hilton H ilton Grand Grand Vacations Vacations Club C lu b Circus C ir c u s C Circus ir c u s

Fashion F a s h io n S Show how Mall M a ll

Treasure T reasure Island Is la n d The T he Mirage M ir a g e Caesars C a e s a rs P Palace a la c e Flamingo F lamingo Road Road Bellagio B e lla g io The T he Cosmopolitan C o s m o p o lit a n City C ity Center C e n te r Monte Carlo Monte C a r lo New N ew York, York, New New Y York o rk

Las L as Vegas Vegas Boulevard B o u le v a r d

Spring S pring Mountain Mountain Road Road

Sands S ands A Avenue venue The Palazzo The P a la z z o The T he Venetian V e n e t ia n Casino C a s in o R Royale o y a le Harrah’s H arrah’s IImperial mperial Palace P a la c e Flamingo F la m in g o Bill’s B ill’s G Gamblin’ a m b lin ’ H Hall all and and S Saloon a lo o n Flamingo F lamingo Road Road Bally’s Bally’s Paris P a r is Planet P la n e t H Hollywood o lly w o o d Harmon H armon Avenue A venue MGM M GM Grand G ra n d

Tropicana T r o p ic a n a A Avenue venue

Tropicana T r o p ic a n a A Avenue venue Excalibur E x c a lib u r

Tropicana T r o p ic a n a

Luxor L uxor Four F our S Seasons, easons, M Mandalay andalay Bay, Bay, THE THE Hotel H o te l Russell R u s s e ll R Road oad

concerns in and around the four-mile stretch of Las Vegas Boulevard depicted in Figure 8.1. Policing Las Vegas Boulevard is complex for a variety of reasons. Defining the “community,” for instance, can be difficult. LVMPD maintains close working relationships with management and security officials in the casinos, but tourists—the bulk of the population—are transitory and come from socially, economically, culturally, and demographically diverse backgrounds. Numerous problems associated with crowds, physical and social disorder, crime, pedestrian traffic, and motorized traffic occur all along the casino-resort corridor. Problems on the Strip are compounded during major events. Las Vegas often welcomes major headliners from the music, sports, and entertainment industries who perform at the resorts’ concert halls and event centers. Typically these events are localized to one casino, but on occasion they will involve a general celebration along Las Vegas Boulevard. On New Year’s Eve, for example, the Strip alone hosts an estimated 400,000 people, making it the second largest New Year’s celebration in the United States.1

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 141

THE POLICE AND MAJOR EVENT PLANNING

141

New Year’s Eve on the Strip presents numerous challenges to LVMPD. The street itself is closed to motorized traffic and crowds are allowed to congregate along Las Vegas Boulevard to celebrate. Public alcohol use is permitted, which contributes to the already hectic and disorderly atmosphere created by the enormous increase in crowd size. Although LVMPD has always had a presence on the Strip during the New Year’s celebration, they historically relied on reactive police techniques, responding to problems in the crowd after they occurred. Beginning in 1998, LVMPD increased the use and subsequent review of proactive crowd management techniques aimed at reducing opportunities for harmful behavior during New Year’s Eve celebrations on the Strip (Schofield 2000). Since then, the department’s strategy has been reviewed and revised each year based on the outcome of specific interventions used the previous year. Today, numerous proactive interventions are used as part of a comprehensive New Year’s Eve crowd management strategy. The strategy used by LVMPD on New Year’s Eve in 2008 was considered a major “success” based on the absence of injuries associated with violence, as well as the limited need for police interference or action within the crowd festivities. Using New Year’s Eve on the Strip as a case study, we attempt to explain the success of LVMPD’s crowd management strategy by relating twenty-five interventions used by the department to specific crime prevention techniques advocated by the situational crime prevention perspective. Although LVMPD did not initially develop their strategy based on the situational crime prevention framework, the congruence between the interventions and the crime reduction principles proposed by the perspective provide a theoretical basis to explain the success of these crowd management techniques. To establish context, we begin by discussing the literature on crowds and their propensity for harmful behavior, then present a review of a theoretical framework rooted in the dimensions of situational crime prevention. We then systematically analyze LVMPD tactics on New Year’s Eve using the situational crime prevention framework.

Harmful Behavior in Crowds Media reports tend to support the perspective that crowds encourage negative behaviors. Violence and other destructive behaviors are often associated with events that bring large numbers of people together (e.g., organized protests, celebratory gatherings, concerts, sporting events). But accurately predicting whether violence will occur at any specific event can be difficult and is generally impossible for those who lack a conceptual understanding of crowd behavior and the role of police in escalating or suppressing violence in crowds. Theory and research suggest that the likelihood of a particular gathering turning violent, absent external controls to prevent such behavior, will depend on

08_Madensen_Ch08.qxd:Madensen

142

1/3/11

10:40 AM

Page 142

PREVENTING CROWD VIOLENCE

a myriad of factors: numerous variables functioning at multiple levels of analysis (Waddington, Jones, and Critcher 1987, 1989; Waddington and Critcher 2000). It is no surprise, then, that some practitioners may find attempts to predict and prevent crowd violence overwhelming and futile; however, such efforts are essential to effective crowd management strategies. Without the ability to predict misbehavior, police tend to regard all crowds and crowd members as potentially problematic.2 However, oversimplification of crowd behavior (e.g., violent or nonviolent) and failure to treat members of a crowd as individuals can produce disastrous results (e.g., excessive use of force against passive crowd members, increasing the potential for crowd violence). Recent research supports this view and contradicts long-held popular assumptions promoted by early social psychologists. Most contemporary scholars refute Gustave Le Bon’s (1895) early and influential view of “mad” crowd behavior. Current perspectives maintain that individuals who stand within or participate in crowd activities do not lose their self-control, values, or ability to choose between alternative courses of action. While people may come together and participate in coordinated group activities because they share a common goal or develop a shared social identity (Reicher 1987, 2001b), individuals do not lose their cognition or ability to act independently. In an attempt to dispel the “Myth of the Madding Crowd,” Clark McPhail (1991) summarized a large body of empirical evidence confirming rational and independent decisionmaking by individuals in crowds. Thus, strategies designed to reduce harmful behavior in crowds should be based on two assumptions: (1) crowds are heterogeneous groups of people with individual agendas, and (2) crowd participation does not extinguish individual decisionmaking capacities. Recent case study research identifies a host of environmental and social factors that appear to increase the likelihood of violence in crowds; for example, extreme temperatures, alcohol availability, overcrowding, reputation of the event location, event significance, and overly aggressive actions of police and security personnel (see Dewar 1979; Russell 1983, 2004; Stott and Adang 2005; Courakis 1999; Connors 2007). However, evaluation research suggests that the influence of these common “triggers” can be mitigated through careful analysis of situational characteristics and proper event planning (e.g., La Crosse, Wisconsin, annual canoe race [Sampson and Scott 2000]; Munich, Germany, annual Oktoberfest [Weiss and Davis 2005]; annual Australian Motorcycle Grand Prix [Veno and Veno 1993]). Documented case studies describe specific crowd management techniques or interventions used to prevent or reduce crowd violence and other harms. Not all techniques and interventions are equal; some are more effective than others. Similarly, some combinations of interventions are presumed to better prevent harm. We will discuss how the situational crime prevention per-

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 143

THE POLICE AND MAJOR EVENT PLANNING

143

spective can be used to explain why some crowd management strategies are more effective than others.

Opportunity Structures for Crowd Violence The situational crime prevention perspective provides a framework for understanding how individuals perceive opportunities for crime. It is grounded in the rational choice perspective, which assumes that all human behavior is purposive and rational (Cornish and Clarke 2008). The perspective insists that opportunity is an important cause of criminal behavior (Clarke 1997a, 2008) and that opportunity is a construct of five dimensions of perception: 1. 2. 3. 4.

the effort required to act, the risk of detection associated with the act, the rewards expected to be gained from the act, the existence of provocations that encourage the act and reduce inhibitions, and 5. excuses that can be used to justify the act. Together, these five dimensions create perceptions of opportunity structures for behavior. The attractiveness of any particular course of action is dependent on the effort, risk, reward, provocations, and excuses associated with that action. In other words, strong opportunities for harmful behavior exist when effort and risk are low, rewards and provocations are high, and when the behavior can easily be justified by the participant. As mentioned previously, social psychologists contend that individuals retain their decisionmaking abilities as crowd participants. This suggests that the dimensions of opportunity outlined by the situational crime prevention perspective can be manipulated to manage crowds. Perceptions, and thus behavior, can be controlled through the implementation of various interventions. Situational crime prevention offers a matrix of twenty-five general prevention techniques that can be used to alter perceptions of attractive opportunities (Cornish and Clarke 2003). Table 8.1 lists the techniques associated with each dimension. The notion that crowd behavior can be managed through situational prevention techniques is gaining acceptance. Situational techniques have been proposed for controlling student party riots and spectator violence in stadiums (Madensen and Eck 2006, 2008). Police in India have adopted preventionbased approaches to crowd rioting by applying a variety of situational techniques, including rule setting, target hardening, access control, and formal surveillance (Verma 2007). Furthermore, situational techniques are frequently

08_Madensen_Ch08.qxd:Madensen

144

1/3/11

10:40 AM

Page 144

PREVENTING CROWD VIOLENCE

Table 8.1

Techniques of Situational Crime Prevention

Increase Effort

Increase Risks

Reduce Rewards

Reduce Provocations

Remove Excuses

Harden targets

Extend guardianship

Conceal targets

Reduce frustrations and stress

Set rules

Control access

Assist natural surveillance

Remove targets

Avoid disputes

Post instructions

Screen exits

Reduce anonymity

Identify property

Reduce emotional arousal

Alert conscience

Deflect offenders

Utilize place managers

Disrupt markets

Neutralize peer pressure

Assist compliance

Control tools and weapons

Strengthen formal surveillance

Deny benefits

Discourage imitation

Control drugs and alcohol

used to control unwanted behavior in places where large numbers of people congregate, such as subways (La Vigne 1997), seasonal events (Björ, Knutsson, and Kühlhorn 1992), tourist resorts (Homel et al. 1997), city-center markets (Poyner and Webb 1997), and Disney World (Shearing and Stenning 1985). A few of the interventions used to manage crowd behavior in the aforementioned settings are summarized in Table 8.2. LVMPD effectively manages the behavior of approximately 400,000 visitors who celebrate New Year’s Eve on the Las Vegas Strip each year. The strategy used by the department is complex, involving the implementation of numerous prevention techniques and partnerships with casinos and other city agencies.3 We dissect the department’s strategy using the situational crime prevention framework to better understand the effectiveness of this approach. This exercise will help determine the degree to which the opportunity structure for crowd violence is altered by their crowd management strategy.

Analysis The enormity of the crowd on New Year’s Eve combined with the limited width of Las Vegas Boulevard (eight lanes at its widest point—see Figure 8.2) present challenges common to other events that involve large numbers of people in confined spaces. Overcrowding, the allowance of alcohol use, and the significance of the event contribute to these concerns. Other challenges relate to unique structural characteristics of the Strip. Some casino-resorts, for example,

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 145

THE POLICE AND MAJOR EVENT PLANNING

Table 8.2

145

Situational Techniques Used to Manage Crowd Behavior

Type of Event/Location

Intervention

Situational Technique

Student parties

“Sanitize” event locations by removing objects and debris that can be thrown or set on fire

Control tools/weapons

Stadiums

Provide adequate and proper placement of facilities (e.g., vendors, restrooms)

Reduce frustrations and stress

Antigovernment demonstrations

Assign additional security personnel to highprofile targets

Strengthen formal surveillance

Subways

Clean graffiti and repair damages from vandalism within twenty-four hours

Discourage imitation

Midsummer’s Eve in Sweden

Refuse admission of high-risk groups

Deflect offenders

Tourist resorts

Promote the discounting of low- or nonalcoholic drinks

Control drugs and alcohol

City-market centers

Reduce pedestrian congestion by increasing the width of shopping aisles

Increase natural surveillance

Disney World

Relay recorded messages concerning appropriate behaviors while entering and exiting rides

Set rules

have distinctive features such as fountains, bodies of water, statues, and other large structures along the roadway. While these features enhance the aesthetic qualities and distinctiveness of each resort, they provide numerous opportunities for vandalism and other acts of disorder during large celebrations. Past policing strategies were problematic and did not effectively address these issues during the New Year’s celebration. Gary Schofield (2000) identifies several problems with LVMPD’s past tactics. For example, officers were typically deployed randomly throughout the crowd, which greatly limited both their visibility and mobility, and contributed to officer stress and fatigue. If an altercation occurred between celebrants (or between an officer and a member of the crowd), considerable time would elapse before police responded— assuming they could see the altercation at all. Moreover, the overall strategy was reactive in nature—while the presence of uniformed officers in the crowd served as some deterrence, they were barely visible to anyone beyond their immediate vicinity. Typically, the only time members of the crowd saw police was when fellow celebrants were being arrested—an activity that, without context, often led to frustration on the part of onlookers who viewed police actions as authoritarian and severe.

08_Madensen_Ch08.qxd:Madensen

146

1/3/11

10:40 AM

Page 146

PREVENTING CROWD VIOLENCE

Las Vegas Boulevard

Wikimedia Commons/Mike Russell, photographer

Figure 8.2

With these challenges in mind, LVMPD began to adopt a more proactive approach to better manage crowds and reduce opportunities for celebrant misconduct in the late 1990s (Schofield 2000). Planning for the event occurs throughout the year. Typically, after-action reports are generated after each New Year’s event between January and March. Evidence concerning the effectiveness of specific interventions is collected, and decisions to eliminate, modify, or replace noneffective tactics are made in April and May. In June, LVMPD begins to coordinate planning efforts with surrounding police departments, city and county agencies, and other external organizations (e.g., Nevada Department of Transportation, Department of Homeland Security, portable toilet companies). Between August and October, LVMPD gathers information concerning specific events scheduled along the Strip. The fourmile corridor is broken down into five sections to assess the level of manpower and resources necessary to maintain safety in these areas. Vast amounts of agency resources are concentrated in this area in an effort to maintain order. All LVMPD officers are accounted for on New Year’s Eve to orchestrate a safe environment for the 400,000 celebrants on the Strip and provide increased levels of service to other areas in the department’s jurisdiction. Meetings within the department and between agencies increase in frequency during November and December until New Year’s Eve day, when specific crowd management interventions are used to facilitate orderly celebrations. We discuss several of these interventions below as they relate to each of the five dimensions of situational crime prevention.

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 147

THE POLICE AND MAJOR EVENT PLANNING

147

Increasing Offenders’ Effort

A clear focus of LVMPD’s interventions is to limit crowd misbehavior by making disorderly activities more difficult to perform. Using the language of situational crime prevention (Clarke 2008), police tactics during New Year’s Eve now involve hardening potential targets, better control over access to critical areas, deflecting offenders away from possible targets, and better control over objects that could be used as instruments of crime or disorder (Table 8.3). Perhaps the single most significant intervention developed by LVMPD in terms of managing the New Year’s celebration is the use of a “barricade tactic” on the Strip (Schofield 2000). On the evening of the event, soon after the street has been closed to motorized traffic, over 4,000 movable barricades are placed along the sidewalks on Las Vegas Boulevard (see Figure 8.3). As the crowd grows throughout the evening, officers on the inside will slowly move the barricades toward the center of the street, keeping the center area clear from all but the police. After the event, as the crowd thins, officers will slowly move the barricades back toward the sidewalks to facilitate the dispersal process. This barricade tactic allows for better management of pedestrian traffic as the crowd size increases. Importantly, however, it also allows police officers

Table 8.3

Interventions to Increase Effort

Situational Technique

Intervention

Behavior Managed

Harden targets

Movable barricades protect police from crowd

Reduces unnecessary physical contact between police and participants Prevents participants from climbing and falling from posts

Grease applied to light posts Control access

Pedestrian walkways closed

Prevents objects from being thrown on officers or the crowd

Deflect offenders

Movable barricades manage pedestrian traffic Ambulances, emergency vehicles, and garbage trucks strategically placed to manage vehicle traffic

Keeps pedestrians from wandering into restricted areas Prevents vehicle traffic from entering pedestrian-designated zones

Control tools/ weapons

Plastic cups or bottles for beverages Trash bins and plastic cups kept within barriers

Removes glass as a potential weapon and safety hazard Eliminates vandalism of bins and potential use as missiles

08_Madensen_Ch08.qxd:Madensen

148

1/3/11

10:40 AM

Page 148

PREVENTING CROWD VIOLENCE

Figure 8.3

Movable Barricades

6:00 p.m. p.m. 6:00

LasVegas Vegas Blvd. closed toto Las Blvd closed vehicletraffic; traffic; barricades barricades keep vehicle keep pedestrians sidewalks pedestrians on on sidewalks

7:00 p.m. p.m. 7:00

Officersmove move barricades barricades Officers towardthe theboulevard boulevard center toward center as crowd size increases as crowd size increases

8:00 p.m. 8:00 p.m.

Pedestrians restricted restricted only Pedestrians onlyfrom from centerlanes lanes as as the center the midnight midnight fireworks display approaches fireworks display approaches

Geographic Features Casino

Las Las Vegas Vegas Blvd. Blvd

Barricade

Crowd

Police officers

to stay separated from the crowd for the duration of the event. This separation serves a number of purposes, including providing a measure of protection for officers. Because the area in the center of the Strip remains off limits to everyone but the police, officers are not engulfed by the crowd, they do not need to fight their way through celebrants when responding to incidents, and they can more readily call for assistance from fellow officers who are also within the barricade. In addition to the barricade tactic, several other interventions have been designed to make disorderly behavior more difficult. Pedestrian walkways over the Strip, for example, are closed during the New Year’s event to prevent objects from being thrown on officers or the crowd. Likewise, grease is now applied to light posts along the Strip to prevent celebrants from climbing the posts and injuring themselves or others. Officers discourage the use of glass containers for beverages (which can be used as potential weapons) by providing plastic cups from within the barricade. Trash bins kept within the barriers minimize litter (and also prevent trash containers from being vandalized or used as weapons). Finally, ambulances, emergency vehicles, and garbage trucks are strategically placed at intersections to allow for their quick deployment, but also to manage vehicle traffic to Las Vegas Boulevard before, during, and after the event. Increasing Offenders’ Risk

LVMPD’s efforts on New Year’s Eve are also designed to limit celebrant misconduct by increasing the chances that those who engage in dangerous or dis-

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 149

THE POLICE AND MAJOR EVENT PLANNING

149

orderly conduct in the crowd will be caught. Interventions are therefore focused on improving the levels of natural surveillance, formal surveillance, and surveillance by place managers (Table 8.4). In addition to protecting officers, the aforementioned barricade tactic enhances surveillance. Officers can better view the overall event, locations on the Strip, and members of the crowd because the barricades separate police from celebrants. The barricade improves visibility and officer mobility, allowing more effective deployment into the crowd in the event of a problem. Furthermore, the barricade separation provides celebrants a better view of police and makes it easier to alert them in the event of difficulty. LVMPD officers enhance their capabilities through more traditional surveillance tactics as well. For example, other police units, such as mounted patrol and correctional staff, augment patrol officers who make up the bulk of the police presence on the Strip.4 Additionally, the police make extensive use of closed-circuit television cameras located at major intersections along Las Vegas Boulevard. Video from the cameras is fed live to off-Strip locations. From there, officers who monitor the video can communicate with police on the ground should any problems arise. LVMPD also conducts a comprehensive planning process with management and security staff at the various casino-resorts along the Strip. Resort personnel are primarily responsible for security within their own buildings, but they also provide important assistance to LVMPD in terms of crowd monitoring. This is especially the case when the casino-resorts have unique structures along the Strip that require special attention during the New Year’s Eve celebration.5 Table 8.4 Situational Technique

Interventions to Increase Risk Intervention

Behavior Managed

Assist natural surveillance

Barricade strategy improves officer ability to observe crowd behavior and increases officer visibility

Serves as a visual deterrent to participants who may try to hide within the crowd

Utilize place managers

Coordination with casino staff to enhance crowd monitoring

Assists police with the implementation of interventions

Strengthen formal surveillance

Live-feed video surveillance

Helps police and other service agencies quickly identify and respond to high-risk situations Acts as a “force-multiplier” by providing additional personnel Increases the likelihood that police will observe misconduct

Use of correctional staff Horse patrol enhances observation and visibility

08_Madensen_Ch08.qxd:Madensen

150

1/3/11

10:40 AM

Page 150

PREVENTING CROWD VIOLENCE

Reducing Offenders’ Rewards

Misbehavior in a crowd setting is less likely to produce tangible rewards than other types of economically motivated crimes (e.g., robbery, burglary). While we can directly reduce the rewards associated with the theft of money or property (e.g., dye packs in stolen cash, installing verification codes for use of laptops), reducing rewards in crowds requires the consideration of indirect benefits associated with misbehavior. The strategy used on the Strip utilizes three techniques aimed at reducing rewards for inappropriate behavior: concealing targets, removing targets, and denying benefits (Table 8.5). Casino exteriors are highly decorative. Some of these decorations represent attractive targets for vandalism. To reduce temptations to damage, deface, or remove these decorations, police and casinos remove or conceal many exterior fixtures for the duration of the New Year’s Eve celebration (e.g., statues, fountains). One of the most obvious rewards associated with participating in the New Year’s Eve celebration on the Strip is simply being part of the crowd. The threat of removal by police reduces the rewards associated with engaging in misbehavior. Forced or voluntary departure from the Strip denies the benefits of crowd participation, and this police response is used to deter potential offenders. Reducing Offenders’ Provocations

A major concern for police at large-scale events is the level of agitation among crowd members. Individuals that experience higher levels of stress, anger, and other negative emotions are more likely to react violently to environmental stressors (Wortley 2001). Thus, effective crowd management strategies must address the potential for provocation by taking steps to limit or remove social and physical instigators that encourage harmful behaviors. The interventions

Table 8.5

Interventions to Reduce Rewards

Situational Technique

Intervention

Behavior Managed

Conceal targets

Conceal property that would invite vandalism

Reduces attractive opportunities for vandalism

Remove targets

Remove exterior casino fixtures that can be moved

Prevents theft of structures or their use as weapons

Deny benefits

Remove highly disruptive participants from the crowd

Helps to deter potential offenders

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 151

THE POLICE AND MAJOR EVENT PLANNING

151

used by LVMPD incorporate all five situational crime prevention techniques for reducing provocations (Table 8.6). One of the many attractions of New Year’s Eve on the Strip is the fireworks display that occurs at the end of the event. People stand in the street along the four-and-a-half-mile stretch of the Strip to watch the show. People who experience an obstructed view of the fireworks could begin pushing and shoving to move to better vantage points within the crowd, thus, increasing crowd frustration and stress and escalating the potential for violence. However, this problem is managed by launching the fireworks from several locations along the Strip. This enhances the visibility of the fireworks for all crowd participants and prevents massive crowd movement toward or away from specific locations. Overly intrusive or aggressive policing tactics can increase crowd violence (Connors 2007) and serve to unite crowd members against police (see discussions of the Elaborated Social Identity Model discussed by Stephen Reicher in Chapter 2 and Clifford Stott in Chapter 3). To “humanize” the officers and reduce the potential for disputes, officers are encouraged to engage in friendly conversation with members of the crowd. Officers interact with crowd participants from behind barriers. This encourages the perception that officers will only intervene if necessary. It also prevents tensions from escalating (e.g., an intoxicated person bumps or falls on an exhausted police officer who may respond in an overly forceful manner). Police also attempt to promote friendly crowd interaction by having officers dress in “soft uniforms” rather than more

Table 8.6 Situational Technique

Interventions to Reduce Provocations Intervention

Behavior Managed

Reduce frustrations and stress

Fireworks displayed evenly along the strip

Prevents crowd members from pushing and shoving to observe firework displays

Avoid disputes

Officers encouraged to converse with crowd Barricades prevent contact between officers and crowd

“Humanizes” both officers and participants Prevents stressed or tired officers from responding with force

Reduce emotional arousal

Officers dressed in “soft” uniforms rather than militarized/riot gear

Promotes a friendly and cooperative atmosphere

Neutralize Off-Strip arrest processing peer pressure

Reduces potential for crowd interference with police actions

Discourage imitation

Reduces potential that harmful behaviors will be imitated

Quick response to disruptive crowd members

08_Madensen_Ch08.qxd:Madensen

152

1/3/11

10:40 AM

Page 152

PREVENTING CROWD VIOLENCE

militaristic riot gear (“riot-ready” police officers and riot response equipment are not visible to the public). This helps to reduce negative emotions that could be aroused by an unnecessary show of police force. When someone is arrested, police quickly escort the person away from the crowd to buses that facilitate arrest processing. These buses are located off the Strip and act as holding facilities for offenders as they await jail transport. Immediately removing the individual from the crowd reduces peer pressure, or the likelihood that other crowd members will attempt to interfere with police action against the offender. Quick action to calm and remove disorderly crowd members also helps to reduce the potential that others will mimic harmful behavior. Although we know that people make rational and purposeful decisions in crowds and retain self-control, observing misbehavior may reduce inhibitions in others who are inclined to engage in similar behavior. Immediate removal of disruptive persons can discourage imitation by other crowd members as long as police actions are perceived as just and legitimate by other crowd members. Removing Offenders’ Excuses

Undesirable behaviors can be managed by making it easier for people to recognize appropriate courses of action. In their strategy, LVMPD attempts to set rules that discourage inappropriate behavior by both crowd participants and officers. The department also provides resources along the Strip that make appropriate behavior a more attractive option (Table 8.7). To control inappropriate officer behavior, supervisors must approve all arrests of crowd participants. Supervisor approval reduces the likelihood of unnecessary arrests or use of force by police. This rule prevents overly fatigued and stressed officers from making impulsive decisions that could elicit a negative crowd response. Police use the fireworks display as a “trigger” for crowd dispersal. Once the show is finished, the crowd is slowly guided back onto the sidewalks and into casinos. Street sweepers are then used to clean the streets, thus signaling Table 8.7

Interventions to Remove Excuses

Situational Technique

Intervention

Behavior Managed

Set rules

Supervisors approve arrests Fireworks signal end of event

Reduces the likelihood of inappropriate officer behavior Facilitates dispersal process

Portable toilets provided Trash bins provided

Reduces public urination Reduces litter

Assist compliance

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 153

THE POLICE AND MAJOR EVENT PLANNING

153

the end of the “street party.” Only after the streets have been cleared of pedestrians and cleaned is vehicle traffic permitted on the Strip. This rule reduces the potential for pedestrian injuries. Public urination may be an attractive alternative for intoxicated persons who do not want to enter a casino, or who encounter excessively long lines at first-floor casino restrooms. This behavior is discouraged through the strategic placement of portable toilets along Las Vegas Boulevard. Similarly, litter is controlled by providing numerous trash bins along the Strip. Intervention Summary

This systematic analysis of twenty-five interventions used by LVMPD suggests that their tactics alter all five dimensions of opportunity as defined by situational crime prevention theory. Overall, seventeen of the twenty-five techniques outlined by the situational crime prevention framework have been incorporated into LVMPD’s crowd management strategy. The specific techniques used by the department are indicated in Table 8.8. Our evaluation reveals that each opportunity dimension (i.e., effort, risk, reward, provocation, and excuses) is altered to a substantial degree by the strategy. This is accomplished through the implementation of multiple techniques related to each dimension, with seven techniques employed through the use of multiple interventions. Table 8.8 suggests this strategy is a relatively balanced approach, with added emphasis placed on reducing crowd provocations. In

Table 8.8

Techniques Used in LVMPD’s Crowd Management Strategy

Increase Effort

Increase Risks

Reduce Rewards

Reduce Provocations

Remove Excuses

Harden targetsa

Extend guardianship

Conceal targetsa

Reduce frustrations and stressa

Set rulesa

Control accessa

Assist natural surveillancea

Remove targetsa

Avoid disputesa

Post instructions

Screen exits

Reduce anonymity

Identify property

Reduce emotional arousala

Alert conscience

Deflect offendersa

Utilize place managersa

Disrupt markets

Neutralize peer pressurea

Assist compliancea

Control tools and weaponsa

Strengthen formal surveillancea

Deny benefitsa

Discourage imitationa

Control drugs and alcohol

Note: a. Specific techniques used by the Las Vegas Metropolitan Police Department.

08_Madensen_Ch08.qxd:Madensen

154

1/3/11

10:40 AM

Page 154

PREVENTING CROWD VIOLENCE

addition to assessing the overall balance of the strategy, this evaluation identifies eight techniques that have not been utilized by the department. This finding may suggest new possibilities to further increase the effectiveness of LVMPD’s crowd management strategy. Success of LVMPD’s Crowd Management Strategy

Earlier, it was noted that the crowd management strategy used by LVMPD on New Year’s Eve in 2008 was considered a major success. Police data indicate that an officer injury has not occurred at this annual event since 2004. Additionally, levels of misconduct and the need for police action declined from 2006 to 2008. Figure 8.4 presents the numbers of citations, misdemeanor arrests, and felony arrests made by LVMPD at the Strip’s New Year’s Eve celebrations from 2006 to 2008. The number of citations issued for minor violations (e.g., minor in possession of alcohol) remained relatively consistent, but there was a dramatic decline in the numbers of misdemeanor and felony arrests made at these events. While it can be difficult to determine whether any particular event should be considered a “success” from a policing standpoint, the general decline in criminal activity coupled with the relatively low numbers of formal police action against crowd members (in relation to the 400,000 participants) suggests that the current crowd management strategy is highly effective.

Figure 8.4

New Year’s Eve Citation and Arrest Statistics

140

116

120

100

80

60

58

55

51

53 42

40

39

27 20

1 0

2006 Citations

2007 Misdemeanor Arrests

2008 Felony Arrests

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 155

THE POLICE AND MAJOR EVENT PLANNING

155

Conclusion Those who manage crowds must identify practical interventions to prevent disorder and violence within highly complex environments. The results of this exercise suggest that a balanced and situational approach to crowd management can effectively prevent harm. Police in Las Vegas have managed the potential for harmful behavior by altering perceptions of opportunity for misconduct among crowd participants. This chapter concludes with a brief discussion of general considerations related to applying situational crime prevention to crowd management and suggestions for future research activities. The Complexity of Crowds

Preventing harm during large gatherings is a complicated and difficult task. This task differs from many other police problem-solving efforts in at least four ways. First, effective crowd management requires the consideration of several types of “harm.” Second, gatherings occur as a result of a process represented by various stages—police must extend their prevention efforts beyond a single time or place. Third, police actions can, if taken without careful planning and an understanding of the crowd’s goals and objectives, escalate violence in crowds. Fourth, rarely can crowds be managed effectively through the use of one or two opportunity reduction techniques. Each of these considerations is discussed briefly below. Numerous, perhaps countless, crime opportunities exist at events that attract large crowds. For example, there are at least six forms of aggression that are commonly reported in crowds (see Lewis 2007; Sharp 1992; Ward 2002; Young 2002): 1. Verbal aggression—singing, chanting, and yelling taunts or obscenities; 2. Aggressive gesturing—signaling to others with threatening or obscene motions; 3. “Missile” throwing—throwing items such as food, drinks, bricks, bottles, broken seats, and cell phones at particular or random targets; 4. Swarming—rushing toward a specific location, such as an entry or exit, which may result in injury or death from trampling; 5. Property destruction—knocking down structures, graffiti, bonfires, and other forms of vandalism; and 6. Physical violence—spitting, kicking, shoving, fistfights, stabbings, and shootings. Crowd management strategies should include techniques that make all types of harmful behavior less attractive to crowd members. When it is not possible to block an opportunity for violence (e.g., verbal aggression at sporting

08_Madensen_Ch08.qxd:Madensen

156

1/3/11

10:40 AM

Page 156

PREVENTING CROWD VIOLENCE

events), strategies should be designed to minimize damage or discourage violent reactions from other crowd members when these incidents occur. Obviously, efforts should be made to discourage violent behaviors among crowd members. However, effective crowd management strategies must also anticipate the potential for harm created by the process and conditions of traveling to and from the event, as well as the environmental features of the location where the crowd will gather (see Chapter 7 by Tamara Madensen and John Eck for a description of the various stages of crowd formation). Even seemingly nonharmful structures can create dangerous conditions, as illustrated by the use of light posts along the Las Vegas Strip. Unintended consequences may result from the use of any particular crowd management tactic. The impact of any specific intervention is highly contextspecific. A tactic that suppresses violence under one set of conditions can escalate violence in another (e.g., asking people to disperse at the end of a concert versus during the middle of a protest). Police agencies should not simply adopt crowd management tactics found to be effective elsewhere. A careful analysis of an event’s social and physical situational conditions must guide the selection of appropriate interventions (see Brown and Scott 2007 for a detailed discussion concerning important considerations for proper implementation of interventions using a police problem-solving approach). Previous case studies have found that specific crimes can be significantly reduced through the implementation of a single intervention: obscene phone calls decline with the use of caller identification (Clarke 1990), ink tags reduce the theft of specific store merchandise (DiLonardo and Clarke 1996), and steering column locks reduce incidents of motor vehicle theft (Webb 1994). However, crowds do not represent a specific target. As noted previously, these are heterogeneous groups of individuals, often with very different agendas.6 Similarly, crowd participants may harm themselves, other crowd members, police and other event staff, or destroy property through various methods. The complexity of crowds demands a more comprehensive approach to harm prevention. Policy Implications for Crowd Management

Our analysis of the LVMPD New Year’s Eve strategy suggests important policy implications for those who police large-scale events. We propose that effective crowd management strategies must accomplish two general objectives: (1) significant alteration of all opportunity dimensions related to engaging in misbehavior and (2) application of interventions that affect behavior at micro, meso, and macro levels. We propose that the LVMPD’s New Year’s Eve crowd management strategy is effective because it accomplishes both objectives. To significantly alter a specific opportunity dimension (i.e., effort, risk, reward, provocation, or excuse), multiple techniques outlined by the situational crime prevention framework should be utilized. Our classification of

08_Madensen_Ch08.qxd:Madensen

1/3/11

10:40 AM

Page 157

THE POLICE AND MAJOR EVENT PLANNING

157

the LVMPD crowd management strategy clearly shows that each dimension is manipulated through multiple techniques and interventions. Furthermore, some of these interventions are target-specific or micro in focus (e.g., greasing light posts, arresting and removing individuals), others affect small groups at an intermediate level (e.g., officer interaction with the crowd, distributing plastic cups), while others influence crowd behavior at a macro level (e.g., barricades that control pedestrian traffic flow, fireworks displays distributed evenly along the Strip). Future Research

Although beyond the scope of this chapter, future research should describe the process through which police develop crowd management strategies. In Las Vegas, as in Madison, Wisconsin (Plant and Scott, Chapter 9 in this volume), police use an iterative problem-solving approach. An “after-action” report is generated by LVMPD following each event. If unanticipated problems occurred at the event or during the execution of a particular intervention, new interventions or techniques are considered and the strategy is modified. A summary of best practices for after-action analysis could increase the effectiveness of crowd management strategies. Theoretical frameworks that assist in the development of crowd management strategies should be extended and developed. A five-stage planning framework (Madensen and Eck 2006) and a violence triangle that categorizes some “triggers” of event violence (Madensen and Eck 2008) have been offered. However, a classification of particular crowd types and corresponding management needs would be useful. Similarly, a comprehensive list of problems commonly faced within particular crowd types could assist strategy formation processes. Situational crime prevention advocates a problem-specific approach to reducing harm (Clarke 1992), and research shows that interventions designed to address specific harms are more likely to be effective. If event planners are aware of common harms, they can manipulate opportunity structures before the problems must be addressed in an after-action analysis. Finally, more case studies on event planning and summaries of interventions used in crowd management strategies should be offered. This effort will help to increase general knowledge of effective, and ineffective, practices. Police, government agencies, and researchers should consider establishing forums to facilitate the distribution of this information.

Notes We would like to thank the Las Vegas Metropolitan Police Department and, in particular, Deputy Chief Gary Schofield for allowing us to observe the crowd management

08_Madensen_Ch08.qxd:Madensen

158

1/3/11

10:40 AM

Page 158

PREVENTING CROWD VIOLENCE

strategy used on New Year’s Eve. Additionally, we thank Tana Gurule for her initial research related to this topic. We are also grateful for the comments and suggestions provided by Johannes Knutsson and attendees of the Stavern conference regarding an earlier version of this chapter. 1. New York City is the largest. 2. This is not an extreme view; police are trained to approach almost all policecitizen interactions in this manner (e.g., traffic stops). 3. Partnerships with other groups and organizations are critical to police success in controlling large crowds. Although not the focus of this chapter, the importance of these partnerships is explored by Knutsson and Clarke (2006) and Chamard (2006). 4. LVMPD serves both a patrol and a corrections function for Clark County, Nevada. Corrections officers employed by LVMPD typically provide services to the jail and court systems of the county, but can be used under certain circumstances to supplement patrol services. 5. For example, several casino-resorts have bodies of water located on the Strip outside of the main structure. Casinos have placed divers in the water during New Year’s Eve to prevent problems related to vandalism or celebrant injury. 6. For example, some people visit the Strip on New Year’s Eve to become intoxicated and party, others join the crowd to people-watch, others want to spread God’s message through signs and bullhorns, and others wish to view the casino-sponsored fireworks.

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 159

9 Trick or Treat? Policing Halloween in Madison, Wisconsin Joel B. Plant and Michael S. Scott

THE CITY OF MADISON IS THE STATE CAPITAL OF WISCONSIN AND home to the flagship campus of the statewide university system, the University of Wisconsin–Madison (UW-Madison). The city has approximately 225,000 permanent residents and nearly 45,000 postsecondary students. Uniquely situated on an isthmus between two large lakes, the downtown area is highlighted by State Street, a well-known and popular pedestrian mall that links the university campus to the state capitol building. State Street is eight blocks long and includes many bars, restaurants, theaters, museums, and unique shops intermingled with residential properties (see Figures 9.1 and 9.2 for descriptive maps). The Madison tradition of a Halloween party on State Street began in 1977 when a crowd of approximately 5,000 people made their way to State Street on October 31 and started a bonfire. The fire was extinguished and the crowd dispersed without significant incident. In 1978, police reported a larger-thantypical crowd on State Street with no notable incidents. Police estimated that the Halloween party crowd exceeded 100,000 people for several years in the 1980s, although, as will be discussed further later, crowd estimation methods were questionable. In 1979, the student government body at the UW-Madison, the Wisconsin Student Association (WSA), harnessed the students’ desire for a fall party and sponsored the first “official” Halloween party, complete with city-issued permits, entertainment stages, and beer sales. WSA recognized revenue from beer sales as a significant fundraiser for the organization and continued its sponsorship of the party through 1987. In 1985, the drinking age in Wisconsin was raised from 18 to 21, which effectively “killed the party,” according to a former WSA leader (Vevea 2008). Tightening of the laws regulating temporary public beer sales brought an end to the sponsored party in 1988 (Schnell 2005). 159

Map of Central Madison Event Zone, Halloween 2006

160

09_Madensen_Ch09.qxd:Madensen

Figure 9.1

1/3/11 10:39 AM Page 160

09_Madensen_Ch09.qxd:Madensen

Map of State Street Area Figure 9.2

1/3/11

10:39 AM

Page 161

161

09_Madensen_Ch09.qxd:Madensen

162

1/3/11

10:39 AM

Page 162

PREVENTING CROWD VIOLENCE

The Halloween party crowds in the early to mid-1980s were large, boisterous, intoxicated, and festive, and required a lot of resources to police, but were on the whole peaceful. Anecdotal reports of property damage from the early years include several small bonfires and broken windows. The most traumatic Halloween incident occurred in 1983 when a man fell from a roof and sustained serious injuries. As paramedics prepared to transport him to a hospital, a Halloween partygoer stole the keys to the ambulance, significantly delaying the victim’s arrival at a hospital. The victim ultimately died from his injuries. After a decade of relative quiet and declining crowds on Halloween, the party began an informal resurgence in 1998 and continued to grow in size. The party date was difficult for police to predict: sometimes it occurred on Halloween itself (October 31), but occasionally it occurred on the Saturday closest to Halloween. Even without a formal sponsor for the event, police reported that the crowds were relatively well behaved and, perhaps most important, dispersed of their own accord in the early morning hours after the party (Vevea 2008). The 2002 event proved pivotal in changing officials’ perceptions of the Halloween gatherings; the rambunctious street party was now seen as a major public safety problem. By night’s end on Saturday, November 2, two fires were set, several officers and citizens sustained significant injuries, public and private property was destroyed, and tear gas was used to disperse the riotous crowd. The primary objective for the 2003 event planning was to avoid a repeat of 2002. The 2002 after-action report concluded that, “Merely increasing police planning and staffing cannot guarantee that a crowd will not become aggressive, damage property, and confront the police. It will require a coordinated effort to minimize the risk of violence in 2003” (Silverwood and Housley 2002, 10). Despite sincere efforts by community leaders and government officials to identify effective responses to the riotous behavior, the 2003 Halloween weekend ended with a large crowd of intoxicated individuals causing injuries and property damage. The police department again deployed large amounts of pepper spray to disperse the crowd.

Defining the Problem Applying the CHEERS test (community, harmful, expectation, events, recurring, similarity; Clarke and Eck 2005) to the Madison Halloween party problem yields the following conclusions. The community was directly and indirectly affected by the Halloween party. The harmful effects of the event included: some partygoers were injured during the event; merchants and other property owners in the party area sustained significant damage or looting; some merchants who did not cater to par-

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 163

TRICK OR TREAT?

163

tygoers’ needs lost business revenue by being forced to close during the event; motorists were inconvenienced by the traffic management plan; and taxpayers incurred higher costs for policing the event. Particularly after the 2002 and 2003 events, the community assertively expressed its expectation—through phone calls, letters, e-mails, and letters to the editor—that the city address the problem. The harmful event was not the Halloween party itself, but rather the climax of the party during which some of the crowd became disorderly and riotous. The problem was recurring in that it coincided with an annual event that had become part of the local culture and, absent some special intervention, the disorder and rioting was unlikely to cease. Indeed, even the rioting was becoming a dangerous part of the local cultural tradition. And although each year’s event might bring unique specific harms, there was an overriding similarity to the problem from year to year in that it occurred on roughly the same date, location, and time of day each year, drawing attendees from a similar demographic, and being preceded by a similar pattern of activity (displaying costumes, consuming a lot of alcohol, crowding, and refusing to go home at the end of the evening). Precisely defined, the problem was the behavior of a group of intoxicated individuals whose disorderly, violent, and criminal actions during the annual Halloween party enticed a larger group to participate in disorderly and destructive behavior, which required a police response that included substantial use of force, generated significant costs to be borne by the public, and resulted in a loss of public support for the annual event and an unwelcome reputation for Madison in the eyes of many civic leaders. The city’s goal has been to reduce and prevent violence and property damage and to control the public costs associated with policing the annual Halloween party on State Street.

An Iterative Problem-Solving Process The effort to improve the policing of Madison’s annual Halloween party was rooted in practical necessity rather than academic interest. Consequently, the findings and conclusions are not the product of a carefully constructed research study on crowd control. Rather, they are the product of much trial and error, reflections on past experiences, and the analysis of data and perceptions gleaned from varied sources. Prior to 2002, Madison’s Halloween party had been managed, for the most part, by local police officials and event sponsors (when applicable) without a significant amount of additional assistance. It was planned and policed in accordance with crowd control principles and practices that had served the police department and city well for the previous thirty years. Not since the riots associated with the Vietnam War protests in the late 1960s and early

09_Madensen_Ch09.qxd:Madensen

164

1/3/11

10:39 AM

Page 164

PREVENTING CROWD VIOLENCE

Figure 9.3

Iterative Problem-Solving Initiative: The “SARA Helix” 2007 analysis

2006 responses 2006 assessment 2005 responses 2005 assessment

2005 analysis

2004 responses 2004 assessment

2004 analysis

2003 responses 2003 assessment 2002 scan

2006 analysis

2003 analysis

1970s did Madison police have cause to question the efficacy of their crowd control approach. Being a college- and state government–centered community, Madison police had extensive experience, and considerable success, planning for and managing large crowds and events, from college football games and sports championship celebrations to political protests, street fairs, and parties. After the 2002 event, however, it was apparent that the growing crowds, lack of a formal sponsor, and significantly increased levels of injuries and property damage required a new approach. In preparation for the 2003 event, police and city event planners analyzed observational data from the 2002 event. A variety of responses were considered, and selected responses were implemented for the 2003 event. In spite of the careful planning and implementation of new responses, the 2003 event ended in much the same fashion as in 2002—with rioting and looting. As the community began preparing for each subsequent Halloween party, the police’s and city’s documentation of the planning process and plan implementation improved, and an iterative problem-solving method emerged. Event planners routinely analyzed both historic and new data. They read the research literature on crowd control and studied reports of other communities’ experiences with similar events. They explicitly considered the successes and failures

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 165

TRICK OR TREAT?

165

from the previous year’s event in their planning of each new event. They formulated a variety of hypotheses about the problem’s causes and contributing factors. They met with a wide range of stakeholders, both formally and informally, to explore new response options. Each year they assessed the success or failure of each new response. At the most basic level, that assessment was simple: Did the event end in rioting and looting? Each year, assessment of the responses signaled the beginning of the next iterative phase of the initiative. The Madison Halloween problem-oriented policing initiative can be characterized as an iterative problem-solving process. This process is best conceptualized as a “SARA helix” (scanning, analysis, response, and assessment; see Figure 9.3). The responses implemented by year can be reviewed in Appendix A. The following synopses of the events from 2002 to 2007 are derived from several sources: the official police after-action report, media reports, interviews with police officers, and the authors’ firsthand knowledge of the event. Halloween 2002 Pre-event planning. The 2002 Halloween party was the twenty-sixth time the Madison Police Department had prepared for Halloween revelry. Since the inception of the party in 1977, the police had identified an array of interventions that they found helpful to some degree. Many of these interventions were again used in 2002, including the following: establishing a command post, deploying a traffic plan to divert vehicular traffic around the event area, mobilizing arrest conveyance teams, asking bar owners to not serve beverages in glass containers, and having the fire department search for and remove combustible material from the event area. New interventions for 2002 sought to enhance safety and reinforce crowd control. Bicycle racks were removed from the event area to eliminate items that could obstruct the crowd and be used as projectiles. Police established a traffic team to control vehicle traffic. The city created temporary taxi stands in two locations to facilitate alternate transportation to and from the event. Event planners met with bar owners whose bars were in the center of the event area to develop plans to reduce outdoor customer queues and crowding. Cell phones were replaced with land-line telephones in the command post to improve the reliability and clarity of communications. Sanitation employees placed large trash bins at two key locations where police officers could dispose of confiscated alcohol and dangerous items. Police staffing levels were increased, with fifteen officers employed from another police agency. Some police officers were assigned to patrol via bicycle. The event. Police estimated the crowd size on both Thursday, October 31, and Friday, November 1, at 20,000 to 25,000. Signaling the beginning of the major crowd event, police deployed their traffic control plan at 10 P.M. Both nights,

09_Madensen_Ch09.qxd:Madensen

166

1/3/11

10:39 AM

Page 166

PREVENTING CROWD VIOLENCE

police made a handful of arrests, primarily for alcohol violations and fighting, but overall, disturbances and property damage were not extraordinary. Saturday, November 2, began routinely, with police reporting a relatively quiet State Street as of 8 P.M. By 9:15 P.M., police commanders decided to keep officers past the end of their regular shifts to assist with the quickly growing crowd, effectively adding thirty-five officers to the thirty-six Special Events Team (SET) officers. By 10 P.M., police had again deployed their traffic control plan and pedestrians began occupying all of State Street. By midnight, police estimated the crowd on State Street at 50,000. Fights erupted frequently. Female partygoers who were baring their breasts to the crowd from hotel and apartment windows were warned by police to stop. By 12:30 A.M., the 500 block of State Street was impassable to police officers. At about 1 A.M., the crowd started two small fires in the 500 block of State Street. By 2 A.M., police estimated the crowd had grown to 60,000 to 65,000 people, with many partygoers appearing to be highly intoxicated. At this hour, time clocks were rolled back to 1 A.M. in recognition of the start of daylight savings time. Shortly after 2 A.M., police stopped a striptease being performed in the window of an apartment in the 600 block of State Street, and the crowd reacted by breaking apartment windows. A few minutes later, police halted another striptease being performed in the window of a hotel in the 500 block of State Street, and again, the crowd reacted by breaking windows. At 2:30 A.M., eighteen SET officers donned riot gear. At 2:40 A.M., paramedics tried to assist and treat a person who had suffered a potentially life-threatening head injury. As police began clearing the area around the victim and paramedics, the crowd began throwing projectiles, including beer cans and bottles, rocks, and pieces of concrete. SET officers moved in to try to disperse the crowd. At 2:50 A.M., police declared the gathering an unlawful assembly and ordered the crowd to leave the area. After issuing the order and observing little compliance, police commanders ordered that pepper spray be used to disperse the crowd. At 3 A.M., as police officers began running out of pepper spray, the crowd’s movement stalled, and partygoers began destroying traffic signals, trash barrels, and newspaper racks. A third fire was started and the crowd threw burning material at police officers. When police officers reported that they were completely out of pepper spray, the event commander ordered the use of CS tear gas to disperse the crowd. At this, the crowd began to retreat from State Street, but vandalized and looted shops as they did so. Finally, at 4:15 A.M., State Street was cleared of all partygoers. Post-event assessment and planning. In the days following the event, event planners collected data on the damage caused and costs incurred to police the event. Twelve businesses sustained a combined $24,000 in damages and one business—a liquor store—lost approximately $8,000 of inventory to looters.

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 167

TRICK OR TREAT?

167

Police made several dozen arrests. Thirteen police officers and an unknown number of partygoers were injured. The public cost for policing the event was estimated at $77,753. These assessments, and post-event assessments for succeeding events, are summarized in Appendix B. To end the post-event assessment stage, a list of recommendations to reduce misbehavior during the next annual event was developed for use in the subsequent 2003 pre-event planning stage. Halloween 2003 Pre-event planning. Shortly after the 2002 event, a group of policymakers, police, and community stakeholders met to begin planning for Halloween 2003. The group included representatives from the common council, city police and fire departments, mayor’s office, student government, student newspapers, chancellor’s and dean of students’ offices, university police, business advocacy groups, liquor license holders, and area residents. The group developed three goals for the 2003 Halloween party: (1) maintain a safe and orderly event; (2) eliminate the opportunity for injury to citizens and officers; and (3) prevent property damage. In order to meet these general goals, an operational plan was developed and implemented, based heavily on the 2002 post-event recommendations. Each of the 2002 interventions was replicated with various modifications based on prior experience (e.g., police increased their stockpile of chemical munitions). The following new responses were implemented in 2003:

• The city created a Halloween event Web page to provide information on traffic, lodging, parking, and behavioral expectations. • Police took arrested lawbreakers into physical custody, rather than citing and releasing them. • Police established an arrest processing center to facilitate processing the scores of arrestees; staff from the UW-Madison Dean of Students’ Office were present at the arrest processing center to initiate university discipline against arrested UW-Madison students. • Police officers from four neighboring police agencies (Capitol Police, UW-Madison Police, Wisconsin State Patrol, and Dane County Sheriff’s Office) reinforced Madison police. • Remote-controlled video cameras were placed in the 500 and 600 blocks of State Street to monitor the crowd and identify law breakers. • Bands provided live entertainment near the 700 block of State Street. • The common council enacted a temporary prohibition against possessing glass containers within the greater event area. • Public works employees used large motorized street sweepers to clean the event area and to signal the end of the party.

09_Madensen_Ch09.qxd:Madensen

168

1/3/11

10:39 AM

Page 168

PREVENTING CROWD VIOLENCE

The event. The 2003 Halloween party was expected to occur on both Friday,

October 31, and Saturday, November 1. On Friday, October 31, police again reported that the crowd was heavily intoxicated and police responded to many fights and calls for service. Police estimated that the crowd size peaked at 65,000 to 70,000. Just before midnight, the crowd in the 500 block of State Street swelled and strained against an iron fence surrounding a private parking lot. Officers secured permission from the property owner to unlock the gate, thereby providing relief to the large crowd. At 12:30 A.M., officers again observed women baring their breasts in the windows of the University Inn hotel and arrested several of them. By about 2:50 A.M., officers moved the crowd from the street onto the sidewalks, and by 4:00 A.M., the crowd had dissipated and officers were dismissed. The party resumed at 6 P.M. on Saturday, November 1, with a live band performing onstage in the 700 block of State Street. As it began to rain at 10 P.M., the traffic control plan was put into effect. At 11:30 P.M., police again received reports of women exposing themselves at the University Inn and several were arrested. At 1 A.M., the band stopped playing, as scheduled. At 1:20 A.M., the crowd formed a “mosh pit” near the now-empty stage and began moving toward the 500 block of State Street.1 At 1:30 A.M., the iron gate at State Street and North Frances Street was opened to relieve pressure from the crowd. At 1:45 A.M., police received a report of a vehicle overturned on Mifflin Street in a nearby student neighborhood. At 2 A.M., police estimated the crowd size at 65,000 to 70,000. Some partygoers got into fights with police officers. Police asked hotel guests to stay away from hotel room windows. Some people climbed trees and light poles. At 2:15 A.M., police officers were pelted with projectiles by the large crowd in the 500 block of State Street. At 2:30 A.M., platoons of SET officers donned protective gear. At 2:30 A.M., paramedics were summoned to treat a person outside a nearby bar who had stopped breathing and had no pulse. At 2:40 A.M., the crowd continued throwing various projectiles like decorative landscaping stones. At about 3 A.M., SET officers deployed pepper spray. By 4:15 A.M., the crowd was gone and police officers were dismissed. Post-event assessment and planning. Event planners again collected data to

quantify the event’s impact on the community. Seven businesses reported damage but no looting was reported. The most heavily damaged property was the University Inn. With the exception of the pulseless nonbreather mentioned above, no police officers or partygoers suffered significant injuries related to the event. Police made 406 arrests related to the event, mostly for alcohol violations. Excluding the costs of officers from other police agencies, the Madison Police Department staffing costs were about $100,000. While the 2003 event’s outcome was arguably an improvement over 2002, police and city officials were not satisfied. For the second straight year,

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 169

TRICK OR TREAT?

169

activities at the University Inn and nearby student house parties remained significant contributing factors to the problem. Event planners asked several key questions as they began planning for 2004: How can an extremely dense, impromptu crowd be managed? How can event planners signal to the crowd that the event is over and encourage dispersement without inciting further disturbances? Why are the 500 and 600 blocks of State Street the site of so many problems? What is known about the most troublesome partygoers? What effect does the sudden emergence of a large contingent of police officers wearing riot gear have on the crowd’s behavior? Is it helpful or antagonistic? The community stakeholder group’s recommendations for 2004 reaffirmed the need for a comprehensive community response to Halloween and again identified alcohol as a major factor contributing to the problem. The group acknowledged that while a small number of people intentionally caused harm at the event, many others joined in destructive and violent actions. The group recognized the need to focus more on the relatively small number of troublemakers and on the specific locations where trouble was repeatedly occurring. Halloween 2004 Pre-event planning. After receiving the recommendations from the community stakeholder group, many of the prior year’s responses were replicated in 2004, some with minor modifications (e.g., all involved officers were trained in special event tactics). The glass container ban was heralded as a success by street department employees responsible for cleaning the event area and by police who noted fewer projectiles thrown at them. Police reported a high degree of compliance with the glass ban in 2003. Video surveillance of the event was expanded, as the live feed was found to be helpful. Police were able to process arrestees more efficiently and securely at the new arrest processing center. The city’s Web page was deemed successful in providing the public event information. Several new responses were planned for the 2004 event:

• The common council increased fine amounts for common event-related violations such as disorderly conduct, underage possession of alcohol, and possession of an open alcohol container in public. • The city formed a partnership with liquor distributors who printed and distributed informational materials discouraging the overconsumption of alcohol. • City officials enlisted the cooperation of the local bar-owners’ association in developing strategies to prevent problems in and around establishments.

09_Madensen_Ch09.qxd:Madensen

170

1/3/11

10:39 AM

Page 170

PREVENTING CROWD VIOLENCE

• City officials asked downtown rental property owners, hotel managers, student residence hall managers, and regional college and high school officials to remind their populations of the consequences of unlawful behavior at the Madison Halloween event. • Online advertisements and social pages were monitored to collect intelligence. • The mayor, police chiefs, sheriff, district attorney, and university chancellor held a pre-event press conference to reinforce the community’s expectations. • The city installed a public address system to communicate with partygoers during the event. • Police formulated a special plan to police house parties. • The public works department brought to the event area two stadiumstyle light trucks to better illuminate the 500 block of State Street and to signal the end of the event. • University officials bought and distributed $15,000 worth of free pizza to partygoers at the end of the event. • The University Inn offered police free use of the hotel as a police staging area. Although recommended by the 2003 stakeholder group, nothing specific was done to restrict the celebration to a single day. Live music—a major feature of the 2003 event—was discontinued in 2004. University student groups argued for live bands but failed to apply in time for a street use permit as required by city ordinance. The event. On Friday, October 29, an estimated 60,000 to 65,000 people

turned out on State Street. By 3 A.M., officers reported that the crowd was diminishing. Mounted police officers proved quite effective at engaging the crowd in a positive manner. On Saturday, October 30, with an estimated crowd of 85,000 in the area, police put the traffic plan into effect at 10:30 P.M. By 11:30 P.M., police reported a large influx of intoxicated partygoers from nearby student neighborhoods. By about 1:30 A.M., the crowd formed mosh pits at North Frances and State streets. Foot patrol and mounted officers unsuccessfully attempted to disperse the crowd. The crowd started throwing projectiles at the officers. At 1:35 A.M., the crowd started a fire at the intersection of North Frances and State streets. The police turned on the bright, stadium-style lights to signal to the crowd that the event was over. Police used pepper spray on some partygoers in order to clear a path for firefighters to extinguish the street fire. At 1:40 A.M., SET officers donned their riot gear. At 1:45 A.M., using the public address system, police declared the assembly to be unlawful. At 2 A.M., police officers in protective gear engaged the crowd and were met with flying projectiles,

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 171

TRICK OR TREAT?

171

ranging from bottles to large rocks. By 4 A.M., the crowd dissipated and officers were dismissed. Post-event assessment and planning. Six businesses reported property damage. One police officer was slightly injured while making an arrest. Police effected 578 arrests for 801 separate offenses, a 42 percent increase in the number of arrests over 2003. Madison Police Department costs totaled $255,536, and outside police agencies providing mutual aid assistance reported costs of $234,009 for a total police cost of $489,545. Overall, 2004 was viewed as moderately successful, as injuries and property damage continued to decline. The growing public cost of the event, coupled with the necessity of declaring an unlawful assembly and deploying pepper spray meant that the city had not yet fully achieved its goal of having a fun and safe event at reasonable public expense. In preparing recommendations for 2005, the stakeholder group developed hypotheses related to the problems with the Halloween event. It was clear that the 500 block of State Street was continuing to serve as “ground zero” for the end-of-event confrontation. In 2003, poor lighting conditions in the area were identified as a contributing factor and, as a result, two light trucks were used to illuminate the area at the end of the party in 2004. Unfortunately, police found that the lights created a “moths to the flame” effect on the crowd, actually increasing the crowd size within minutes of turning on the lights. Planners also concluded that the primary antagonists at the end of the event were arriving late, were heavily intoxicated, and typically were not in costume.

Halloween 2005 Pre-event planning. Immediately after the 2004 event, the mayor publicly entertained the possibility of putting an end to the whole Halloween event. The mayor’s frustration was stoked by some unwelcome national media attention about the 2004 party. Specifically, Sports Illustrated mentioned Madison’s Halloween riots, which city officials felt encouraged readers to travel to Madison to participate (Duerson 2004). While the mayor’s comments were met with some skepticism, the mechanisms necessary to declare emergency powers were put into place for the 2005 party. Event planners identified four major issues needing attention: the impact of nearby house parties, the prevalence of people from outside Madison contributing to the problems, continuing concentration of problems in the 500 block of State Street, and the impact that the state-mandated bar closing time—2:30 A.M. on weekend nights—had on the problem. In August 2005, the city and university jointly hired a full-time alcohol policy coordinator in recognition of the centrality that alcohol policies and practices play in a range of municipal and campus problems. Also that year,

09_Madensen_Ch09.qxd:Madensen

172

1/3/11

10:39 AM

Page 172

PREVENTING CROWD VIOLENCE

the Center for Problem-Oriented Policing was nearing completion of its publication of a guide for police on student party riots (Madensen and Eck 2006) and an advance draft was shared with the Madison police and the alcohol policy coordinator, specifically to help inform its Halloween event planning. These two developments injected into the iterative problem-solving initiative greater attention to crowd control theory and research, the experiences of other communities in addressing problems similar to Madison’s Halloween events, and data collection and analysis. The discussions of Madison’s Halloween event planners were now framed by the problem analysis triangle (Center for Problem-Oriented Policing 2009), based on routine activity theory (Cohen and Felson 1979). This theory posits that crime or disorder occurs when likely offenders and suitable victims or targets come together in time and space, in the absence of capable controllers (i.e., target guardians, offender handlers, and place managers—see Center for Problem-Oriented Policing 2009). Applying this framework to the Madison Halloween context, event planners recognized that police, university officials, bar owners, and responsible partygoers all acted as handlers in some respects. Shop owners, city officials, responsible partygoers, and police acted as guardians of potential victims and targets; and police, city officials, and shop and hotel owners acted as place managers. A new central insight was that the event lacked capable nonpolice guardianship—to better manage the place, protect targets, and handle offenders. With police acting by default as the predominant guardians of places, targets, and offenders, they had in fact become targets themselves, the central object of partygoers’ hostility. This insight pointed to the need for a nonpolice guardian of Madison’s Halloween event to emerge. The mayor articulated his goals and specific success indicators for the 2005 event: (1) no serious injuries or deaths related to the party, (2) prevention of crowd behavior that would necessitate that police use protective gear or deploy pepper spray on a wide scale, and (3) dispersal of the crowd peacefully and in an orderly fashion at the party’s end. As in years past, many of the responses from previous Halloween parties were replicated or modified for the 2005 event. Light trucks were again used in the 500 block of State Street but were now turned on throughout the event. More mounted patrol officers were deployed and were used to prevent the crowd flow from stagnating, particularly in the 500 block of State Street. Police doubled the number of house-party patrol teams from two to four, and added a fire inspector to each team. The University Inn was again used as a police staging area, although the owners now required the city to pay for its use. The city extended the glass ban to a larger area. All Madison police officers were ordered to work the Halloween party in some capacity alongside officers from other police agencies. The public address system was enhanced

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 173

TRICK OR TREAT?

173

with better equipment and prerecorded messages.2 More alcohol purveyors cooperated with event planners in controlling alcohol sales. The mayor requested that the city attorney and police chief produce a memorandum outlining options both for reducing Halloween-related problems and for eliminating the Halloween party altogether (May and Wray 2004). The options they outlined included: • announcing a plan to stop or scale back the Halloween party by blockading access to the city at major intersections, blocking access to the 500 block of State Street, or closing State Street businesses; • gating the event and admitting only UW-Madison students; • eliminating beer keg sales; and • cancelling university classes on the Friday and Monday surrounding the Halloween weekend. The community stakeholder group recommended several new responses for the 2005 event, including: • deploying paramedics on foot to facilitate faster response times; • implementing a no-guest policy at UW-Madison dormitories, preventing residents from housing friends not affiliated with the university; • installing flexible fencing to restrict pedestrian access to State Street from North Frances Street so as to alleviate crowd pressure;3 • assigning the police crime analyst to the command post to evaluate information from field officers and identify emerging trouble spots; • using city buses as preliminary arrest booking stations; • requesting that bar owners allow and encourage patrons to exit from the rear of bars to reduce the crowd emptying onto State Street; and • requesting bar owners to voluntarily close their bars at 2:30 A.M. rather than taking advantage of the extra service hour allowed due to daylight savings time. The university’s free pizza distribution was discontinued in 2005 because partygoers tended to fight while waiting in line for pizza. The event. Police estimated the State Street crowd to number 15,000 to 20,000 on Friday, October 28. There were no major incidents reported, although one enormous fraternity house party involving more than 600 partygoers drew a lot of police and fire department attention.4 By 3 A.M., the crowd had dispersed and police officers were dismissed. On Saturday, October 29, approximately 200 police officers—an unprecedented number for the Halloween event—marshaled and patrolled on and

09_Madensen_Ch09.qxd:Madensen

174

1/3/11

10:39 AM

Page 174

PREVENTING CROWD VIOLENCE

around State Street. Police put the traffic control plan into effect at 10 P.M. The crowd was reportedly jovial until at least midnight. By 1 A.M., police estimated the crowd size at 65,000 to 70,000 and noticed that crowd flow was stagnating, especially in the 500 block of State Street. Commanders ordered all plainclothed officers off the streets because the crowd was too dense and aggressive. At 1:15 A.M., mounted patrol officers attempted to dislodge the stagnant crowd. At 1:25 A.M., a partygoer set off a smoke bomb. At 1:30 A.M., the crowd began chanting at and challenging officers. Foot patrol and mounted officers attempted to divide the crowd, but the crowd immediately closed in behind the officers as they worked their way through the crowd. At 1:35 A.M., partygoers began throwing projectiles. At 1:40 A.M., police aired the first prerecorded message over the public address system (“Thank you for coming . . . it is time to leave the area”). At 1:55 A.M., police extinguished a small fire in the street. At 2 A.M., police aired the second prerecorded message over the public address system (“You are participating in an unlawful assembly. You must leave the area immediately”). At 2:06 A.M., SET officers in protective gear sought to clear the street with pepper spray. At 2:18 A.M., mounted patrol officers requested assistance from SET officers. At 2:31 A.M., the crowd was throwing rocks at the officers, who responded with more pepper spray. By 2:40 A.M., all remaining partygoers had dispersed, and by 4 A.M., police officers could be dismissed. Post-event assessment and planning. Continuing the trend of incremental improvement, the 2005 event resulted in no significant injuries to officers or civilians and only one reported broken window within the event area. Costs, however, continued to rise, with Madison police reporting personnel, equipment, and rental costs of $349,092 and outside police agencies reporting costs of $255,753, for a cumulative total of $604,845, a 24 percent increase in policing costs for the two-day event. Police officers effected 468 arrests, a 19 percent reduction from 2004. Event planners collected additional data about arrestees. Their average age was twenty years old. There were 65 percent of arrestees claiming some collegiate affiliation, with 20 percent claiming they were UW-Madison students. Their mean blood alcohol content was .11, significantly higher than the .08 limit for driving a vehicle. Twenty-eight arrestees had blood alcohol contents higher than .20. The police house party patrol teams issued warnings at thirty-six houses. At fifteen houses police took official enforcement action. Forty-nine people from these fifteen houses were issued a total of 457 citations. The hosts at the fraternity house party with more than 600 guests faced fines totaling nearly $100,000.5 The annual Halloween party had become extremely expensive and potentially violent and still had no official sponsor. While the local press acknowledged that violence and property damage were decreasing, the story’s focus

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 175

TRICK OR TREAT?

175

turned to the event’s financial costs. The community stakeholder group voiced the frustrations being felt throughout the community. Postassessment recommendations focused mainly on enhancing those prior years’ responses that seemed to be working. Although the policing of the Halloween party had been improving incrementally from year to year, event planners seemed resigned that their efforts had reached a plateau. They felt there was a growing consensus in the larger community that the Halloween party should change more dramatically, but there was no consensus as to how to do so. Halloween 2006 Pre-event planning. In the early morning hours of October 30, 2005, as the Halloween event finally ended, the police chief asked the mayor at the command post whether or not the community was willing to end the event with the mass deployment of pepper spray each year. After considering the chief’s comments, the mayor outlined three goals for the 2006 event: (1) have the event end safely, (2) reduce the event’s public cost, and (3) reduce alcohol overconsumption and its consequential problems. The mayor asked the staff to consider every feasible option for fundamentally changing the city’s approach to the Halloween event. After four consecutive years of violence, property damage, pepper spray deployment, and growing public costs, the mayor and the community at large had had enough. As has been noted by other problem-oriented policing experts, sometimes a dramatically different response to a problem must await its moment of opportunity (Goldstein and Susmilch 1982). Such an opportunity had just presented itself for the policing of Madison’s Halloween. The mayor’s office proposed two approaches to reduce problems at the 2006 party: (1) charging property owners in the greater event area a special tax levy for the additional municipal services related to the event (referred to as the “Charge Plan”), or (2) creating a concert-style environment involving gating, paid ticketing, and a variety of entertainment options (referred to as the “Gating Plan”). The Charge Plan. The Charge Plan was based on statutory authority given to municipalities to impose a “special charge against real property for current services rendered” (Wisconsin Statutes).6 The Charge Plan called for identifying an event impact area and developing a novel cost recovery scheme based on a complex matrix of factors. Planners took into account the size, use, hours of operation, and zoning classification. Planners wanted to ensure that those properties benefiting most from the additional services rendered by the city during the Halloween party—such as bars, restaurants, liquor stores, and latenight entertainment venues—pay a larger share of the public costs, and conversely, those properties benefiting least from the additional services—such as

09_Madensen_Ch09.qxd:Madensen

176

1/3/11

10:39 AM

Page 176

PREVENTING CROWD VIOLENCE

residential, religious, and public facilities like libraries—pay a smaller share of the costs. The Gating Plan. The Gating Plan was premised on several hypotheses and

principles related to crowd control and situational crime prevention. The primary hypothesis guiding the development of the Gating Plan was that by rebranding the event and creating a concert-style environment with which partygoers would be familiar, violent, disorderly, and destructive behavior would be reduced. A secondary hypothesis was that the event could be transitioned into—and recognized as—a one-night event by defining specific hours for the event itself. For these purposes, a ticket would be required for admittance only on Saturday evening. All event advertising included a beginning and ending time. The conclusion of the live entertainment was intended to signal the event’s end. Event planners recognized that gating a public street and charging for admission to it would be controversial. The city attorney was asked to provide a legal opinion on the issue and he ultimately concluded that the proposed gating and ticketing mechanisms were permissible so long as the intent was to control the crowd and recover some portion of the public cost of the event (May 2006). The city attorney also concluded that the annual Halloween party constituted the quintessential unlawful assembly because the presence of tens of thousands of persons on State Street precluded normal pedestrian traffic.7 Given the high likelihood of an unlawful assembly, the city attorney advised that the police chief’s obligation to maintain order permitted the closure of public streets and the assessment of a fee to enter.8 The opinion concluded with the assertion that, “Steps reasonably calculated to control the crowd, to protect property and persons, and, as I understand part of the policy behind this idea, to change the nature of the Halloween party so that it may be attractive for a private sponsor, all seem within the City’s broad police power reflected in sec. 62.11(5), Wis. Stats.” (May 2006). Selecting and implementing the new framework. The mayor committed to

moving forward with either the Charge Plan or the Gating Plan but was interested in building consensus among key stakeholders before moving ahead with the planning process. A meeting was held in early February 2006 with representatives from the business community, the city’s common council, university administration, and relevant staff from a variety of city agencies. In presenting the two options, the Charge Plan was purposely presented first in an attempt to convince the assembled group that the mayor was indeed prepared to enact a fundamental change to the 2006 Halloween party. The proposal was not a ruse—in fact, it was a feasible, legal, and well-developed option that could be enacted ahead of the 2006 event. After the Charge Plan was presented, the lone question from the stakeholder group was, “You said there were two options?”

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 177

TRICK OR TREAT?

177

After presenting the Gating Plan, there was unanimous, albeit tepid, support for shelving the Charge Plan and moving forward with the Gating Plan. While many stakeholders expressed significant skepticism about the possibility of success, all agreed that a new framework was required for the annual party. While the police and fire departments again planned for the usual contingencies, event planners began working on the four major elements of the Gating Plan: logistics, entertainment, event staffing, and branding. The following sections discuss each of these elements in turn. Logistics. As expected, the notion of gating off a six-block stretch of a public

street filled with restaurants, bars, and retail establishments and requiring a ticket for admission raised concerns throughout the business community, although, remarkably, no formal opposition . The business community seemed to concur that the historic approach to the Halloween event was not resulting in a successful outcome. The city engineering and parks departments took the lead on designing lightweight, mobile, and low-cost fencing partitions. They incurred nominal costs to purchase orange plastic construction fencing and rent weighted orange construction barrels. A full field survey was conducted in order to place the fencing and gating at the most appropriate locations. Before tickets could be printed and distributed, the city needed to determine exactly how many tickets could safely be sold. The fire marshal worked with geographic information systems experts in the engineering department to determine a maximum capacity for the six-block area. Their calculations showed that a crowd of 50,000 on State Street would result in three people per square meter, building to building, for the entire length of the street. This data immediately called into question the accuracy of prior crowd size estimates, especially those nearing or exceeding 100,000 people. Based on this relatively simple calculation, the fire marshal authorized the sale of up to 50,000 tickets. Tickets were priced at $5 and event planners sought out venues through which to sell them. A handful of retail establishments volunteered to serve as ticket outlets on a consignment basis. The parks department developed a plan to use its utility vans and a construction trailer as ticket booths on the day of the event. Additionally, planners developed a satisfactory method for distributing complimentary tickets to residents and merchants within the event area. Entertainment. A critical piece of the Gating Plan was the reintroduction of live entertainment to the event. In earlier years and under the sponsorship of the student government, entertainment was a common component of the annual Halloween party. Once the party lost its formal sponsors, entertainment was tried only once, in 2003, with questionable impact. The Gating Plan called for entertainment to be brought back to the event as part of the effort to create a concert-style environment for the partygoers.

09_Madensen_Ch09.qxd:Madensen

178

1/3/11

10:39 AM

Page 178

PREVENTING CROWD VIOLENCE

Ideally, a private professional event promoter would have seized the opportunity to help improve the tarnished image of the annual Halloween party. Not surprisingly, however, there was little interest from promoters in the 2006 efforts and even less interest from private corporations that might have provided sponsorship funds. With the assistance of several interested students, the mayor’s staff identified a local music promoter who agreed to obtain the necessary permits, secure the required insurance policies, and book local bands for the inaugural event. The only stipulation was that the city had to pay the costs for the event permits, insurance, stages, sound equipment, entertainment, gating supplies, and private event staff. After a lengthy planning process, a street use permit was secured and two stages were strategically erected. Care was taken to avoid placing entertainment in the perennial flashpoints in the 500 and 600 blocks. The bands were selected from the local music scene, with a total cash payout of only $300 for all thirteen acts. In an attempt to offset the effects of alcohol consumption, event planners arranged for licensed food vendors to set up in a food court at the end of State Street nearest the university. Event staff. As part of the full gating and ticketing approach, the city pursued private event security staff to serve as the first points of contact for partygoers as they entered the event area. During the planning process, students generally acknowledged that they were comfortable dealing with concert-style security and would prefer not to see sworn police officers immediately upon entering the event area. Police officials wholeheartedly agreed, and an official request for proposals was released in late summer. A local security firm was selected to provide staff for ticket-taking, screen partygoers as they entered the event, provide security around the entertainment stages, and generally supervise partygoers. Unfortunately, the firm chosen was ultimately unable to produce the required number of staff. Fortunately, the city was able to supplement the available event staff with city employees under a standard on-call system. When the gates opened, partygoers were welcomed to the event by private event staff commingled with truck drivers, equipment operators, common council members, and mayoral staff members. Branding. The single most important aspect of the effort to change the Hal-

loween party was based not in policing but in public relations. After event planners committed to the Gating Plan, mayoral staff members had many conversations with local media and marketing experts. Initially, none was interested in taking on the Halloween project. Eventually, a group of young marketing and advertising professionals called Ad2-Madison developed a name, logo, and advertising campaign for the event on a pro bono basis. Ad2-Madison created “Freakfest,” an irreverent and catchy name about which the mayor was initially

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 179

TRICK OR TREAT?

179

dubious. Working with city staff members and interested students, Ad2 developed a full-color logo, press packets, and a grassroots marketing campaign designed to draw attention to the event’s costumes and festival atmosphere.9 The marketing itself was aimed at students from area colleges and included posters, flyers, stickers, chalking, and many forms of electronic media. While some community members disliked the new name for the Halloween party, it clearly indicated that the event was being redefined. Displacement concerns. During the discussions surrounding the Gating Plan, the possibility of displacement was commonly raised as a reason not to make fundamental changes to the Halloween party. Some community members were concerned that the individuals interested in causing trouble would simply avoid the State Street area and converge on another part of town. As if on cue, a handful of UW-Madison students started a Facebook page devoted to purposely displacing the Halloween party to Langdon Street, a residential street one block from State Street, full of fraternities, sororities, and private student rental properties. The organizers of this “Halloween on Langdon Street” initiative did not themselves live on Langdon Street and failed to garner the support of those students who did. Through direct intervention by the district’s common council member, mayoral staff, and police commanders, the plan to move the Halloween event quickly fizzled out. Interestingly, the individuals who started this effort ultimately established a formal student organization to assist Ad2Madison in marketing the legitimate party. As a precaution, police established an extra presence on Langdon Street during the main Halloween event on State Street. As the Freakfest weekend approached, media attention focused on the myriad of opportunities for the plan to break down. Some community members worried that the new plan might backfire, and some university officials openly predicted failure. The event. On Friday, October 27, mounted officers arrived on State Street at

8:55 P.M. At approximately 9:30 P.M., police began making arrests in the student neighborhoods. By 10:10 P.M., a crowd began forming in the 500 block of State Street. At 12:10 A.M., the traffic plan was deployed. By 2:15 A.M., officers were able to move pedestrians back onto the sidewalks and begin the process of ending the event. At 3:15 A.M., all officers specially assigned to the event were dismissed and the command post was closed. On Saturday, October 28, the command post was reopened by 6 P.M. and the full traffic plan was put in effect by 6:55 P.M. Signage problems caused minor backups at several entry gates. By 9 P.M., more than 22,000 tickets had been sold. At about 11 P.M., police made eighteen arrests at a private residence

09_Madensen_Ch09.qxd:Madensen

180

1/3/11

10:39 AM

Page 180

PREVENTING CROWD VIOLENCE

hall after projectiles were thrown out of windows. By 12:40 A.M., the county detoxification center was full. At 1:05 A.M., officers reported being hit with small thrown objects in the 500 block of State Street. At 1:20 A.M., police intercepted a group of partygoers wearing gas masks as they walked toward State Street. At 1:45 A.M., a group began forming in the 500 block. At 1:58 A.M., mounted officers reported being hit with small thrown objects, but immediately thereafter, a State Capitol Police mounted officer defused the crowd tension by snatching a strand of party beads from the air, twirling them above his head, and placing them around his neck. This simple, friendly, humorous act effectively quelled the potentially dangerous behavior of several hundred people in the 500 block of State Street. By 3:00 A.M., the Special Events Team was dismissed. Post-event assessment and planning. For the first time since 2001, no chemical munitions were used to disperse an unlawful crowd. For the second straight year, no officers or citizens were injured during the event. No damage was reported within the event area. Despite an increase in police staffing levels, the city’s net cost for the event increased by only 2 percent over 2005, thanks to the recovery of $144,300 through ticket sales. With an overall attendance estimated at 34,000, the crowd was smaller than in years past. Arrests decreased 50 percent from 2005, with 233 individuals being arrested for 374 separate charges (342 ordinance violations, 16 misdemeanors, 7 administrative code violations, and 2 felonies). Of the arrestees, 71 percent had an alcohol intoxication reading of .08 or higher. The aggressive stance police took toward house parties seemed successful as officers reported fewer large, unruly gatherings. Police commanders reported that the adjustment of the fencing materials to prevent a backflow of partygoers to the 500 and 600 blocks of State Street was effective. Furthermore, the gating mechanisms and the use of private event staff seemed to help reduce the prevalence of carry-in alcohol.

Halloween (and UW Homecoming) 2007 Pre-event planning. After what was widely considered an enormous success in

2006, event planners had an easier time improving the event plan. Most critically, the successful 2006 event encouraged local businesses to openly express support for the event and assume greater responsibility for organizing and managing it. A reputable professional promotion company based in Madison stepped forward to organize the entertainment and handle the ticketing duties. This promoter secured significant private funding for the event, further offsetting the public cost. While certain local media reporters continued to draw attention to the event’s potential drawbacks, the promoter and the city were able to garner positive media attention for Halloween for the first time in nearly a decade.

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 181

TRICK OR TREAT?

181

Some longstanding elements of the event plan such as the arrest processing center, remote-controlled surveillance cameras, mounted officers, and lighting trucks were preserved, and only a handful of modifications were planned for 2007, including the following: • Close attention was paid to the selection of a different private event staffing firm. • More attention was given to promoting the event in a positive light via local media coverage, grassroots public relations, and professional marketing. • Entertainment options were expanded by adding a third stage. • The day-of-event ticket price was raised from $5 to $7 both to encourage pre-event sales and to enable the promoter to get higher-caliber music entertainment. • The food court was expanded. • Pedestrian-level event signage was improved. The event. On Friday, October 26, no significant incidents were reported in or around the State Street area. The traffic plan was not implemented and officers were dismissed just after bar closing time at 2:30 A.M. On Saturday, October 27, the University of Wisconsin football team played its homecoming game in the afternoon, resulting in a huge crowd of fans on and near State Street during the early evening hours. The command post was operational by 5:50 P.M. and the traffic plan was fully implemented by 7 P.M. At the same time, more than 100 private event staff members familiar with concert environments were deployed throughout the event area to handle ticket-taking and entrance screening. By 11:30 P.M., officers reported several small house parties in adjoining neighborhoods. At 12:25 A.M., the main entry gates were adjusted to allow easier ingress by partygoers. At 1:00 A.M., the entry gates were transitioned to exit-only gates and the on-stage entertainment began its tiered conclusion with the band on the stage closest to the 500 block of State Street ending first, and the band on the main stage near the State Capitol ending last. At 2:00 A.M., daylight savings time ended and the clocks were turned back to 1:00 A.M. At 1:02 A.M., the street sweepers began cleaning the streets. At 1:28 A.M., the SET officers were dismissed. At 1:36 A.M., the traffic posts were closed and the streets reopened. Post-event assessment and planning. By the end of the event, 179 individu-

als had been arrested on 259 separate charges (247 ordinance violations, 9 misdemeanors, and 3 felonies), representing a 23 percent decrease in arrests compared to 2006. Three-fourths of the arrestees had an alcohol intoxication reading of .08 or higher. Eighty-nine arrestees were college students from

09_Madensen_Ch09.qxd:Madensen

182

1/3/11

10:39 AM

Page 182

PREVENTING CROWD VIOLENCE

thirty different colleges. Thirty-eight of the 89 (43 percent) reported being UW-Madison students. Public costs for the event finally began dropping thanks to a combination of more efficient police deployment and an increase in ticket revenues. The total net public cost decreased 7 percent from 2006, but still amounted to more than $577,000. Halloween 2008 Pre-event planning. The private promoter of Freakfest in 2007 again

obtained a street use permit, secured special event insurance, and entered into a mutually beneficial cost-sharing agreement with the city for the 2008 event. The entertainment options continued to grow, with three stages and a giant inflatable movie screen being planned for the party. Several weeks before Halloween, city officials learned of another attempt to move the party, this time to Mifflin Street, a neighborhood dominated by student rental housing and which is the site of another annual student street party that regularly demands police attention.10 Again, the individuals promoting the displaced party did not actually live in the neighborhood they were targeting. UW-Madison administrators, the common council, mayor’s office, and police commanders quickly intervened. Again, the students actually residing on the targeted street made known their displeasure with the notion of having 40,000 partygoers on their front porches. This second attempt at displacement was quickly thwarted. While the event planners had previously dealt with daylight savings time and home football games as confounding variables, 2008 threw yet another variable into the mix: Halloween fell on a Friday but the event was being planned for Saturday. Early in the planning process, officials discussed whether to switch the party to Friday to accommodate the actual Halloween holiday. This suggestion was quickly rejected, based heavily on the city’s desire to limit the annual party to one day. Moreover, the event promoter argued that a Saturday event allowed it more planning flexibility. Police planned for the standard deployment of officers for the Friday before Freakfest. Based on 2007 postassessment recommendations, several cost-saving measures were implemented in 2008, including reducing non-Madison police staffing. Furthermore, only one portable light truck was used, with planners instead relying on the lighting from the three stages to supplement the existing light sources. The event. Friday, October 31, saw reasonable crowds on State Street. The traffic plan was not implemented, and the media reports took note of the unusually creative costumes.

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 183

TRICK OR TREAT?

183

Saturday, November 1, began as the two previous Freakfests had begun: the command post was operational by 6 P.M. and the private event staff were in place by the time the gates opened at 7 P.M. With a national headline act playing on the main stage, ticket prices were raised slightly and larger crowds were expected. Throughout the evening, officers reported friendly crowds and few arrests. The evening’s most tense moments came when the headline band began to play and thousands of people rushed to the area surrounding the stage. To alleviate the pressure created by the crowd, the police scene commander ordered a quick modification to the fencing near the stage. The fencing mechanism was manipulated to allow more space for the crowd near the main stage and immediately resulted in a safer, more-controlled environment. Once the pressure was relieved, officers and event staff reported that the festive crowd left the event area peacefully at the end of the show. Post-event assessment and planning. For the fourth consecutive year, no offi-

cers or civilians were injured during the event. For the third consecutive year, there was no unlawful assembly and no chemical munitions were required to disperse the crowd. A city-owned information kiosk was damaged near the main stage but was repaired by 9 A.M. the following morning. Gross expenditures for the Madison Police Department were down nearly 8 percent thanks to a tightly structured deployment plan, and estimated partner agency costs were down 3 percent. Once the city’s share of the increased ticket revenue was accounted for, the 2008 net costs represented a 12 percent decrease from 2007. Police continued to make fewer arrests. In 2008, 135 arrests were made for 195 individual charges (184 ordinance violations, 7 misdemeanors, and 4 felonies), representing a 25 percent decrease from 2007. Just more than half the arrestees (53 percent) had an alcohol intoxication reading of .08 or higher. While still remarkably high, this figure is notably lower than the portion of arrestees at this level of intoxication in 2006 and 2007 (71 percent and 75 percent, respectively).

Reflections on the Problem-Solving Process The City of Madison’s efforts to address its Halloween event problems did not start as a formal problem-oriented policing project, but it took the form of one— in fits and starts. The project was implemented within the SARA problemsolving framework. Problems were identified through careful and detailed scanning during postassessment evaluations. Analysis of the factors contributing to the identified problems guided the selection of appropriate interventions. Responses selected to reduce problems were adopted, modified, or eliminated each year based on assessments of effectiveness.

09_Madensen_Ch09.qxd:Madensen

184

1/3/11

10:39 AM

Page 184

PREVENTING CROWD VIOLENCE

Despite previous concerns, the strategy used to facilitate a safe Halloween event on State Street did not result in the displacement of misbehavior to other areas of Madison. Although not confirmed quantitatively, qualitative evidence suggests that the most recent response plans provided some diffusion of benefits (Brantingham and Brantingham 2003). Discussions in the command post during the 2006, 2007, and 2008 Halloween events each included comments from commanders regarding the “unusually quiet” Saturday nights citywide.

Lessons Learned Throughout the seven-year process discussed in this chapter, many assumptions were made, many theories were field tested, and many failures were recognized. Over time, these failures were used to develop a successful model for dealing with a problem that seemed insurmountable just a few years prior. Along the way, four fundamental lessons were learned. First, it is possible for a community to be in general agreement that an annual event is problematic and that there should be significant change, and yet be reluctant to act. For many years, bar and liquor-store owners, students, college administrators, politicians, and private sector leaders resisted notions that there may be some role for them in addressing the Halloween problems. This was, some believed, simply a problem for the police to handle. Only after these stakeholders were actively engaged in the problem-solving process for several years did they begin to take ownership of the problem and assume responsibility for its solution. Second, it is important to explore all feasible options for addressing a problematic event. By proposing a drastic response that proved unattractive to many stakeholders, the city persuaded stakeholders to give more serious consideration to the slightly less unattractive alternative. No one believed the mayor was prepared to charge property owners to recover the public cost of the Halloween party until he unveiled a specific and complete plan to do so. Despite the disparate opinions about how best to handle the party, proposing the Charge Plan compelled key stakeholders to support the alternative Gating Plan. Third, a lack of empirical data analysis hampers the search for viable alternative responses and makes it more difficult to persuade stakeholders to become engaged in problem-solving efforts. Absent data indicating how many UW-Madison students were being arrested during the event, the dean of students would not likely have dedicated staff to the arrest processing center. Without analysis related to chronic house-party addresses, property owners and tenants would not likely have been convinced that they were at risk of incurring significant fines. Without the benefit of even the simplest event cost analysis, common council members would have been hard pressed to support the radical notion of closing six blocks of a public street for a Halloween party.

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 185

TRICK OR TREAT?

185

Over the years, better data specific to the Halloween event in Madison were collected and analyzed, including more-detailed accounting of arrests, alcohol intoxication readings, school affiliations, ages, costs, and house-party addresses. There remains, however, room yet to further improve data analysis. While some informal surveying was conducted in 2007, better instruments could be developed to better understand how partygoers, business owners, and residents perceive the event. These data might lead to better collaboration in terms of event logistics, better marketing of the event, or enhanced community support for the party. Strategic interviewing of offenders might yield better information about where alcohol is procured and consumed, how decisions to attend and to offend are made, and what measures might be effective at preventing and deterring offending. Fourth, and perhaps most importantly, the solution to the problems related to the annual Halloween party could not be solved through police efforts alone. This assertion is not a surprise to those who work in the public safety arena, but this problem-solving initiative illustrated the importance of a comprehensive effort. Rather than continuing to police the event through conventional crowdcontrol approaches, event planners set out to change crowd behavior. The environment was changed from an unsponsored free-for-all to a concert-style setting that was both familiar and fun. Expectations were clearly communicated through well-understood mechanisms: ticketing, gating, posting of rules, defined beginning and ending times, and the presence of concert-style event staff. Experts in event promotion, production, advertising, and public relations were engaged in the process. By legitimizing, rather than squelching or terminating, the event, the net result has been to all but eliminate the harm to the community.

Conclusion In 2002, the Madison community and its city government were caught flatfooted by the Halloween party. The event was understaffed, available intelligence was ignored, and nonpolice guardians were all but absent. Beginning in 2003, vigorous, earnest, and ultimately successful planning efforts were undertaken. The early efforts focused almost solely on the police and the possible responses that they could implement. Through the scanning and early analysis phases, the city became adept at predicting when and where the Halloween-related problems would manifest. Event planners began to collect data, seek community input, and systematically improve their understanding of what was happening at the event and why. After several cycles of implementing packages of well-intended but ultimately ineffective responses, the entire event was reconceptualized in 2006 in an effort to regain control of the party and more broadly distribute responsibility for the event throughout the community. This entailed involving more stakeholders, seeking their input and

09_Madensen_Ch09.qxd:Madensen

186

1/3/11

10:39 AM

Page 186

PREVENTING CROWD VIOLENCE

soliciting their assistance, and then holding every stakeholder accountable for their own piece of the response package. It is admittedly difficult to determine with scientific certainty that the interventions adopted over the seven years discussed caused the recent positive outcomes, but it is also difficult to refute. Given that the focus of the 2006 plan was specifically to change the physical and social nature of the event, it is unrealistic to assert that the implemented responses had no effect on crowd behavior. Every Halloween will present new challenges for Madison’s Freakfest event planners, but each year that the event proves festive and peaceable, rather than riotous and dangerous, helps solidify its new reputation and shed its old one. To Halloween’s defining question, “Trick or treat?” the Madison community can once again answer, “Treat.”

Appendix A: Halloween on State Street— Responses Implemented by Year (2002–2008)

Sanitize event area Fire department scans area for combustibles Arrest conveyance teams Bars asked to limit glass containers Traffic plan Dumpsters Port-a-potties Command post Temporary cab stands Meetings with bar owners Additional officers Chemical munitions Stakeholder planning group Arrest processing center Special event team officers Definition of event goals Pro-arrest enforcement policy Multi-agency staffing plan Remote-controlled security cameras On-street videotaping Street sweepers to signal end of event Live entertainment Website (traffic, lodging, parking, behavioral expectations) Glass ban Stadium-style lighting Promote event as local, one-night party Increased forfeiture amounts (disorderly conduct, underage alcohol, open container) Partnership with Tavern League and liquor distributors regarding information distribution

2002

2003

2004

2005

2006

2007

2008

x x x x x x x x x x x x

x x x x x x x x x x x x x x x x x x x x x x x

x x x x x x x x x x x x x x x x x x x x x

x x x x x x x x x x x x x x x x x x x x x

x x x x x x x x x x x

x x x x x x x x x x x

x x x x x x x x x x

x x x x x x x

x x x x x x x

x x x x x x x

x

x x x

x x

x

x x x

x

x x x x

x x x x

x x x x

x x x x

x x x x

x

x

x

x

x

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 187

187

Appendix A: cont. 2002 Letter to rental property owners, hotel guests, private residence halls, regional colleges, and high schools Pre-event press conference Temporary public address system House party plan developed and implemented Free food distribution University Inn used as police staging area Mounted officers Monitor online advertisements and social pages for related intelligence Publicize possibility of ending the event Mechanism to employ emergency powers followed Formal problem-oriented approach deployed Prerecorded messages for use on public address system Formal assessment of legal options to address disorder related to annual party Paramedics deployed on foot throughout event No-guest policy within all UW-Madison dormitories Some strategic use of flexible fencing to restrict pedestrian access Problem analyst in command post City buses used to stage arrested individuals Bars asked to utilize rear exits Bars asked to voluntarily close at the “first” 2:30 A.M. Develop cost-recovery model Partner with media Rebrand the event (via advertising and marketing experts; name, logo, press conferences, etc.) Establish event perimeters Deploy signage Use private security/event staff Ticket event Street use permit (including special event insurance policy) Food court outside event area Promoter to coordinate entertainment Optimize staffing and deployment plans for efficiency Scale back non-Madison law enforcement involvement Private sponsorship

2003

2004

2005

2006

2007

2008

x

x

x

x

x

x x x x x x x

x x x

x x x

x x x

x x x

x x x

x x x

x x x

x x x

x x

x x

x x

x x

x x

x x

x x

x x

x

x

x

x

x

x

x

x

x

x

x

x

x

x

x x x x

x x

x x

x x

x x x

x

x

x x x x x

x x x x x

x x x x x

x x

x x x

x x x

x

x

x

x

Injuries

Property damage

Arrests

Afraid. “Riot.”

2006

2007

2008

1 officer; no significant injuries to civilians

No significant injuries to officers or civilians

No significant injuries to officers or civilians

None

None

6 businesses (5 broken windows, 1 broken door)

1 broken window

None

None

1 broken kiosk case near main stage

578

468

233

179

135

$255,536 MPD; $234,009 partner agencies; $489,545 net

$349,092 MPD; $255,753 partner agencies; $604,845 net

$376,900 MPD; $386,822 partner agencies; $763,722 gross; $144,300 revenue; $619,422 net

$276,498 MPD; $402,315 partner agencies; $678,813 gross; $101,717 revenue; $577,096 net

Appalled. Police reacted too slowly. “Madness.” “Crazy.”

Worried that this is “the way it will always be.” Police reacted too quickly. “Public drunkfest.”

Frustrated and concerned about rising costs. Acknowledge continuous improvement. “Let those who party pay the bills.” “There has to be a better way.”

Skeptical at first— will the new plan work? “Success.” “Rebranded.” “Peaceful.”

Some are still skeptical—was 2006 a fluke? “Freakfest 2, Rioters 0.”

House parties begin to be recognized as contributor to end-of-event problems.

House parties addressed via formal teams (police).

House parties addressed via formal teams (police and fire).

House parties addressed via formal teams; concerted effort to affirmatively displace party to Langdon Street, one block from State Street, thwarted by stakeholders.

House parties addressed via formal teams; neighboring residential areas relatively quiet.

$255,150 MPD; estimated $390,000 partner agencies; $645,150 gross; cost recovery $137,602; $507,548 net Embrace the rebranded event. Begin to focus on the positive aspects of the safe and fun party. “Three-peat.” “Tamed.” “Turned the corner.” House parties addressed via formal teams; concerted effort to affirmatively displace party to Mifflin Street, five blocks from State Street, thwarted by stakeholders.

Page 188

Displacement Unlikely but unknown; flashpoint was the epicenter of event; no formal attention given.

2005

10:39 AM

Community support and media sentiment

13 officers; unknown No significant civilians injuries to officers or civilians 12 businesses; 7 businesses; $24,000 in damages; approximately $8,000 in looting $13,000 in damages; no looting Several dozen 406 (specific data not collected) $77,753 Madison $101,823 MPD; no Police Department data collected for (MPD) partner agencies

2004

1/3/11

Public costs

2003

188

2002

09_Madensen_Ch09.qxd:Madensen

Appendix B: Assessment of Six Success Indicators (2002–2008)

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 189

TRICK OR TREAT?

189

Notes 1. For the uninitiated reader, a “mosh pit” is an area in front of a stage where concertgoers “dance” by slamming their bodies into one another. “Moshing” is supposed to be fun, but obviously, people can get hurt doing it. 2. The first message thanked the partygoers for attending the festivities and informed them when it was time to leave. The second message was to declare an unlawful assembly and to order people to leave the area immediately. 3. Fencing materials consisted of orange snow fencing fastened to orange highway construction barrels weighted with sand. 4. This house party ultimately led to the revocation of the host fraternity’s charter. 5. Under a city attorney’s policy adopted in 2005, a significant portion of the fines were commonly held in abeyance and dismissed if there were no repeat offenses within a set time period. 6. The special charge is expressly not a special assessment, a mechanism used primarily to fund capital improvements. Wisconsin case law provides that public safety services are covered by the statute, but notes that such services need to be actually performed, not simply be available for performance. Town of Janesville v. Rock County, 153 Wis. 2d 538, 451 N.W.2d 436 (Ct. App. 1989). 7. See Koss v. State, 217 Wis. 325, 258 N.W. 860 (1935) and Cassidy v. Ceci, 320 F. Supp. 223 (E.D. Wis., 1970). 8. See, Wis. Stat. §§62.09(13), 62.11(5). 9. For their work on Freakfest, Ad2-Madison won multiple national awards. 10. The annual Mifflin Street Block Party occurs each May as an impromptu endof-school celebration for university students. It sparked serious rioting in the late 1960s (largely from anti–Vietnam War protesting) and again in the 1990s (largely from alcohol-fueled revelry). Interestingly, for the first time in several decades, the 2009 Mifflin Street Block Party was a legitimate and legal party with a local sponsor who secured a street-use permit, set up a stage in the street, and utilized food vendors throughout the event area. The editors of the local daily newspaper praised the improvement—modeled after the Halloween success—shortly after the event. See Wisconsin State Journal, Tuesday, May 5, 2009, p. A10.

09_Madensen_Ch09.qxd:Madensen

1/3/11

10:39 AM

Page 190

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 191

10 Dialogue Policing: A Means for Less Crowd Violence? Stefan Holgersson and Johannes Knutsson

The Salem demonstration, December 8, 2007, 12:46. About 200–300 demonstrators have gathered on a football ground, preparing for their march. One of the organizers greets them with a welcome speech: “It is nice to be here, that is, not at all. At four o’clock this afternoon in the borough of Salem the Nazis will march. . . .” (A number of speeches with strong political leftist character followed.) 12:59. Two social workers in yellow jackets inform a police officer in civilian clothes and a yellow vest marked “Dialogue Police,” that they just have seen two slightly drunk young males, one of them a wellknown delinquent. They are making Hitler greetings on a slope above the demonstrators. The officer phones his colleague, who is among the demonstrators with an organizer, to inform them that uniformed police will become visible when they take care of the potentially disturbing and provoking persons. Is that agreeable? Seemingly after a confirmative answer, three uniformed police officers approach the young males. They are requested to leave the area and are helped by the police to climb up the slippery and steep slope. One of the officers picks up a bottle from the ground and empties it when they are gone. He leaves at 13:08. Stockholm, the Swedish National Day, June 6, 2008, 17:50. A uniformed police officer in a yellow vest marked “Dialogue Police,” leaves the mobile command center and marches away into the crowd of nationalists, some of them neo-Nazis. It is estimated to be about 600 of them. They have ended their march and speeches are made. The officer—the coordinator of the dialogue police officers—is on his way to inform the organizer that those who have made Hitler greetings (about 100) will be arrested. His question is: “Will it be carried out in an orderly fashion or will it be messy?” The answer is: “Your potential for violence is currently greater than ours, so it will be orderly.” 18:30. Three large buses with the arrested persons leave for the police headquarters for processing and investigation.

THESE ARE EXCERPTS OF FIELD NOTES FROM AN EVENT WHERE ANTIracists (in police jargon, “the left”) demonstrated against nationalists (in police jargon, “the right”), and from a right-wing demonstration. In the first instance, the police commander had beforehand, via the dialogue police, made an agree191

10_Madensen_Ch10.qxd:Madensen

192

1/3/11

10:25 AM

Page 192

PREVENTING CROWD VIOLENCE

ment that, as long as the protesters demonstrated in an orderly manner, uniformed police would keep out of sight during the march. The fundamental idea is to create self-policing among the protesters. In the other, the main task for the police was to keep the left and right apart to prevent violent attacks from leftists, protecting the right to a lawful demonstration. But, as shown by the example, illegal behaviors are not tolerated. In both cases, dialogue police officers, well known by the organizers, acted as a communication link between demonstrators and police commanders. In the new Swedish approach for crowd management and control, dialogue policing is an integrated and crucial component. In this chapter, we will describe its background and explain its principles. The preventive aspects of the tactics will also be explored. Finally, we will discuss some dilemmas of dialogue policing.

Background—The Riots in Gothenburg, 2001 As a member state of the European Union, Sweden was responsible for a number of EU activities in 2001, including the European Union summit that took place in Gothenburg, June 14–16. A considerable complication was a visit by President George W. Bush. A lot of different groups, many of them representing left-wing opinions and some of them nonparliamentary, anarchistic, or “autonomous,” wanted to use the opportunity to protest against the EU politics and against President Bush. This was, of course, an enormous challenge for the police—both to secure the meeting and the visit of the president, as well as to protect the fundamental democratic right to demonstrate and the freedom of speech. During the meeting, about 25,000 people demonstrated in different places. The meetings could, on the whole, be carried out as intended and the demonstrators’ planned marches were for the most part carried out in an orderly manner. However, some of the protesters were extremely violent and severe riots occurred in the center of Gothenburg with vandalism and assaults of the police, where large amounts of paving stones were thrown at them. In one critical incident, there was a mass arrest of 459 persons (see, e.g., Peterson and Oskarsson 2006). During the event, three demonstrators were injured by police gunfire. About 150 people ended up needing medical care, of whom about 50 were police officers. In the aftermath, a number of protesters received jail sentences, the chief commander of the police operations was tried in court but acquitted for the decision to make the mass arrest, and in a parliamentary report the national police commissioner was criticized for his inaction, having not coordinated police resources (SOU 2002). To summarize, the event became a national trauma.

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 193

DIALOGUE POLICING

193

To understand what occurred during the event and the processes behind the actions of the police, knowledge of some characteristics of the Swedish police will be helpful. Sweden has a national police service, responsible to the Ministry of Justice. The police are organized in twenty-one independent police forces. The National Police Board has an administrative and supervisory function for the police service and is headed by the national police commissioner who is appointed by the government. Among other things, the National Police Board is responsible for the development of new methods and technological support. According to its instruction, it also has a coordinating function for large events, where resources of one police force will not suffice. To plan for the event in Gothenburg and to act as chief commanding officer, the commissioner of the uniformed branch in the local county police force was appointed.1 Discussions with representatives from the board and the security police during the planning process were not successful and were carried out in what might be characterized as an antagonistic atmosphere. One major issue of disagreement was the need for reinforcement from other police forces. From the start, the chief commander assumed that he could manage the event with 500 officers (there are about 2,500 police officers in the local force). However, the summits in Seattle in 1999, in Prague in 2000, and in Nice in 2000 had all ended up with riots and great difficulties for the police. Information from the security police and other sources indicated that violent demonstrators planned to come to Gothenburg. Despite this information, the chief commander, just before the meeting was to begin, still hoped for a friendly event with smiling police officers. Since big events of this kind are rare in Sweden and local forces had developed their own methods of crowd management, there was no common, nationally understood tactic. Furthermore, many officers had no training in or experience of crowd control. However, it should be mentioned that during the preparation for the European football championship that took place in Sweden in 1992, police officers from different police forces had been trained in a common tactic. To sustain this competence, however, continual training and replacement of officers is required. This would be an obvious task for the National Police Board. However, the national police commissioner decided to take no action. In the end, when intelligence indicated a considerable risk for unlawful and violent actions of protesters, the chief commander asked for substantial reinforcement of officers from other forces, ending up with a total of 2,500 officers to police the event. Many were badly trained or not trained at all, and a large portion had no adequate equipment or no equipment at all for this kind of task. Even though Swedish police use commission tactics, where an objective is given and it is up to the officer in charge to solve the assignment, in practice, it was the chief commander who tried to exercise strict control. To make things

10_Madensen_Ch10.qxd:Madensen

194

1/3/11

10:25 AM

Page 194

PREVENTING CROWD VIOLENCE

worse, the radio communications system did not function during critical phases, partly because it was disturbed by activists. Even if it is easy to be wise after the fact, the way the event was policed seems to have been doomed to fail. Soon after the event, a number of investigations of the police were carried out, the most important by a committee appointed by the government (SOU 2002). The massive report, almost 800 pages long, consists to a large extent of a thorough description of what happened, from the decision to organize the meeting in Gothenburg, the planning, the different happenings during the event, to the aftermath. A large number of documents, as well as interviews with a sample of demonstrators and police officers with different ranks and duties, created the basis for the description. The report concluded with a number of critical remarks, especially concerning the passive role of the national police commissioner. Of special importance in this context is the conclusion that Sweden was in dire need of a common national tactic to police large events and to manage and control crowds. A nonconfrontational and de-escalating tactic was recommended, one that includes violence preventive strategies. To avoid mass arrests, it should further prioritize interventions against single persons. The officers working in crowd control should have a common training and a common command vocabulary. They also need knowledge of different political movements and their new manners of expressing opinions. Foul language directed at demonstrators by police officers should be banned. Commanders at different levels need adequate training. The report mentioned positively the dialogue that specially appointed contact police officers established with representatives of the protesters during the event. It was suggested that the new approach should include such a function. The Gothenburg Riots in a Research Perspective

In a research project, the Gothenburg event was followed up in three substudies (Granström 2002). In the first, attitudes among demonstrators and attending police officers were studied before and after the event, making possible the study of changes in attitudes as well as differences between the groups. The studied demonstrators were characterized as peaceful and the police officers as ordinary, not specifically trained for crowd control. In the second study, focus group discussions were used with three groups of demonstrators and two groups of police officers from special units trained in crowd control. And lastly, a media content analysis was carried out of how the event was depicted in five major daily newspapers. Two overriding social processes were identified in the analysis of the combined data from the different sources. These processes either support peaceful demonstrations and interactions between demonstrators and the police or the reverse—create or reinforce conflict and confrontations (see

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 195

DIALOGUE POLICING

195

Chapter 5 by Hylander and Granström). Processes aggravating conflict are provocation, creating chaos, and negative stereotyping. The mirrored processes for mitigation are disarming, peaceful organizing, differentiation, and positive stereotyping. In Gothenburg, the aggravating social processes were evident. The way the exclusionary tactic was practiced by the police, where they tried to control violent protesters, had a huge impact on social processes and affected otherwise peaceful demonstrators, changing their mind-sets. As pointed out by the authors of the Gothenburg study, this interpretation is in accordance with the Elaborated Social Identity Model, which predicts that otherwise peaceful protesters may be incited by forceful actions by the police, making them prepared to act aggressively toward the police (Reicher 1996; Stott and Reicher 1998a; Stott and Drury 2000). In the concluding remarks of the Gothenburg study, a prediction was made that demonstrations in the future would more often become violent because small militant groups had learned how to create chaos and to provoke by challenging nonviolent demonstrators and the police. That the police fundamentally would change their tactic was something the authors of the report could hardly foresee.

The Special Tactic for Crowd Control The National Police Board decided in 2004 to implement a new national approach for crowd management and crowd control—the Special Police Tactic. Altogether there are about 1,200 police officers trained in the tactic (Sweden has about 20,000 police officers). It is not a full-time assignment, which means that all officers, including the commanders, work in their ordinary functions until the need arises. Only the three largest forces with Sweden’s biggest cities—Stockholm, Gothenburg, and Malmö—have tactical units. All personnel are trained in their specific assignments, but the commanders are specifically educated in issues connected to demonstrations and crowd control. After completed training, the commanders are certified at different levels according to rank and assignment. An important principle of the tactic is to delegate decisions to adequate operative levels. Commission tactics are thus used where subordinate chiefs are given objectives, and within given preconditions, are working in the spirit of their commander, free to decide how to carry out the assignment. All Swedish officers are taught a National Basic Tactic. Two aspects of this tactic are of special importance for the Special Police Tactic. The aim is to teach officers to have a mental readiness for hazardous situations and a capability to cope with stress, and to form a communicative mind-set, with an ability to actively listen to counterparts and solve conflicts in de-escalating manners.

10_Madensen_Ch10.qxd:Madensen

196

1/3/11

10:25 AM

Page 196

PREVENTING CROWD VIOLENCE

Principles of the Special Police Tactic

The principles behind the Special Police Tactic can be described through a few key terms, namely facilitation, dialogue, counterpart perspective, differentiation, signal value, and state, which are explained in Table 10.1. The strategy is either deferred, where the police wait and see how events evolve, or offensive where police attempt to take charge over the event. To perform the Special Tactic, a dedicated organization has been formed that consists of highly mobile units of uniformed police officers, civilian arrest units, arrestee transportation units, and dialogue police officers. The core of the Special Tactic is the mobile concept that was more or less taken over from Denmark. Danish police had to develop a new tactic after an incident in 1993 where police shot and wounded eleven demonstrators during a riot in Copenhagen. The Danes, in turn, were inspired by the Dutch police. The basis of the mobile concept consists of secured vehicles, strong enough to withstand attacks from protesters, such as bombardments with paving stones. Each vehicle carries a group of uniformed officers. The officers protected inside can rest and swiftly be moved where they are needed. They can also stand behind the vans and be protected and be ready to intervene. An impor-

Table 10.1

Principles of the Special Police Tactic

Principle

Definition

Facilitation

Support demonstrators to reach legitimate goals with aim to prevent conflicts and to achieve self-policing.

Dialogue

To communicate, make agreements and mediate between commanders and demonstrators.

Counterpart perspective

To adapt police actions by considering possible reactions in crowds, attempting to avoid escalation and to achieve de-escalation.

Differentiation

To differentiate police actions according to conduct of subgroups and individuals in crowd.

Signal value

Through various means, display to crowds police readiness to use force.

State

There are three states indicating moods of crowds and level of risk for confrontation: 1. Green—a relaxed and controlled situation with no confrontation. 2. Yellow—a tense situation that may develop into a confrontation. 3. Red—confrontation with physical attacks and use of force.

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 197

DIALOGUE POLICING

197

tant task for the vehicles is to function as devices to physically funnel, stop, round up, or disperse demonstrators. Civilian arrest groups also use secured vehicles, although they are not marked. The main mission of these units is to apprehend demonstrators who commit offenses within a crowd. They enable the police to improve their capability to swiftly intervene against offenders, thereby minimizing the opportunities for them to influence the rest of the crowd or to get away by hiding in the crowd. Another essential reason for the creation of the groups is to avoid mass arrests. A mass arrest where crowds or parts of crowds are rounded up and put in temporary custody or shipped away from the sensitive place is a blunt instrument. Innocent persons who happen to be in the wrong place might be swept along. The arrest group is thus one means to get differentiation, where the police adapt their actions to specific conditions with the intention to avoid affecting a whole collective. The transport units take care of arrested persons, thereby freeing arresting officers to return to their respective duties. Through dialogue, the police want to help demonstrators reach their legitimate goals and avoid conflicts and confrontations. By entering a dialogue, conditions for demonstrations can be explained, discussed, and negotiated. A desired objective is to stimulate self-policing among demonstrators. During demonstrations, dialogue police officers work in pairs, normally in civilian clothes, although distinguishable by yellow vests marked “Dialogue Police.” Their key function in the course of a demonstration is to act as a communication link between demonstrators and police commanders (see next section for a detailed account). By using a counterpart perspective, where the police consider group processes between the police and demonstrators as well as within the police and among protesters, the police strive to avoid an escalation with increased hostility and aggression in the crowd as an effect of police actions. Both the deferred tactic and differentiation are methods to steer clear of an escalation. Within the approach, communication is a central element. Dialogue is just one of them. However, it is not only verbal communication that is used; other means are also employed. Through intentional and consequent use of signals, the police willingness to use force is continuously conveyed to the crowd. The visibility of vehicles, how they are positioned, if sirens and flashlights are used, formation of uniformed officers, how protection gear is used, and whether means of force are displayed have different signal values. The range stretches from indicating a calm and friendly situation to a readiness to exercise maximum force. During demonstrations, the police constantly assess the mood of the crowd and the risk for confrontations and outbursts of violence. A scale indicating different states is used, where green is a desired default value signifying

10_Madensen_Ch10.qxd:Madensen

198

1/3/11

10:25 AM

Page 198

PREVENTING CROWD VIOLENCE

a controlled and relaxed situation. Yellow indicates risk for confrontation and red indicates high risk or ongoing confrontation. The relationship between state and signal value might be illustrated by the way the uniformed officers use their helmets. In calm situations (green) the officers usually do not have helmets on. They are either stored in the vehicles or hang from the officers’ waistbands. If suitable, during the assembly of a march or its end point, the officers mingle in pairs in the crowd, trying to enter friendly conversations. In yellow states, the helmets might be put on and telescope batons may be drawn. As soon as a tense situation has calmed down, helmets are typically taken off and if batons are displayed, they are put back in the belt. The shift is done fast and simultaneously in the affected units, thereby immediately signaling to the crowd that the police are relaxed and in control. In red situations, the officers are engaged in or prepared for confrontations with helmets on, visor down, batons out, and so on. When in control again, police, through different visible means, display the condition to the crowd. The new tactical approach is dynamic in two senses. There is an effort to continuously adapt police actions to the current situation and to the objectives of the operation. Secondly, experiences of practicing the tactic contribute to its refinement. Through education of leaders and officers, new knowledge influencing practice is also incorporated.2 Another essential source for modification is, of course, changes in the behavior of protesters. The current execution of the tactic differs in many ways from when it first was put into practice. However, there is no guarantee that relapses into the old tactic or elements of it will not occur. One reason is that some from the “old guard” still are around.

The Dialogue Police Concept Dialogue policing has become a vital element of the Special Police Tactic. Its origin can be traced to Gothenburg and the riots during the EU summit in 2001. Much of the work in developing the model has since been carried out by the Stockholm police force. Early Experiences

Following the recommendation of the Gothenburg committee, the Stockholm police formed a dialogue policing group in early 2002. It consisted of a few officers working as police negotiators. At that time, there was much skepticism and resistance to the whole idea. The old police crowd-control tactic was still practiced while the process of deciding about its successor was ongoing. Dialogue policing was just added on to the old tactic. The initial creation of the unit might actually be seen as an exercise in paying lip service to requests by politicians, where police management could show that the police had tried dia-

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 199

DIALOGUE POLICING

199

logue, but that it was ineffective (Holgersson 2010; see also Holgersson 2005). One reason for this belief is that the appointed officers were supposed to work as dialogue police officers in addition to their normal tasks, and they hardly got the time necessary to work with dialogue. There had already been some experiences of dialogue between the police and right-wing groups when the unit began its work. The police officer in question, a negotiator who had been appointed to work with dialogue policing, continued his activity, but now had a formal position and support from his colleagues. The prospect for a meaningful dialogue with right-wing groups is favorable for two reasons. They often need protection by the police from attacks by left-wing groups during their demonstrations. Besides that, the structures of these groups are hierarchical, which means that there are leaders with authority to influence their members. To get a meaningful dialogue with the left-wing groups is generally more complicated. Lack of trust and negative opinions of the police in combination with the nonhierarchical organizational structure are huge obstacles. It often was (and still is) difficult for the police to find a counterpart for dialogue with these groups. Another massive problem surfaced when the dialogue police officers (Stefan Holgersson was one of them) tried to create a channel of communication between representatives of left-wing groups and the police commanders. Due to past experiences of dealing with them, there was almost a total lack of acceptance within the police organization for entering a dialogue. Police had for a long time associated some left-wing groups’ activities with a need to use massive force. Dialogue policing was a completely new way of thinking—to try to solve or reduce a problem by means of communication. Furthermore, many police officers perceived it as a sign of weakness. Threats to law and order should be met with force. As experiences of dialogue have accumulated, the number of police officers in the Special Police Tactic organization having this opinion has, however, decreased substantially. There was also an apprehension about the security of the dialogue police officers, since they moved in pairs close to or inside groups of demonstrators. It was considered too risky for them. To assess risk is, however, one of the tasks of the dialogue officers. If their experience tells them that the situation is threatening, they remove themselves. Early experiences indicated clearly that there was no foundation for the fear of their safety. In traditional police tactics, groups of demonstrators have commonly been treated as homogeneous. If a large group includes some who act violently and aggressively, all members are at risk of being treated as having violent tendencies. Instead of decreasing the number of persons acting violently toward the police, the effect can often be the reverse. Initially it was difficult for commanders and police officers to see their actions as part of the problem, namely that police tactics can make a group more homogeneous and thereby increase the inclinations to behave aggressively and destructively (Reicher 1996; Stott

10_Madensen_Ch10.qxd:Madensen

200

1/3/11

10:25 AM

Page 200

PREVENTING CROWD VIOLENCE

and Reicher 1998a; Stott and Drury 2000; Reicher et al. 2004; see also Holgersson 2010). Initial negative reactions from both sides were predictable when dialogue police officers tried to mediate between the commanders’ interests to control demonstrations and the goals of the demonstrators. A common attitude in the police organization was that dialogue police officers were some kind of traitors; they had taken sides with the demonstrators. It was seen as an example of the Stockholm syndrome, where identification with an extreme group made the dialogue police officers promote the protesters’ interests. Activists, on the other hand, perceived the dialogue police as an instrument to deceive them, enabling the police to obtain control. A widespread opinion was that dialogue police officers acted as intelligence officers, a view still strongly held in some groups. A significant reason for the gradual acceptance of the dialogue police function was positive experiences among commanders and demonstrators. The chances of fulfilling their respective objectives were generally higher when dialogue was practiced. To begin with, some commanders dared to try it out. When they found out that dialogue had positive effects—expected outbursts of violence for instance did not occur—other commanders’ willingness to try the tactic increased. Protesters, on their side, discovered that the focus could shift from their acts of violence to their message. Over time, the dialogue police unit has gradually acquired acceptance and a clearer organizational role and status within the Stockholm police force. When the new Special Tactic, where dialogue policing is an integrated function, started to be put into service, dialogue had already been tried and to a large extent approved. An indication of acceptance is the growth of the group from only a small number to twelve members and the appointment of a fulltime coordinator reporting directly to the commanders. Dialogue police officers are increasingly engaged in many different situations, such as riots, when groups make religious statements, in ethnic conflicts, and in other efforts to avoid, limit, or solve conflicts between different groups. In 2008, dialogue police officers were used in 100 missions by the Stockholm police. It ought to be emphasized that the experience of introducing dialogue policing in the police organization indicates that it requires police officers that have courage to stand up for their beliefs, being willing to challenge traditional decisionmaking within the police as well as conventional police wisdom. In a hierarchical organization, where orders are supposed to be followed without arguing, such a posture can be rather awkward. Sources of Inspiration for the Concept

The development of dialogue policing has to a large extent been influenced by practical experiences of officers. Because the first dialogue police officers

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 201

DIALOGUE POLICING

201

were trained and functioned as police negotiators, negotiation techniques played an important role in the first steps of creating the dialogue model. Dialogue police officers’ personal encounters with ethnic conflicts from international peacekeeping missions in the Balkans have also played an important role, especially experiences of mediating between groups with very different interests and opinions. Because a substantial number of the officers had studied at university, subjects like political science and psychology were also sources of inspiration. Research studies of crowd behavior have, of course, been crucial, and the Elaborated Social Identity Model has been of especially great value (Reicher 1996; Stott and Reicher 1998a; Stott and Drury 2000). Another source of research input has been the Field Research on Organization and Groups (FOG) from Linköping University that has been studying the interplay between protesters and police in a number of situations (see, e.g., Granström 2002). It must be emphasized that the dialogue concept is under continual development, where new experiences and knowledge contribute to its refinement. It has to be a dynamic concept since some protesters, especially those that are hostile to the police, will try to adapt to police tactics. Goals of Dialogue Policing

The ultimate goal of dialogue policing is to facilitate freedom of speech and the right to demonstrate. Its fulfillment will contribute to another important objective—to decrease confrontations between crowds and police and between opposing groups of demonstrators. The expected result is less violence against the police, less force used by the police, less destruction of property, and less violence between demonstrators. If effective, fewer members of crowds, as well as police officers, will be injured by acts of aggression and police use of force. However, in some instances, it is impossible to prevent acts of aggression and destruction, and for the police to abstain from use of force. Even when there is a considerable risk of an outburst of aggression and attacks against the police, and the dialogue officers have not been able to establish a dialogue with the demonstrators, dialogue policing can still have an important role. It might increase the chances that the police will be able to restrict their use of force, especially toward participants not looking for a confrontation, and to prevent acts of violence against police. If violence occurs after efforts to initiate dialogue, it is sometimes regarded as unsuccessful, even if the attempt might have reduced the extent of aggression. Such a black-and-white attitude is rather common in the police organization as well as in media reporting. External pressure on the police to act forcefully can be very problematic. Insistence on it risks influencing tactical decisions by commanders, and if they comply, it can result in long-term difficulties in creating dialogue with the

10_Madensen_Ch10.qxd:Madensen

202

1/3/11

10:25 AM

Page 202

PREVENTING CROWD VIOLENCE

affected groups. When the response is increased use of force, an escalation can occur instead of de-escalation. There are many historical examples of failures when the police have acted in this manner (e.g., see P. A. J. Waddington 1994; Bessel and Emsley 2000). External pressure can also affect the police organization’s way of adhering to the principles of human rights and the freedom of speech in negative ways (P. A. J. Waddington 1991, 1994; della Porta 1995; D. Waddington and Critcher 2000; see also Holgersson 2010). Principles of Dialogue Policing

Dialogue policing can be divided into five elements: negotiation, mediation, suggesting, communication, and monitoring (see Table 10.2). As will be evident, dialogue policing should not be seen as an isolated function. It is highly dependent on the rest of the police organization. There are some similarities as well as differences between ordinary police negotiation and negotiation carried out by dialogue police officers. One similarity is the fundamental principle that a commander never negotiates and that a negotiator never commands (see, e.g., FBI 1995; McMains and Mullins 1996). Dialogue police officers thus never make the decisions and commanders do not negotiate. A crucial difference is that police negotiators rarely have to negotiate with the same person twice. Trickery that is sometimes used by police negotiators to solve a difficult situation can work in a single episode, but not in a more ongoing relationship. Dialogue policing cannot succeed

Table 10.2

Five Fundamental Elements of Dialogue Policing

Function

Definition

Negotiation

Facilitate compromises and agreements between the police commanders’ interest in getting the most preferable tactical conditions and the demonstrators’ interest to get most preferable terms for their goals.

Mediation

Explain the police point of view to groups of demonstrators and the demonstrators’ view to the police in order to increase mutual understanding and to avoid negative stereotyping.

Suggesting

To come up with possible solutions to avoid or minimize risk for conflicts and confrontations.

Communication

In an accessible way, act as link between demonstrators and commanders in exchanging information.

Monitoring

Read moods and preparedness for action in the group of demonstrators and how that is affected by police activities, and inform commanders of consequences of different courses of action in a short- and long-term perspective.

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 203

DIALOGUE POLICING

203

without mutual trust and respect between dialogue police officers and demonstrators. One instance of distrust can destroy the possibilities for a successful dialogue for a very long time. By negotiating, dialogue police officers facilitate compromises and agreements between police and demonstrators. An example is a demonstration where the police for tactical reasons want the protesters to take a particular route and the march to end well away from a sensitive object like an embassy. The demonstrators want to use another route to get more attention and make their speeches closer to the embassy. By putting forward arguments from each side, it might be possible to make some adjustments or to come up with an alternative solution that is acceptable to both parties. A rather common arrangement with police and hostile groups is that uniformed police keep out of sight, provided that the demonstrators exercise control over the demonstration. The expected effect in both cases is increased willingness by demonstrators to police themselves: the organizers prevent disorder without interference of the police (Marx 1981; P. A. J. Waddington 1994; D. Waddington and Critcher 2000; Reicher et al. 2004). Since dialogue police officers know about police tactics and about the commanders’ plans in addition to understanding how demonstrators may react to different suggestions, the officers do not merely act as go-betweens: they can actively explain the possible consequences of different courses of action to both parties. The intention of mediation is to enhance mutual understanding of the counterparts’ motives for their actions and thereby increase acceptance and respect. By explaining to the demonstrators in detail why certain conditions for a demonstration have been made and what might happen if they are broken, it might be possible for the demonstrators to accept the imposed restrictions and thereby decrease the risk of a confrontation. Likewise, the dialogue police officers’ effort to describe and elucidate the reasons for a certain group’s demonstration and how it is carried out can create a greater understanding and acceptance of their behavior among police. “Group-thinking,” a phenomenon that otherwise can lead to negative stereotyping and counterproductive actions, might be decreased (Stenaasen and Sletta 2000). In their rather free role and through using or creating new contacts with activists or representatives of the local community, dialogue police officers can come up with or suggest solutions that will decrease conflicts. An example is organizing a meeting with local leaders representing a group of immigrants, trying to engage the local community in controlling rioting youths with an immigrant background. Before demonstrations start and during the demonstrations, dialogue police officers move in the crowds or get close to them. The officers wear civilian clothes but are easily identifiable in their yellow vests marked “Dialogue Police.” Especially in events with groups skeptical or hostile to the

10_Madensen_Ch10.qxd:Madensen

204

1/3/11

10:25 AM

Page 204

PREVENTING CROWD VIOLENCE

police, they may be the only visible representatives of the police. Their nonthreatening appearance in combination with their mere presence might very well have a calming and de-escalating effect. By being present, the dialogue police officers can function as direct communicative links between organizers of demonstrations or their contact persons and the commanders. An example is demonstrators who during a march have detected members of a hostile and opposing group close by. Instead of attacking them and trying to drive them away, they inform the dialogue police, who notify the commander, who in his turn might order uniformed police to remove the hostile group. The position within a crowd usually creates good opportunities to perceive police behavior like the demonstrators do, as well as to sense the atmosphere inside the crowd. Dialogue police officers do not act physically toward demonstrators and can therefore concentrate on assessing a situation. Since the officers often follow actions of a group of demonstrators closely during an extended period, it is easier for them to interpret an action by the protesters, compared to someone that enters into a situation unprepared and is able only to observe for a short period. The aim of dialogue police officers is to be near critical places, enabling vital information, assessments, and feedback about problems and the police way of acting to be continuously passed on to the commanders. Police actions and interventions can also be explained to demonstrators. An important function is to try to influence a plan or ongoing activity by the police that may be perceived as provocative by the demonstrators. In acute situations of attacks on the police, the dialogue officers withdraw. They continue, however, to observe and can still be reached. Commanders naturally need and request intelligence, something that can create pressure on dialogue police officers to deliver such information. To have a genuine dialogue, the dialogue police officers must be trusted by demonstrators. It is therefore imperative clearly to define the dialogue police function as different from that of gathering and disseminating intelligence. Dialogue Police Activities Before, During, and After a Demonstration

When describing and analyzing the role of dialogue police officers, it is helpful to distinguish activities before, during, and after an event. The different stages are closely related to each other, where the last phase mostly will comprise preparation for the next event (Holgersson 2010). Far too often, extensive police actions have been considered isolated events (Adang and Stott 2004). Typical activities before, during, and after a demonstration are summarized in Table 10.3. Before an event. It is essential that the police are sensitive when they make

statements in the media. A multifaceted presentation can have a positive effect

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 205

DIALOGUE POLICING

Table 10.3

205

Activities Before, During, and After a Demonstration

Before

During

After

Try to influence statements in media by the police.

Check that agreements are kept and facilitate new agreements.

Try to influence statements in media by the police.

Facilitate agreements between demonstrators and police.

Act as communication link between commanders and organizers of demonstration.

Get feedback on police operations from organizers of demonstration.

Exchange information between commanders and organizers of demonstration.

De-escalate by being present and prevent hazardous incidents or decrease their seriousness.

Explain police activities that occurred during the demonstration to demonstrators.

Create scenarios for commanders.

Read moods in crowds and among police.

Give feedback from police on behavior of demonstrators.

Involve and inform external actors.

Inform external actors and coordinate their activities.

Exchange information with external actors.

on the trust toward the police. Negative stereotyping, on the other hand, might worsen the prospects for an orderly demonstration (Granström 2002). The ease in making contact with organizers or other key persons differs strongly between groups. This means that the method to achieve a dialogue and the possibilities to reach agreements varies considerably between and within groups. Some are skeptical and even hostile to the police and have a loose nonhierarchical structure. Positive or negative prior experiences will naturally have consequences on the willingness of groups to enter a dialogue. However, the threat of police involvement can sometimes be an effective motivator for groups hostile to police to have a dialogue about how to accomplish self-policing. Irrespective of difficulties in making contact, it is important that the dialogue work starts well in advance of an expected problematic situation (SOU 2002; Abrahamsson and Johansson 2002; Oskarsson 2002, 2005). An important task for the dialogue police officers is to make scenarios. Their experience creates good possibilities to predict how an event might progress. Commanders can use this information to make improved, informed decisions. In some situations (for example, when there is a risk of riots by juveniles), it can be useful to engage external actors. The presence and commitment from others, such as religious leaders, sports leaders, social workers, local politicians, and different types of volunteer organizations can have a soothing effect. The dialogue police often try to arrange meetings in advance with such actors, informing them about and discussing plans and assumed scenarios.

10_Madensen_Ch10.qxd:Madensen

206

1/3/11

10:25 AM

Page 206

PREVENTING CROWD VIOLENCE

During an event. Normally the dialogue police officers meet organizers or other key persons when an activity starts. In Genoa, as in Gothenburg, communication between the police and the protesters often failed. Organizers were incapable of making contact with the police in crucial situations (della Porta and Reiter 2006). Dialogue policing is a means of avoiding such occurrences. Just as before a demonstration, the possibilities to enter a dialogue with organizers or other key persons differ during an event. A main task for the dialogue police officer is to act as a link between participants in demonstrations and their organizers and the commanders of the police operation. Even if a meaningful dialogue is not established, the dialogue police officers usually have an important role to play by being present. One important task for the dialogue police officers during an event is to inform and, as far it is possible, to coordinate activities of external groups such as local leaders along with other key persons and those from volunteer organizations. However, sometimes commanders fail to understand that the police cannot decide what these organizations should do. The dialogue police officers can present relevant information and emphasize what is sensible to do in a situation to minimize violence and vandalism. During large events, mass text messages can be sent via mobile phones where coordinators from different organizations and dialogue police can simultaneously inform participants from different organizations about what is occurring. After an event. In democratic countries, the media cannot be controlled, but

it is possible to appeal to them and to provide them with information. Media focus on a riot can, for example, result in an escalation, since it can encourage further violence (Granström 2002). It is important that the police steer clear of manipulated information and try to avoid messages that can be perceived as provocations. When a constructive dialogue has taken place, it is useful to point out that it was beneficial for all involved parties. Through meetings or other methods of contact with the involved parties, different viewpoints can be discussed and actions by the police explained. This can be an effective way to stop rumors that frequently spread in all phases of a demonstration. The stage after a demonstration is often preparatory for the next, where things that worked will be put into effect again or things that did not might be avoided. Some Words About the Dialogue Police Officers

As with all police officers in the organization for the Special Police Tactic, dialogue police officers are not engaged in their function on a full-time basis. (An exception is the coordinator in Stockholm who has a full-time position. Stockholm also has a special arrangement where the officers are able to set aside 25 percent of their time for assignment as dialogue police officers.) The ordinary functions of these officers vary—some work as uniformed police officers, com-

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 207

DIALOGUE POLICING

207

munity police officers, investigators, or civilian narcotics officers. It is interesting to note that four are former special weapons and tactics (SWAT) team members. One of them, with negative experiences from the Gothenburg riots, stated that he, in his wildest dreams, could not have foreseen that he would end up as a dialogue police officer engaged in discussions with protesters. In Sweden, it is a requirement for police in outdoor service to carry a service pistol. The officers need permission to be unarmed. Some of the dialogue police officers, including former SWAT team members, do not carry firearms. Their argument is that it is impossible to use firearms in crowds and that if attacked by violent demonstrators, their arms may get taken away from them. Besides, working unarmed underlines their mission and trustworthiness.

The Preventive Dimension Is the new Special Police Tactic more effective in preventing collective violence than its predecessor? Notwithstanding its importance, the question is very hard to answer. Simple comparisons with the former tactic, for example in numbers of injured protesters or police officers during demonstrations, are far too crude. To do this would, among other things, not take into account changes among demonstrators such as the establishment of new groups, their ideologies, and how that affects behavior during demonstrations. In common with many preventive schemes and measures, traditional evaluation methods and procedures are of limited value (Tilley 2006, 2009). One possibility is, by assistance of a well-founded prevention scheme, to identify the aspects of the tactic likely to give rise to the expected effects and then make a judgment as to whether the effects are likely to occur. For this purpose, the situational crime preventive approach, with its focus on the situation and how its characteristics affect perceptions and actions, will be applied. Since its conception in the late 1970s, the approach has been continuously refined, and there is now an extensive literature documenting its qualities (see, e.g., Clarke 1980, 1997b; Wortley and Mazerolle 2008; Guerette 2009).3 In the situational approach, there is a general reluctance to dwell on characteristics of offenders. Nevertheless, in this context and for this phenomenon, we argue that it is the proper thing to do, since protesters with various inclinations to behave violently act and interact in a common context where they affect each other. Depending on circumstances, even normally peaceful demonstrators may commit acts of aggression. Such circumstances are described by the Elaborated Social Identity Model, which focuses on dynamic inter- and intragroup processes where actions by police play a key role (Reicher 1996; Stott and Reicher 1998a; Stott and Drury 2000). There are five basic techniques in the situational approach that are manipulated in order to achieve crime preventive effects. It should be underlined that perception of the measures is essential. Offenses may thus be prevented by (1)

10_Madensen_Ch10.qxd:Madensen

208

1/3/11

10:25 AM

Page 208

PREVENTING CROWD VIOLENCE

increasing the effort, (2) increasing the risk, (3) decreasing the reward, (4) removing excuses, and (5) reducing provocations.4 Even though, as mentioned above, there is a disinclination to make a distinction between offenders, distinctions are nevertheless made between the expected effectiveness of the techniques for different categories of offenders. Thus, three “ideal types” of offenders are envisaged (Cornish and Clarke 2003, 56–79). The first category is the antisocial predator, who is portrayed as goal directed, mostly free of moral scruples, seeking to obtain satisfactions at least risk and effort. Situations are used for crime-commission purposes, where the offender selects situations for the opportunities they provide. Because of their motivation, the basic instrumental techniques best suited to prevent their offenses are to increase perceived effort, to increase perceived risks, and to reduce anticipated rewards. The second category, the mundane offender, is influenced by moral considerations in his or her decisionmaking. In essence, they are basically lawabiding people but may commit offenses that are easily neutralized or excused. Mundane offenders are in some situations released from moral reservations and permitted to offend. Compared to predators, they are reactive instead of proactive. They do not actively seek, manipulate, or invent opportunities, but may more passively choose to enter situations that may allow offenses to happen. Situational cues that affect permissibility and excusability are crucial, which means that removal of excuses might be an appropriate technique in order to prevent offenses committed by this offender category. And lastly, there is the provoked offender. In contrast to the two other types of offenders, who enter the situations with a high (predators) or a low (mundane offenders) degree of motivation, the provoked offender has no prior motivation to commit offenses. The motivation is evoked as reactions to averse stimuli in a situation, often due to an involuntary exposure. Without provocations, the offender will not be ready to offend and no crime will occur. By changing situations in order to decrease or remove cues perceived as provocations, offenses may be prevented. To return to crowds and violence, a crowd could very well be perceived to be composed of fractions of these three “ideal types” of offenders. The equivalent to predators would thus be notorious troublemakers looking for opportunities to act aggressively toward the police or toward protesters or others perceived as belonging to an opposing group. The mundane offenders would be demonstrators ending up in confrontations with the police or members of opposing groups, if circumstances allow it. And third there are the normally peaceful demonstrators who only under very special conditions commit acts of aggression. The troublemakers have a key role since they are looking for opportunities to create or to intensify confrontations, partly with the intention of inspiring other demonstrators to act violently toward the police or to cause damage. Whether the rabble-rousers are successful or not in their

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 209

DIALOGUE POLICING

209

endeavors depends to a large extent upon how the police respond to their actions. Situational Techniques

The situational techniques listed in Table 10.4 are those that, compared to the old tactic, are new or that are more strongly emphasized and systematically employed. The former tactic can be characterized as static; officers were to a large extent positioned in closed formations protected by shields and, when attacking demonstrators, indiscriminately used their batons as the primary means of force. One experienced operative commander pointedly summarized the tactic in the following way: “It was an aggressive wall of meat.” The lists are not exhaustive and, in common with such lists, it may, because of ambiguities, be difficult to align a particular technique to one basic technique. In some cases, similar or identical activities have ended up under the headline of two basic techniques. The reason is that they may be perceived differently by different groups of protesters or the police may have different intentions of using them, depending on context. Some techniques are performed by uniformed police or civilian arrest groups as part of the Special Tactic and some after dialogue between protesters and police. In many instances, the active party is not the police, but demonstrators or external actors. There are also examples of activities that depend on cooperation. The measures are highly context dependent. To give an example, in some instances a show of force can, on average, have a positive preventive effect, but in other circumstances, can provoke protesters, stimulating them to commit acts of violence. Compared to the old tactic where show of force was more or less a “one size fits all” measure, it currently is employed only after having weighed pros and cons. The techniques ought, thus, to be seen as sensitive instruments often in need of careful consideration and preparation before they are put into effect. This underlines the need to make assessments and followups to assess how effective they were. It follows from our arguments that the first three basic techniques are primarily aimed at militant aggressive protesters. Experience indicates clearly, however, the importance of employing mechanisms that prevent the fundamentally law-abiding demonstrators from becoming motivated to engage in aggressive acts. It highlights the importance of being careful about employing some of the first three basic techniques, since they may be perceived as provocations and facilitate aggression or neutralize inhibitions to strike against the police. This becomes evident by examining the techniques under the headlines decrease rewards, remove excuses, and reduce provocations, since they indicate that some techniques under the headlines increase efforts or increase risks can in varying degrees be counterproductive. Generally speaking, many of the

10_Madensen_Ch10.qxd:Madensen

Table 10.4

1/3/11

10:25 AM

Page 210

Situational Techniques Used in the Special Police Tactic

Increase Efforts

Increase Risks

Decrease Rewards

Remove Excuses

Reduce Provocations

Get demonstrators to follow a certain route, to have a particular assembling place and/or start time for a march that makes assaults from opposing groups difficult.

Have civilian arrest groups arrest offenders in crowds.

Make absence of visible uniformed police conditional on self-policing.

Promote selfpolicing to render rule breaking more problematic among sympathizers.

Beforehand, explain and elicit acceptance of police actions.

Remove objects that can be used as weapons from the assembling place, demonstration route, and/or end point.

Use special arrest techniques (rapidly jumping out of vehicles, apprehending, and immediately returning to vehicle) to intervene against offenders in crowds.

Ask organizers to publicly dissociate themselves from violent demonstrators.

Facilitate attainment of legitimate goals.

Afterward, explain police actions to demonstrators.

Put up checkpoints or perform sector patrol to make searches on demonstrators to find and remove objects that may be used as weapons.

Perform sector patrol and intense searches to detect protesters carrying dangerous objects.

Instruct police not to respond to provocations by confrontational protesters.

Make demonstrators perceive officers as individuals, by uniformed officers mingling in crowds.

Maximize use of signal values to convey calm and peaceful policing.

Urge protesters to use sticks for flags and banners that cannot be used as missiles.

Facilitate detection of movements of opposing groups by sector patrol or civilian police officers.

Present multifaceted image to the media, playing down clashes.

Focus restraining and repressive actions on misbehaving protesters to avoid creating motives in crowd to counterattack police.

As much as possible, avoid inflammatory signals.

Use secured vehicles or other means as physical barriers to stop crowds from entering areas with opposing groups present.

Separate peaceful demonstrators physically from violent crowd members, thereby facilitating detection and arrest.

Change focus to demonstrators’ messages by encouraging them to abstain from confrontations.

Focus restraining and repressive actions on misbehaving protesters, avoiding affecting whole crowd.

Scatter crowds, making coordinated actions against opposing groups difficult.

Detect and remove opposing groups via information from demonstrators.

Find alternatives to confrontational protests for delivering message.

Make arrests when unnoticed by crowd.

Get trains or subway trains not to stop at certain stations to keep opposing groups of demonstrators apart.

Intervene against misbehaving demonstrators on information from organizers.

Ban demeaning language toward demonstrators.

continues

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 211

DIALOGUE POLICING

Table 10.4 Increase Efforts

211

Continued Increase Risks

Decrease Rewards

Remove Excuses

Reduce Provocations

Detect and remove violent demonstrators via information from cooperating external actors.

Cut short police activities perceived as provocative.

Use a show of force.

Keep units perceived as provocative out of sight. Engage units perceived as nonprovocative to solve acute problems for demonstrators. Give multifaceted descriptions in media, avoiding negative stereotyping. Support selfpolicing by assisting organizers’ efforts.

specific techniques may be considered “soft.” It will often take courage and imagination to make use of them. As pointed out earlier, it is extremely difficult to ascertain a view based on hard facts about the effectiveness of the new special tactic. However, the fact that compared to the old, the new tactic includes a number of measures and techniques that have proved to be effective for other types of offenses and that they are employed to prevent confrontations in demonstrations is in our opinion an indication that the new tactic is superior to the old. It should also be emphasized that the measures intended to remove excuses and to reduce provocations are in accordance with a well-substantiated theory of crowd behavior, namely the Elaborated Social Identity Model (Reicher 1996; Stott and Reicher 1998a; Stott and Drury 2000). They are also in agreement with processes that are considered to be mitigating between groups (Granström 2002). In fact, these techniques could be conceived of as concrete applications of these theories. Application of these procedures are, not surprisingly, in agreement with recommendations by researchers arguing for knowledgebased public order policing (Reicher et al. 2007). It could also be argued that since many of the techniques are counter to conventional police wisdom, the increased utilization might in itself be perceived as

10_Madensen_Ch10.qxd:Madensen

212

1/3/11

10:25 AM

Page 212

PREVENTING CROWD VIOLENCE

a criterion of success. Their own personal experiences have made commanders and police officers, who were initially skeptical of the new tactic, change their opinion. And lastly, there are some indications that confrontations have become fewer and less intense as a result of the new tactic (Holgersson 2010).

Dilemmas for Dialogue Policing The principles of dialogue policing are not really new for the police. Many forces have more or less consistently acted in accordance with all, or at least some, of them. The innovation is that specially appointed police officers for an extended period of time have been assigned to develop and try out the philosophy, and furthermore that it has become integrated in the tactics for crowd management and control. The endeavor should actually be seen as an effort to define a new function within the police—a function that should be as natural as the police negotiation function. Since concepts like this challenge conventional police wisdom and many beliefs about what policing should be, experience indicates that there will initially be suspicion and skepticism in police management and among police officers. To implement dialogue policing is thus not an easy or straightforward process. That problems are encountered is not unique to dialogue policing, but seems to occur generally when new knowledge-intense methods and philosophies of how to conduct police business are introduced. A well-documented example is the introduction of problem-oriented policing (see, e.g., Scott 2000; Knutsson 2003, 2009; Bullock, Erol, and Tilley 2006; Knutsson and Clarke 2006). As for dialogue policing, we will describe some problems and dilemmas. For one thing, dialogue policing makes the image of reality more complex. Police officers often seem to prefer to see the world as black and white, where some groups are just bad and some good. To question that perception may create uncertainty and frustration (Wahlström 2007). The dialogue police concept also questions the definition of good police work. The focus is not to win battles—it is to avoid them. Traditionally, when police officers have to resort to force they get the feeling that something has been accomplished. When confrontations do not occur, there is a risk that some police officers conclude that there was no point in them being present. The ambition to avoid conflicts is against their conception of what constitutes good police work. Dialogue policing makes for more sensible decisionmaking. A main issue is the fact that some actions of the police cause demonstrators to act more violently. One very important task for the dialogue police officers is continuously to warn about such risks—an undertaking that is not always appreciated. It seems to be difficult for some police officers and commanders to understand that the dialogue police officers’ role in presenting the perspective of the

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 213

DIALOGUE POLICING

213

demonstrators does not mean that the dialogue police officers agree with the opinions they describe. There is a danger that arguments and points of view will be rejected because commanders believe that the dialogue police officers have taken sides against the police and for the protesters. The dialogue police must adopt a noncombative role. It is not possible to engage actively in confrontations and use force, and at the same time be trusted by protesters. A closely related issue is the requirement that police officers report crimes that come to their attention. To unconditionally adhere to this obligation would, for example, make it very hard to enter a dialogue with organizers of illegal demonstrations. If the result would be charges, demonstrators would, needless to say, be unwilling to allow attendance of dialogue police officers. A circumscribed presence would reduce the chances of establishing a meaningful dialogue. It must be recognized by the police organization that dialogue police officers do not have a repressive function. There is a risk that commanders will put pressure on dialogue police officers to obtain intelligence. In their course of service, dialogue police officers will receive important information. To achieve a true dialogue requires honest intentions. The protesters must have confidence in the dialogue police officers. It is the task of specially appointed intelligence officers to gather intelligence. In order to negotiate agreements, dialogue police officers need to be granted some discretionary power. It is not uncommon that commanders want to have total control. Ultimately, it is the commanders that are in charge, but to be able to reach a fruitful dialogue, the officers must be given authority to discuss alternative solutions and to make preliminary agreements. Otherwise, there are no reasons to talk to dialogue police officers. Wider discretionary power increases the chances of establishing a successful dialogue and having nonviolent responses. The function and role of dialogue police officers put them in a situation that highlights an ethical dilemma. At the end of the day, there is an issue of to whom they are accountable. For dialogue policing to be undertaken, it is essential to acquire confidence of the groups with whom the officers interact. If dialogue police officers only act as willing message boys of the commanders, it will be impossible to gain trust. Some demonstrators will instinctively have negative opinions of dialogue police officers simply because they are police officers. They hate all police, including dialogue police officers. An aim of the dialogue police officers is to obtain a role as some sort of objective and independent observer. This requires a strong sense of integrity and independence. Two important outcomes of such a role should be pointed out. First, it might help in making aggressive demonstrators accept the presence of dialogue police officers. Second, it might help in restraining unacceptable behavior among police officers. To begin with, when dialogue policing was something new, negative opinions of dialogue police officers were common both in aggressive groups of

10_Madensen_Ch10.qxd:Madensen

214

1/3/11

10:25 AM

Page 214

PREVENTING CROWD VIOLENCE

demonstrators and among police officers with a tendency to use too much force. Both groups disliked the attendance of dialogue police officers. The perspective of dialogue police officers, both in the sense that they are physically close to demonstrators during demonstrations, and they observe what goes on as bystanders, trying to see how police actions affect the demonstrators, will put them in a situation where they will be able to witness and be sensitive to instances of unacceptable behavior by police officers. Group thinking and the code of silence will normally mean that such instances remain unnoticed, unreported, and unaddressed. If dialogue police officers speak out, it is not graciously received. On the contrary, they will be regarded more or less as traitors, even if formal complaints are not made. However, there has recently been an interesting development. The focus of the new tactic to restrict use of force and to use other methods to make demonstrators comply has made uniformed police officers more sensitive regarding the behavior of their colleagues. There are some instances where uniformed police officers have expressed very firm opinions about decisions to use force—it was considered to be unnecessary. Officers have also confronted units that they deemed to have acted unacceptably. This was earlier the role of the dialogue police, who can now avoid these internal conflicts and instead concentrate on other tasks. The progress can be brought back to the new tactic with its ambition to restrict forceful actions and to de-escalate tense situations. It illustrates the importance of involving the whole organization in the enterprise. Instead of complaining, dialogue police officers can now often inform the rest of the organization of solid and professional police conduct. This is not to say that everything is as it should be. There are still some units that focus on winning battles and that act inconsiderately. If they are supported by police management who avoid or suppress internal conflicts, there is a risk that the dialogue police unit will gradually be transformed to complying marionettes.

Notes We are grateful to the dialogue police officers and others in the Special Police Tactic organization who willingly shared their experiences and to those at the Stavern meeting for their comments and suggestions on this chapter. We are also indebted to Nick Tilley for useful comments and suggestions. 1. Source for the description is the report by the committee that was appointed by the government to examine the event (SOU 2002) supplemented by Oskarsson (2005) and conversations and interviews with different key persons. 2. The National Police Board is in charge of a project—Development of Knowledge Within the Special Police Tactic—with the ultimate objective of supporting and developing the new tactics. Another aim is to stimulate the interaction between research, education, and practice.

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 215

DIALOGUE POLICING

215

3. There is in fact a dedicated series to promote situational crime prevention: Crime Prevention Studies, formerly published by Criminal Justice Press, now published by Lynne Rienner Publishers, available at www.rienner.com. 4. The order of the last two mechanisms are changed compared to the usual way of presenting them.

10_Madensen_Ch10.qxd:Madensen

1/3/11

10:25 AM

Page 216

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 217

Bibliography

Abrahamsson, H., and Johansson, G. (2002). Dialog om dialogen som blev konfrontation. In M. Löfgren and M. Vatankhah (Eds.), Vad hände med Sverige i Göteborg? Stockholm: Ordfront. ACPO (2003). Manual of guidance on keeping the police. Bramshill, UK: National Police Training. Adang, O. M. J. (1988). Voetbalvandalisme geobserveerd. Tijdschrift voor Criminologie, 4, 302–312. Adang, O. M. J. (1991). Gedrag van mensen tijdens rellen. In P. B. Defares and J. D. van der Ploeg (Eds.), Agressie. Assen, Netherlands: Van Gorcum. Adang, O. M. J. (1998). Hooligans, autonomen, agenten. Geweld en politie-optreden in relsituaties. Alphen aan den Rijn, Netherlands: Samsom. Adang, O. M. J. (2001). Friendly but firm: The maintenance of public order. Paper presented at the 2001 meeting of the American Political Science Association, San Francisco, August 30–September 2. Adang, O. M. J. (2006). Utilisation-focused evaluation of large-scale police operations: International evaluation teams. In W. Stol and R. van der Wal (Eds.), Politiekundige verkenningen 2005–2006. Apeldoorn, Netherlands: Politieacademie. Adang, O. M. J., and Brown, E. (2008). Policing football in Europe. Experiences from peer review evaluation teams. Apeldoorn, Netherlands: Politieacademie. Adang, O. M. J., and Cuvelier, C. (2001a). Policing Euro2000. Apeldoorn, Netherlands: Dutch Police Academy. Adang, O. M. J., and Cuvelier, C. (2001b). Policing Euro2000: International police cooperation, information management and police deployment. Ubbergen, Netherlands: Felix-Tandem. Adang, O. M. J., and Stott, C. (2004). Evaluation of public order policing during Euro 2004. Tijdschrift voor Veiligheid en Veiligheidszorg, 4, 46–47. Anderson, B. (1983). Imagined communities. London: Verso. Barrows, S. (1981). Visions of the crowd in late nineteenth century France. New Haven and London: Yale University Press. BBC News (2009, November 1). Police quash nude pumpkin run. Retrieved November 1, 2009, from http://news.bbc.co.uk. Berk, R. (1972). The controversy surrounding analyses of collective violence: Some methodological notes. In J. F. Short and M. E. Wolfgang (Eds.), Collective violence. Chicago: Aldine/Atherton.

217

11_Madensen_RM.qxd:Madensen

218

1/3/11

10:24 AM

Page 218

BIBLIOGRAPHY

Bessel, R., and Emsley, R. (Eds.) (2000). Patterns of provocation. Police and public disorder. Oxford, UK: Berghan Books. Björ, J., Knutsson, J., and Kühlhorn, E. (1992). The celebration of Midsummer Eve in Sweden—A study in the art of preventing collective disorder. Security Journal, 3 (3), 169–174. Brantingham, P. J., and Brantingham, P. L. (1981). Notes on the geometry of crime. In P. J. Brantingham and P. L. Brantingham (Eds.), Environmental criminology. Beverly Hills, CA: Sage Publications. Brantingham, P. J., and Brantingham, P. L. (2003). Anticipating the displacement of crime using the principles of environmental criminology. Crime prevention studies, vol. 16. Monsey, NY: Criminal Justice Press. Brown, R., and Scott, M. S. (2007). Implementing responses to problems. Washington, DC: US Department of Justice, Office of Community Oriented Policing Services. Bullock, K., Erol, R., and Tilley, N. (2006). Problem-oriented policing and partnerships. Implementing an evidence based approach to crime reduction. Cullompton, UK: Willan Publishing. Bunyan, T. (1981/1982). The police against the people. Race and Class, 23, 153–170. Center for Business and Economic Research (2009). Metropolitan Las Vegas Tourism Statistics. Las Vegas: University of Nevada, Las Vegas. Center for Problem-Oriented Policing (2009). The problem analysis triangle. Retrieved April 28, 2009, from http://www.popcenter.org/about/?p=triangle. Chamard, S. (2006). Partnering with businesses to address public safety problems. Washington, DC: US Department of Justice, Office of Community Oriented Policing Services. Clarke, R. V. (1980). Situational crime prevention: Theory and practice. British Journal of Criminology, 20, 136–147. Clarke, R. V. (1990). Deterring obscene phone callers: Preliminary results of the New Jersey experience. Security Journal, 1 (3), 143–148. Clarke, R. V. (1992). Introduction. In R. V. Clarke (Ed.), Situational crime prevention: Successful case studies. Albany, NY: Harrow and Heston. Clarke, R.V. (1995). Situational crime prevention. In M. Tonry and D. P. Farrington (Eds.), Building a safer society: Strategic approaches to crime prevention. Crime and Justice, vol. 19. Chicago: University of Chicago Press. Clarke, R. V. (1997a). Introduction. In R. V. Clarke (Ed.), Situational crime prevention: Successful case studies (2nd ed.). Albany, NY: Harrow and Heston. Clarke, R. V. (1997b). Situational crime prevention: Successful case studies. New York: Harrow and Heston. Clarke, R. V. (2004). Technology, criminology and crime science. European Journal on Criminal Policy and Research, 10 (1), 55–63. Clarke, R. V. (2008). Situational crime prevention. In R. Wortley and L. Mazerolle (Eds.), Environmental criminology and crime analysis. Portland, OR: Willan Publishing. Clarke, R. V., and Eck, J. (2005). Crime analysis for problem solvers in 60 small steps. Washington, DC: US Department of Justice, Office of Community Oriented Policing Services and Center for Problem-Oriented Policing. Clarke, R. V., and Newman, G. R. (2006). Outsmarting the terrorist. Westport, CT: Praeger Security International. Cohen, L., and Felson, M. (1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44, 588–608. Connors, E. (2007). Planning and managing security for major special events: Guidelines for law enforcement. Washington, DC: US Department of Justice.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 219

BIBLIOGRAPHY

219

Cornish, D. B. (1994). The procedural analysis of offending and its relevance for situational prevention. In R. V. Clarke (Ed.), Crime prevention studies, vol. 3. Monsey, NY: Criminal Justice Press. Cornish, D. B., and Clarke, R. V. (Eds.) (1986). The reasoning criminal: Rational choice perspectives on offending. New York: Springer-Verlag. Cornish, D. B., and Clarke, R. V. (2003). Opportunities, precipitators, and criminal decisions: A reply to Wortley’s critique of situational crime prevention. In M. J. Smith and D. B. Cornish (Eds.), Theory for practice in situational crime prevention. Crime prevention studies, vol. 16. Monsey, NY: Criminal Justice Press. Cornish, D. B., and Clarke, R. V. (2008). The rational choice perspective. In R. Wortley and L. Mazerolle (Eds.), Environmental criminology and crime analysis. Portland, OR: Willan Publishing. Courakis, N. (1999). Football violence: Not only a British problem. In R. Hood and N. Courakis (Eds.), The changing face of crime and criminal policy in Europe. Oxford: University of Oxford. Cronin, P., and Reicher, S. (2006). A study of the factors that influence how senior officers police crowd events: On SIDE outside the laboratory. British Journal of Social Psychology, 45, 175–196. Cronin, P., and Reicher, S. D. (2009). Accountability processes and group dynamics: A SIDE perspective on the policing of an anti-capitalist riot. European Journal of Social Psychology, 39, 237–254. della Porta, D. (1995). Social movements, political violence and the state: A comparative analysis of Italy and Germany. Cambridge: Cambridge University Press. della Porta, D., and Fillieulle, O. (2004). Policing social protest. In D. A. Snow, S. A. Soule, and H. Kriesi (Eds.), The Blackwell companion to social movements. Oxford, UK: Blackwell. della Porta, D., Peterson, A., and Reiter, R. (Eds.) (2006). The policing of transnational protest. Aldershot, UK: Avebury. della Porta, D., and Reiter, H. (1998a). Policing protest: The control of mass demonstrations in Western democracies. Minneapolis: University of Minnesota Press. della Porta, D., and Reiter, H. (1998b). The policing of protest in Western democracies. In D. della Porta and H. Reiter (Eds.), Policing protest: The control of mass demonstrations in Western democracies. Minneapolis: University of Minnesota Press. della Porta, D., and Reiter, H. (2006). The policing of global protest: The G8 at Genoa and its aftermath. In D. della Porta, A. Peterson, and H. Reiter (Eds.), The policing of transnational protest. Aldershot, UK: Avebury. Dewar, C. K. (1979). Spectator fights at professional baseball games. Review of Sport and Leisure, 4 (1), 12–25. Diener, E., Fraser, S. C., Beaman, A. L., and Kelem, R. T. (1976). Effects of deindividuation variables on stealing among Halloween trick-or-treaters. Journal of Personality and Social Psychology, 33 (2), 178–183. DiLonardo, R. L., and Clarke, R. V. (1996). Reducing the rewards of shoplifting: An evaluation of ink tags. Security Journal, 7 (1), 11–14. Drury, J., and Reicher, S. D. (1999). The intergroup dynamics of collective empowerment: Substantiating the social identity model of crowd behaviour. Group Processes and Intergroup Relations, 2, 381–402. Drury, J., and Reicher, S. D. (2000). Collective action and psychological change: The emergence of new social identities. The British Journal of Social Psychology, 39, 579–604.

11_Madensen_RM.qxd:Madensen

220

1/3/11

10:24 AM

Page 220

BIBLIOGRAPHY

Drury, J., and Reicher, S. D. (2005). Explaining enduring empowerment: A comparative study of collective action and psychological outcomes. European Journal of Social Psychology, 35, 35–58. Drury, J., Stott, C., and Farsides, T. (2003). The role of police perceptions and practices in the development of “public disorder.” Journal of Applied Social Psychology, 33, 1480–1500. Duerson, A. (2004). Road trip: University of Wisconsin. Retrieved on September 7, 2009, from http://sportsillustrated.cnn.com/2004/sioncampus/10/27/road_trip 1028/index.html. Eck, J. E. (1994). Drug markets and drug places: A case-control study of the spatial structure of illicit drug dealing. Baltimore: University of Maryland. European Court of Human Rights (2009). Case of Giuliani and Gaggio v. Italy (Application no. 23458/02) Judgment Strasbourg, August 25, 2009. Farrell, N. (2004). Mussolini: A new life. London: Phoenix. Federal Bureau of Investigation (FBI) (1995). A practical overview of crisis negotiation. Critical incident response group, Crisis management unit. Quantico, VA: FBI Academy. Felson, M. (1986). Linking criminal choices, routine activities, informal control, and criminal outcomes. In D. B. Cornish and R. V. Clarke (Eds.), The reasoning criminal: Rational choice perspectives on offending. New York: Springer-Verlag. Felson, M. (1996). Redesigning Hell: Preventing crime and disorder at the Port Authority Bus Terminal. In R. V. Clarke (Ed.), Preventing mass transit crime. Crime Prevention Studies, vol. 6. Monsey, NY: Criminal Justice Press. Felson, M., and Clarke, R. V. (1998). Opportunity makes the thief: Practical theory for crime prevention police research. Series Paper 98. London: Home Office. Fernandez, L. A. (2008). Policing dissent: Social control and the anti-globalization movement. New Brunswick, NJ: Rutgers University Press. Ferwerda, H., and Adang, O. M. J. (2005). Hooligans in beeld. Van veel blauw naar slim blauw. Tijdschrift voor de Politie, 12, 18–20. Festinger, L., Pepitone, A., and Newcomb, T. (1952). Some consequences of deindividuation in a group. Journal of Abnormal and Social Psychology, 47, 382–389. Frosdick, S., and Marsh, P. (2005). Football hooliganism. Cullompton, UK: Willan. Gigerenzer, G., and Todd, P. M. (1999). Fast and frugal heuristics: The adaptive tool box. In G. Girgerenzer, P. M. Todd, and ABC Research Group (Eds.), Simple heuristics that make us smart. New York: Oxford University Press. Goffman, E. (1959). The presentation of self in everyday life. London: Penguin. Goldstein, H. (1979). Improving policing: A problem-oriented approach. Crime and Delinquency, 25 (2), 236–258. Goldstein, H. (1990). Problem-oriented policing. New York: McGraw-Hill. Goldstein, H. (2003). On further developing problem-oriented policing: The most critical need, the major impediments, and a proposal. In J. Knutsson (Ed.), Problemoriented policing: From innovation to mainstream. Crime Prevention Studies, vol. 15. Monsey, NY: Criminal Justice Press. Goldstein, H., and Susmilch, C. (1982). Experimenting with the problem-oriented approach to improving police service: A report and some reflections on two case studies. Photocopy of Vol. 4 of the Project on Development of a Problem-Oriented Approach to Improving Police Service. Madison: University of Wisconsin Law School. Gorringe, H., and Rosie, M. (2006). “Pants to poverty”? Making poverty history, Edinburgh 2005. Sociological Research Online, 11 (1), http://www.socresonline.org .uk/11/1/gorringe.html.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 221

BIBLIOGRAPHY

221

Gorringe, H., and Rosie, M. (2008a). It’s a long way to Auchterarder! “Negotiated management” and mismanagement in the policing of G8 protests. British Journal of Sociology, 59 (2), 187–205. Gorringe, H., and Rosie, M. (2008b). The policing of “global” protest: Policing protest at the G8 in Scotland. Current Sociology, 56, 691–710. Gottfredson, M., and Hirschi, T. (1990). General theory of crime. Palo Alto, CA: Stanford University Press. Granström, K. (Ed.) (2002). Göteborgskravallerna—Identitetsskapande och attitydförändringar genom deltagande i fredliga demonstrationer och våldsamma upplopp. Rapport 187. Stockholm: Styrelsen för psykologiskt försvar. Granström, K. (2008). Soziale Identität und Massenereignisse. Freie Assoziation, 3, 7–14. Granström, K., Guvå, G., Hylander, I., and Rosander, M. (2005a). Reclaim the streets. En analys av strategiska händelser i samband med olovliga gatumanifestationer [Reclaim the streets. An analysis of strategic events in connection to street demonstrations with no authorization]. FOG-report no. 52. Linköping, Sweden: Linköping University, Department of Behavioural Sciences. Granström, K., Guvå, G., Hylander, I., and Rosander, M. (2005b). Salemmanifestationerna 2004—Ett samspel med geografiska och ideologiska fokus [Salem manifestations— interplay with geographical and ideological focus]. FOG-report no. 56. Linköping, Sweden: Linköping University, Department of Behavioural Sciences. Granström, K., Guvå, G., Hylander, I., and Rosander, M. (2009). Riots and disturbances. How riots start and how order is maintained. FOG-report. Linköping, Sweden: Linköping University, Department of Behavioural Sciences. Granström, K., Guvå, G., Hylander, I., Rosander, M., Emilsson, K., and Lind, M. (2002). The Gothenburg riots—identity formation and attitude changes through participation in peaceful demonstrations and violent tumult. In S. Hovmark, P. Bergman, B. Edlund, and U. Stoetzer (Eds.), The group as a paradox. Proceedings from a conference on group and social psychology, May 2002. Stockholm: Stockholm University. Granström, K., and Rosander, M. (2008). Friedensbewahrende Intergruppenprozesse zwischen Polizei und Demonstranten bei den “Reclaim the street”-Protesten. Freie Assoziation, 3, 51–64. Guerette, R. T. (2009). The pull, push, and expansion of situational crime prevention evaluation: An appraisal of thirty-seven years of research. In J. Knutsson and N. Tilley (Eds.), Evaluating crime prevention initiatives. Crime Prevention Studies, vol. 24. Monsey, NY: Criminal Justice Press. Guvå. G. (2008). Hilfe oder Selbstkontrolle—über Intragruppenprozesse bei der Polizei bei Massenereignissen. Freie Assoziation, 3, 65–83. Guvå, G., and Hylander, I. (2008). Aggravation und Friedensprozesse. Freie Assoziation, 3, 33–51. Hall, A., and de Lint, W. (2003). “Policing labour in Canada.” Policing and Society, 3 (13), 219–234. Hau, S. (2006). Communication as the most important police strategy at the football World Cup Final 2006. FOG-report no 61. Linköping, Sweden: Linköping University, Department of Behavioural Sciences and Learning. Her Majesty’s Inspectorate of Constabulary (2009). Adapting to protest. http:// inspectorates.justice.gov.uk/hmic. Hoggett, J., and Stott, C. (2010a). Crowd psychology, public order police training and the policing of football crowds. Policing: An International Journal of Police Strategies and Management, 33, 2.

11_Madensen_RM.qxd:Madensen

222

1/3/11

10:24 AM

Page 222

BIBLIOGRAPHY

Hoggett, J., and Stott, C. (2010b). The role of crowd theory in determining the use of force in public order policing. Policing and Society, 20 (2), 1–14. Holgersson, S. (2005). “Yrke: POLIS.” Dissertation, Institute for Computer Science, Linköping University. Holgersson, S. (2010). Dialogue police: Experiences, observations, and possibilities. Stockholm: The Swedish National Police Board. Homel, R., Hauritz, M., McIlwain, G., Wortley, R., and Carvolth, R. (1997). Preventing drunkenness and violence around nightclubs in a tourist resort. In R. V. Clarke (Ed.), Situational crime prevention: Successful case studies (2nd ed.). Guilderland, NY: Harrow and Heston. Hopkins, N., and Reicher, S. (1997). The construction of social categories and processes of social change. In G. Breakwell and E. Lyons (Eds.), Changing European identities. London: Butterworth. Hylander, I., and Granström, K. (2009). Organizing for peace at a mass event, an example from a large football competition. Submitted manuscript. Hylander, I., and Guvå, G. (2008). Aggravation and peace processes in crowd events. Submitted manuscript. IJzerman, P. D. (1982). UC-Almelo 1981, van relbestrijding naar relbeheersing. Algemeen Politieblad, 1, 3–7. Institute of Race Relations (1987). Policing against black people. London: Institute of Race Relations. Jefferson, T. (1987). Beyond paramilitarism. British Journal of Criminology, 2, 47–53. Jefferson, T. (1990). The case against paramilitary policing. Milton Keynes, UK: Open University Press. Jefferson, T., and Grimshaw, R. (1984). Controlling the constable. London: Frederick Muller. King, M., and Waddington, D. (2005). Flashpoints revisited: A critical application to the policing of anti-globalization protest. Policing and Society, 15 (3), 255–282. King, M., and Waddington, D. (2006). The policing of transnational protest in Canada. In D. della Porta, A. Peterson, and H. Reiter (Eds.), The policing of transnational protest. Aldershot, UK: Avebury. Knutsson, J. (Ed.) (2003). Problem-oriented policing: From innovation to mainstream. Crime Prevention Studies, vol. 15. Monsey, NY: Criminal Justice Press. Knutsson, J. (2009). Standard of evaluations in problem-oriented policing-projects. Good enough? In J. Knutsson and N. Tilley (Eds.), Evaluating crime prevention initiatives. Crime Prevention Studies, vol. 24. Monsey, NY: Criminal Justice Press. Knutsson, J., and Clarke, R. V. (Eds.) (2006). Putting theory to work. Implementing situational prevention and problem-oriented policing. Crime Prevention Studies, vol. 20. Monsey, NY: Criminal Justice Press. La Vigne, N. (1997). Safe transport: Security by design on the Washington Metro. In R. V. Clarke (Ed.), Preventing mass transit crime. Crime Prevention Studies, vol. 6. Monsey, NY: Criminal Justice Press. Le Bon, G. (1895). The crowd: A study of the popular mind. London: Ernest Benn. Trans. 1926. London: Unwin. Lewis, J. (2007). Sport fan violence in North America. Lanham, MD: Rowman and Littlefield. LVMPD (2008). Annual Report. Las Vegas, NV: Las Vegas Metropolitan Police Department.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 223

BIBLIOGRAPHY

223

Madensen, T. D., and Eck, J. E. (2006). Student party riots. Washington, DC: US Department of Justice, Office of Community Oriented Policing Services and Center for Problem-Oriented Policing. Madensen, T. D., and Eck, J. E. (2008). Spectator violence in stadiums. Washington, DC: US Department of Justice Office of Community Oriented Policing Services. Madensen, T. D., Fisher, B. S., and Eck, J. E. (2004). Cinco de Mayo student disturbance: University of Cincinnati follow-up student survey. Funded by the Ohio Office of Criminal Justice Services through the Ohio Service for Crime Opportunity Reduction Project. Submitted to the University of Cincinnati Student Affairs Administration. Martin, E. D. (1920). The behaviour of crowds. London: Ernest Benn. Marx, G. (1981). Ironies of social control: Authorities as contributors to deviance through escalation, nonenforcement and covert facilitation. Social Problems, 3, 221–246. Mawby, R. C. (2002). Policing images: Policing, communication, and legitimacy. Cullompton, UK: Willan Publishing. May, M. (2006). Opinion 06-001. Halloween party cost assessments and legal liability issues. January 5, 2006. Madison City Attorney’s Office. May, M., and Wray, N. (2004). Report to Madison mayor David Cieslewicz on options to eliminate the Halloween party or reduce Halloween problems. December 15, 2004. McCarthy, J. D., and McPhail, C. (1998). The institutionalization of protest in the United States. In D. S. Meyer and S. Tarrow (Eds.), The social movement society: Contentious politics for a new century. Lanham, MD: Rowman and Littlefield, 59–82. McClelland, J. S. (Ed.) (1970). The French right: From de Maistre to Maurras. London: Jonathan Cape. McClelland, J. S. (1989). The crowd and the mob: From Plato to Canetti. London: Unwin. McMains, M., and Mullins, W. (1996). Crisis negotiations managing. Critical incidents and hostage situations in law enforcement and corrections. Cincinnati, OH: Anderson Publishing. McPhail, C. (1991). The myth of the madding crowd. New York: Aldine de Gruyter. McPhail, C. (1997). Stereotypes of the crowds and collective behavior: Looking backward, looking forward. In D. E. Miller, M. A. Katovich, and S. L. Saxton (Eds.), Constructing complexity: Symbolic interaction and social forms. Greenwich, CT: JAI Press. McPhail, C., Schweingruber, D., and McCarthy, J. (1998). Policing protest in the United States: 1960–1995. In D. della Porta and H. Reiter (Eds.), Policing protest: The control of mass demonstrations in Western democracies. Minneapolis: University of Minnesota Press, 49–69. Milgram, S., and Toch, H. (1969). Crowds and social movements. In G. Lindzey and E. Aronson (Eds.), The handbook of social psychology, vol. 4. (3rd ed.). New York: Addison-Wesley. Mitchell, D., and Staeheli, M. A. (2005). Permitting protest: The fine geography of dissent in America. International Journal of Urban and Regional Research, 29 (4), 796–813. Möller, R., Sander, U., Schäfer, A., Villányi, D., and Witte, M. (2009). Motive structures and violence among young globalization critics: A statistical typlogy of the motives for protest at the 2007 G8 summit. International Journal of Conflict and Violence, 3, 124–142.

11_Madensen_RM.qxd:Madensen

224

1/3/11

10:24 AM

Page 224

BIBLIOGRAPHY

Newman, G. R., and Clarke, R. V. (2003). Superhighway robbery: Preventing e-commerce crime. Portland, OR: Willan Publishing. Noakes, J. A., and Gillham, P. F. (2004). Policing protest in the US and Europe: The failure of negotiated management and the emergence of selective incapacitation post-Seattle. Paper presented at the “Policing Protest After Seattle” Workshop, Fiskebackskil, Sweden, May 2–5, 2004. Noakes, J. A., and Gillham, P. F. (2006). Aspects of the “New Penology” in the police response to major political protests in the United States, 1999–2000. In D. della Porta, A. Peterson, and H. Reiter (Eds.), The policing of transnational protest. Aldershot, UK: Avebury. Noakes, J. A., Klocke, B., and Gillham, P. F. (2005). Whose streets? Police and protester struggles over space in Washington DC, September 2001. Policing and Society, 15 (3), 235–254. Novelli, D., Drury, J., and Reicher, S. D. (2010). Come together: Two studies concerning the impact of group relations on “personal space.” British Journal of Social Psychology, 49 (2), 223–236. Nye, R. (1975). The origins of crowd psychology. Beverly Hills, CA: Sage. Oskarsson, M. (2002). Dialog med förbehåll. Polisens kontaktgrupp, demonstrationsnätverken och det tragiska dilemmat i Göteborg. In M. Björk and A. Peterson (Eds.), Vid politikens yttersta gräns. Perspektiv på EUtoppmötet i Göteborg 2001. Stockholm: Brutus Östlings Bokförlag Symposion. Oskarsson, M. (2005). Lag eller ordning. Polisens hantering av EU-toppmötet i Göteborg 2001. Doktorsavhandling, Kriminologiska institutionen, Stockholms Universitet. Stockholm: Jure Förlag AB. Otten, M. H. P., Boin, R. A., and van der Torre, E. J. (2001). Dynamics of disorder: Lessons from two Dutch riots. The Hague: Crisis Research Center, Leiden University. Peterson, A., and Oskarsson, M. (2006). The police riots in Gothenburg: The European Union summit in Gothenburg, June 2001. In A. Peterson and M. Björk (Eds.), Policing contentious politics in Sweden and Denmark. Maastricht: Shaker Publishing. Police Cooperation Working Party of the Council of the European Union (2006). European Union handbook on international police cooperation and measures to prevent and control violence and disturbances in connection with football matches with an international dimension. Official Journal of the European Union (English edition), 49, C32. Postmes, T., and Spears, R. (1998). Deindividuation and anti-normative behaviour: A meta-analysis. Psychological Bulletin, 123, 238–259. Poyner, B. (1986). A model for action. In K. Heal and G. Laycock (Eds.), Situational crime prevention. London: Her Majesty’s Stationery Office. Poyner, B. (1991). Situational crime prevention in two parking facilities. Security Journal, 2 (1), 96–101. Poyner, B., and Webb, B. (1997). Reducing theft from shopping bags in city centre markets. In R. V. Clarke (Ed.), Situational crime prevention: Successful case studies, 2nd ed. Guilderland, NY: Harrow and Heston. Prayag Magh Mela Research Group (2007). Living the Magh Mela at Prayag: Collective identity, collective experience and the impact of participation in a mass event. Psychological Studies (Special Issue on the Social Psychology of Collectivity), 52, 293–301. Reddy, W. M. (1977). The textile trade and the language of the crowd at Rouen, 1752–1871. Past and Present, 74, 62–89.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 225

BIBLIOGRAPHY

225

Reicher, S. D. (1982). The determination of collective behaviour. In H. Tajfel (Ed.), Social identity and intergroup relations. Cambridge: Cambridge University Press, and Paris: Maison des Sciences de l’Homme. Reicher, S. D. (1984). The St. Paul’s riot: An explanation of the limits of crowd action in terms of a social identity model. European Journal of Social Psychology, 14, 1–21. Reicher, S. D. (1987). Crowd behaviour as social action. In J. C. Turner, M. A. Hogg, P. J. Oakes, S. D. Reicher, and M. S. Wetherell (Eds.), Rediscovering the social group: A self-categorization theory. Oxford: Blackwell. Reicher, S. D. (1996). “The Battle of Westminster”: Developing the social identity model of crowd behaviour in order to explain the initiation and development of collective conflict. European Journal of Social Psychology, 26, 115–134. Reicher, S. D. (1997). Collective psychology and the psychology of the self. BPS Social Section Newsletter, 36, 3–15. Reicher, S. D. (2001a). Crowds and social movements. In M. A. Hogg and R. S. Tindale (Eds.), Blackwell handbook of social psychology: Group processes. Oxford: Blackwell. Reicher, S. D. (2001b). The psychology of crowd dynamics. In M. A. Hogg and R. S. Tindale (Eds.), Blackwell handbook of social psychology: Group processes. Oxford: Blackwell. Reicher, S. D., Haslam, A., and Hopkins, N. (2005). Social identity and the dynamics of leadership: Leaders and followers as collaborative agents in the transformation of social reality. Leadership Quarterly, 16, 547–568. Reicher, S. D., and Haslam, S. A. (2010). Beyond help: A social psychology of collective solidarity and social cohesion. In M. Snyder and S. Sturmer (Eds.), The psychology of prosocial behavior. Chichester: Wiley-Blackwell. Reicher, S. D., Stott, C., Cronin, P., and Adang, O. M. J. (2004). An integrated approach to crowd psychology and public order policing. Policing: An International Journal of Police Strategies and Management, 27, 558–572. Reicher, S. D., Stott, C., Drury, J., Adang, O. M. J., Cronin, P., and Livingstone, A. (2007). Knowledge-based public order policing: Principles and practice. Policing, 1, 403–415. Rosie, M., and Gorringe, H. (2009). “The anarchists’ world cup”: Respectable protest and media panics. Social Movement Studies, 8 (1), 35–53. Rudé, G. (1964). The crowd in history: A study of popular disturbances in France and England, 1730–1848. London: Wiley. Russell, G. (1983). Crowd size and density in relation to athletic aggression and performance. Social Behavior and Personality, 11 (1), 9–15. Russell, G. (2004). Sport riots: A social-psychological review. Aggression and Violent Behavior, 9, 353–378. Russell, G., and Arms, R. (1998). Towards a social psychological profile of would-be rioters. Aggressive Behavior, 24, 219–226. Russell, G., Arms, R., and Mustonen, A. (1999). When cooler heads prevail: Peacemakers in a sport riot. Scandinavian Journal of Psychology, 40, 153–155. Sampson, R., Eck., J. E., and Dunham, J. (2010). Super controllers and crime prevention: A routine activity explanation of crime prevention success and failure. Security Journal, 23 (1), 37–51. Sampson, R., and Scott, M. S. (2000). Tackling crime and other public-safety problems: Case studies in problem-solving. Washington, DC: US Department of Justice Office of Community Oriented Policing Services.

11_Madensen_RM.qxd:Madensen

226

1/3/11

10:24 AM

Page 226

BIBLIOGRAPHY

Schelling, T. C. (1978). Micromotives and macrobehavior. New York: W. W. Norton. Schnell, D. (2005). Madison’s Halloween history. Retrieved on April 28, 2009, from www.badgerherald.com/news/2005/10/26/madisons_halloween_h.php. Schofield, G. (2000). Controlling boisterous celebrations. The Police Chief, 67 (6), 51–62. Schreiber, M., and Adang, O. M. J. (2010). Fiction, facts, and a summer’s fairy tale–Mixed messages at the World Cup 2006. Policing and Society, 20 (2), 237–255. Schweingruber, D. (2000). Mob sociology and escalated force: Sociology’s contribution to repressive police tactics. The Sociological Quarterly, 41 (39), 371–389. Scott, M. S. (2000). Problem-oriented policing. Reflections on the first 20 years. Washington, DC: US Department of Justice, Office of Community Oriented Policing Services. Scott, M. S., Eck, J. E., Knutsson, J., and Goldstein, H. (2008). Problem-oriented policing and environmental criminology. In R. Wortley and L. Mazerolle (Eds.), Environmental criminology and crime analysis. Collumpton, UK: Willan Publishing. Scott, M. S., and Goldstein, H. (2005). Shifting and sharing responsibility for public safety problems. Washington, DC: US Department of Justice. Sharp, A. (1992). Rock concerts. Law and Order, 40, 149–154. Shearing, C. D., and Stenning, P. C. (1985). From the Panopticon to Disney World: The development of discipline. In A. N. Doob and E. L. Greenspan (Eds.), Perspectives in criminal law: Essays in honour of John Ll. J. Edwards. Toronto: Canada Law Book. Sheptycki, J. (Ed.) (2002). Issues in transnational policing, London: Taylor and Francis. Sighele, S. (1892). La foule criminelle, essai de psychologie collective. Paris: Alcan. Silverwood, G., and Housley, W. (2002). Halloween—After-action report. Compiled for Madison Police Department Chief Richard Williams. December 5, 2002. Sivanandan, A. (1981/1982). From resistance to rebellion: Asian and Afro-Caribbean struggles in Britain. Race and Class, 23, 111–152. Sivanandan, A. (1983). Challenging racism: Strategies for the 1980’s. Race and Class, 25, 1–11. SOU (2002). Göteborg 2001. Betänkande från Göteborgskommittén. Statens offentliga utredningar report 2002:122. Stockholm: Fritzes förlag. Stensaasen, S., and Sletta, O. (2000). Grupprocesser. Om inlärning och samarbete i grupper. Stockholm: Natur och kultur. Stott, C. J., and Adang, O. M. J. (2004). “Disorderly” conduct: Social psychology and the control of football “hooliganism” at “Euro2004.” The Psychologist, 17 (6), 318–319. Stott, C. J., and Adang, O. M. J. (2005). Crowd dynamics, policing, and “hooliganism” at “Euro2004.” UK Economic and Social Research Council research report. Grant reference RES-000-23-0617. Stott, C. J., and Adang, O. M. J. (2009). Understanding and managing risk: Policing football matches with an international dimension in the European Union. Copenhagen: Bavnebanke Press. Stott, C. J., Adang, O. M. J., Livingstone, A., and Schreiber, M. (2007). Variability in the collective behaviour of England fans at Euro2004: “Hooliganism,” public order policing and social change. European Journal of Social Psychology, 37, 75–100. Stott, C. J., Adang, O. M. J., Livingstone, A., and Schreiber, M. (2008). Tackling football hooliganism: A quantitative study of public order, policing, and crowd psychology. Psychology, Public Policy, and Law, 14 (2), 115–141.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 227

BIBLIOGRAPHY

227

Stott, C. J., and Drury, J. (1999). The intergroup dynamics of empowerment: A social identity model. In P. Bagguley and J. Hearns (Eds.), Transforming politics: Power and resistance. Basingstoke, NY: Macmillan. Stott, C. J., and Drury, J. (2000). Crowds, context, and identity: Dynamic categorization processes in the “poll tax riot.” Human Relations, 53, 247–273. Stott, C. J., Hutchison, P., and Drury, J. (2001). “Hooligans” abroad? Inter-group dynamics, social identity, and participation in collective “disorder” at the 1998 World Cup finals. British Journal of Social Psychology, 40, 359–384. Stott, C. J., Livingstone, A., and Hoggett, J. (2007). Policing football crowds in England and Wales: A model of “good practice”? Policing and Society, 18, 258–281. Stott, C. J., and Pearson, G. (2007). “Football hooliganism,” policing and the war on the “English disease.” London: Pennant Books. Stott, C. J., and Reicher, S. (1998a). Crowd action as intergroup process: Introducing the police perspective. European Journal of Social Psychology, 28, 509–529. Stott, C. J., and Reicher, S. D. (1998b). How conflict escalates: The inter-group dynamics of collective football crowd “violence.” Sociology, 32, 353–377. Taine, H. (1876). L’Ancien régime “les origines de la France contemporaine.” Trans. John Durand. London: Daldy, Ibister and Co. Tajfel, H., and Turner, J. (1979). An integrative theory of intergroup conflict. In W. Austin and S. Worchel (Eds.), The social psychology of intergroup relations. Monterey, CA: Brooks/Cole, 33–47. Thompson, E. P. (1971). The moral economy of the English crowd in the eighteenth century. Past and Present, 50, 76–136. Tilley, N. (2006). Knowing and doing: Guidance and good practice. In J. Knutsson and R. V. Clarke (Eds.), Putting theory to work. Implementing situational prevention and problem-oriented policing. Crime Prevention Studies, vol. 20. Monsey, NY: Criminal Justice Press. Tilley, N. (2009). What’s the “what” in “what works”? Health, policing, and crime prevention. In J. Knutsson and N. Tilley (Eds.), Evaluating crime prevention initiatives. Crime Prevention Studies, vol. 24. Monsey, NY: Criminal Justice Press. Tilly, C., Tilly, L., and Tilly, R. (1975). The rebellious century 1830–1930. Cambridge: Harvard University Press. Turner, J. C. (2005). Explaining the nature of power: A three-process theory. European Journal of Social Psychology, 35, 1–22. Turner, J. C., Oakes, P., Hogg, M. A, Reicher, S., and Wetherell, M. (1987). Rediscovering the social group: A self-categorisation theory. Oxford: Basil Blackwell. Turner, R. H., and Killian, L. M. (1987). Collective behavior (3rd ed.). Englewood Cliffs, NJ: Prentice Hall. van den Brug, H. (1986). Voetbalvandalisme: Een speurtocht naar verklarende factoren. Haarlem: De Vrieseborch. van de Valk, P., and Linckens, P. J. (1988). Voetbalvandalen: Hun contacten met politie en justitie. Tijdschrift voor Criminologie, 4, 313–320. Veno, A., and Veno, E. (1993). Situational prevention of public disorder at the Australian Motorcycle Grand Prix. In R. V. Clarke (Ed.), Crime prevention studies, vol. 1. Monsey, NY: Criminal Justice Press. Verma, A. (2007). Anatomy of riots: A situational crime prevention approach. Crime Prevention and Community Safety, 9 (3), 201–221. Vevea, B. (2008). Halloween history: Student government once sponsored big, nonviolent festivities. Wis. Stat. sec. 66.0627(2). Retrieved on April 28, 2009, from www.badgerherald.com/news/2008/10/31/halloween_history_st.php.

11_Madensen_RM.qxd:Madensen

228

1/3/11

10:24 AM

Page 228

BIBLIOGRAPHY

Vitale, A. S. (2005). From negotiated management to command and control: How the New York Police Department polices protests. Policing and Society, 15 (3), 283–304. Waddington, D. (1992). Contemporary issues in public disorder: A comparative and historical approach. London: Routledge. Waddington, D. (2007a). Policing public disorder: Theory and practice. Cullompten, UK: Willan Publishing. Waddington, D. (2007b). Seattle and its aftershock: Some implications for theory and practice. Policing, 1 (4), 380–389. Waddington, D., and Critcher, C. (2000). Policing pit closures, 1984–1992. In R. Bessel and C. Elmsley (Eds.), Patterns of provocation: Police and public disorder. Oxford: Berghahn Books. Waddington, D., Jones, K., and Critcher, C. (1987). Flashpoints of public disorder. In G. Gaskell and R. Benewick (Eds.), The crowd in contemporary Britain. London: Sage. Waddington, D., Jones, K., and Critcher, C. (1989). Flash points: Studies in public order. London: Routledge. Waddington, D., and King, M. (2007). The impact of the local: Police public order tactics and strategy during the G8 Justice and Home Affairs Ministerial Meeting in Sheffield, England, June 2005. Mobilization: An International Quarterly, 12 (4), 417–430. Waddington, P. A. J. (1987). Towards paramilitarism? Dilemmas in policing civil disorder. British Journal of Criminology, 27, 37–46. Waddington, P. A. J. (1991). The strong arm of the law: Armed and public order policing. Oxford, UK: Clarendon. Waddington, P. A. J. (1993a). Dying in a ditch: The use of police powers in public order. International Journal of the Sociology of Law, 21, 335–353. Waddington, P. A. J. (1993b). The case against paramilitary policing considered. British Journal of Criminology, 33, 353–373. Waddington, P. A. J. (1994). Liberty and order, public order policing in a capital city. London: UCL Press. Waddington, P. A. J. (1998). Controlling protest in contemporary historical and comparative perspective. In D. della Porta and H. Reiter (Eds.), Policing protest. The control of demonstrations in Western democracies. Minneapolis: University of Minnesota Press, 117–140. Waddington, P. A. J. (2003). Policing public order and political contention. In T. Newburn (Ed.), Handbook of policing. Cullompten, UK: Willan Publishing. Wahlström, M. (2007). Forestalling violence: Police knowledge of interaction with political activists. Mobililization: The International Quarterly, 4, 309–402. Ward, R. E. (2002). Fan violence: Social violence of moral panic. Aggression and Violent Behavior, 7, 453–475. Webb, B. (1994). Steering column locks and motor vehicle theft: Evaluations from three countries. In R. V. Clarke (Ed.), Crime prevention studies, vol. 2. Monsey, NY: Criminal Justice Press. Weiss, J., and Davis, M. (2005). Policing the Oktoberfest. Law and Order, 53, 77–80. Wilson, M., and Daly, M. (1985). Competitiveness, risk-taking, and violence: The young male syndrome. Ethology and Sociobiology, 6, 59–73. Wortley, R. (2001). A classification of techniques for controlling situational precipitators of crime. Security Journal, 14 (4), 63–82. Wortley, R., and Mazerolle, L. (Eds.) (2008). Environmental criminology and crime analysis. Collumpton, UK: Willan Publishing.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 229

BIBLIOGRAPHY

229

Wortley, R., and Smallbone, S. (2006). Situational prevention of child sexual assault. Crime Prevention Studies, vol. 19. Monsey, NY: Criminal Justice Press. Young, K. (2002). Standard deviations: An update on North American sports crowd disorder. Sociology of Sport Journal, 19, 237–275. Zimbardo, P. G. (1969). The human choice: Individuation, reason, and order versus deindividuation, impulse, and chaos. In W. J. Arnold and D. Levine (Eds.), Nebraska symposium on motivation, vol. 1. Lincoln: University of Nebraska Press.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 230

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 231

The Contributors

Otto M. J. Adang is a behavioral scientist and chair of public order management at the Police Academy of the Netherlands where he heads the international research program “Managing Dangerous Situations.” He has published extensively on violence and use of force, public order management, and football hooliganism and consults with police forces across Europe. John E. Eck is a professor at the University of Cincinnati’s School of Criminal Justice. He specializes in police effectiveness and the prevention of crime at places. With Tamara D. Madensen, he has written problem-specific guides for police on preventing stadium violence and preventing college drinking riots, and he has been instrumental in the development, testing, and advancement of problem-oriented policing. Kjell Granström, professor emeritus in educational psychology, is a reader in communication and licensed psychologist. His main research interests concern communication and dynamics within and between groups, such as school classes, staff teams, football supporters, and groups in demonstrations. Stefan Holgersson works as a police officer and has a Ph.D. in information system development. He is an assistant professor at Linköping University. One of his research interests has been to study police officers’ varying work performance and the causes of these variations. Ingrid Hylander is associate professor in educational psychology at Linköping University, senior researcher at Karolinska Institutet, and licensed psychologist. Her research interest is in group psychology, teamwork, and professional consultation and collaboration.

231

11_Madensen_RM.qxd:Madensen

232

1/3/11

10:24 AM

Page 232

THE CONTRIBUTORS

Johannes Knutsson is a professor of police research at the Norwegian Police University College. He also has a part-time position as professor at the Swedish National Police Board and is a visiting professor at University College London. His research interest has primarily been crime prevention and evaluation of crime preventive measures. Tamara D. Madensen is an assistant professor in the Department of Criminal Justice at the University of Nevada, Las Vegas. Her research interest is in crime prevention, especially as it relates to reducing opportunities and risks of criminal victimization. She has coauthored crowd management policing guides with John E. Eck and conducted numerous evaluations of police crime prevention strategies. Joel Plant is the chief of staff to the Milwaukee Police Department. Previously, he was a Madison, Wisconsin, mayoral aide responsible for public safety and neighborhood issues. He was Madison’s first alcohol policy coordinator, a research assistant for the Center for Problem-Oriented Policing, and a crime analyst in the Tempe (Arizona) Police Department. Stephen Reicher is professor of psychology at the University of St. Andrews. His work is concerned with the relationship between social identities and collective action, and he has some 150 publications covering such phenomena as crowd behavior, nationalism, intergroup conflict, leadership, and the psychology of tyranny. Michael S. Scott is the director of the Center for Problem-Oriented Policing and clinical associate professor at the University of Wisconsin Law School. He served in various police administrative positions in Florida, St. Louis, and New York City and as a police officer in Madison, Wisconsin; was a senior researcher at the Police Executive Research Forum in Washington, DC; and was the 1996 recipient of the Gary P. Hayes Award for innovation and leadership in policing from the Police Executive Research Forum. William H. Sousa is an associate professor in the Department of Criminal Justice at the University of Nevada, Las Vegas. He is currently involved in research on police order-maintenance practices and community crime prevention in Las Vegas neighborhoods. His most recent publications appear in the Journal of Experimental Criminology and Police Practice and Research. Clifford Stott is a senior lecturer at the University of Liverpool. His research focuses on the dynamics of intergroup interaction, social identity processes, and the development of violent norms during crowd events. His work has

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 233

THE CONTRIBUTORS

233

played a major role in both the development of crowd theory and government and police policy across Europe. David Waddington is professor of communications and head of the Communication and Computing Research Centre, Sheffield Hallam University, UK. His primary research interests are the sociology of industrial relations, the decline and regeneration of former industrial areas, theories of crowd violence, and the policing of public disorder.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 234

12_Madensen_index.qxd:Layout 1

1/3/11

10:23 AM

Page 235

Index

Action space, 133 Adang, Otto M. J., 2–3, 17 After-action report, 146, 157, 162, 165 Aggravation and Mitigation (AM) Model: Gothenburg riots, 81–83; Mitigation strategies, 88–92; Peaceful assembly, 84–86; Processes, 73–74 AM Model. See Aggravation and Mitigation (AM) Model Awareness space, 127 Barricade tactic, 147–149 Berk, Richard, 47 Bounded rationality, 62, 64 Branding, 16, 176–178, 187–188 Bubbling, 7. See also Corralling; Kettling Cascade briefings, 105 Categorizing, 73–74; Aggravating relational conditions, 83; Mitigating relational conditions, 85–86; Using active and passive mitigation strategies, 88–89, 92–93 Center for Problem-Oriented Policing, 172 Chaos: Avoiding, 88, 91, 93; Creating, 74–75, 80, 83, 94, 195 Charge Plan, 175–177, 184

CHEERS, 162 Clarke, Ronald V., 120, 127, 130, 158 Coercive police tactics, 29, 38, 43 Collective mind, 9, 31. See also Group mind; Racial unconscious Command vocabulary, 194 Community policing, 131–132; in special police units, 40, 42 Contagion, 13, 26 Contextual: Influences, 46, 103–104, 108–111; Level of analysis, 98, 100, 112, 133 Controllers, 125–126, 133, 172. See also Guardians; Handlers; Place managers; Routine activity theory Cornish, Derek, 130 Corralling, 7, 96–97, 99, 109–110. See also Bubbling; Kettling Counterguardianship, 126 Crime pattern theory, 116, 123, 127 Crowd-related crime, 116, 122; Elements, 125 Cultural: Influences, 103–104, 108–109, 111; Intelligence, 18; Level of analysis, 98–100, 113, 133; Norms/values, 34, 99; Understandings, 13 De-escalation of crowd violence, 17, 41, 73–74, 91, 196, 202, 204–

235

12_Madensen_index.qxd:Layout 1

236

1/3/11

10:23 AM

Page 236

INDEX

205, 214; Policy, 37; Tactics 4, 194–195 Deindividuation: Theory, 27, 48, 68; Effect, 64 della Porta, Donatella, 31, 100 Dialogue Police: Dynamic risk assessment, 5; Goals and tactics, 45, 70, 201, 207; Improving crowd relations, 3–4; Obstacles, 199–200, 212–214; Officer functions, 41–42, 90, 192, 197, 203–206; Officer perceptions of, 42, 199–200; Origins, 198, 200–201 Preventing violence, 41; Principles, 202 Differentiation: Between crimes 122, 124; Between crowd types and settings 122, 124; Between groups 19, 39, 74–75, 79, 81, 85–90, 92–93, 122, 124, 136, 195; Failure to use, 22, 83; Police use of force, 34, 37, 41; Tactical approaches, 108, 197 Disarming actions, 74–75, 78, 88–90, 94, 195. See also Mitigation Displacement, 179, 182, 184, 188 Dynamic risk: Assessment, 5, 35, 39; Theoretical model, 35 Eck, John E., 3, 5, 156 Elaborated Social Identity Model (ESIM), 3–4, 28–29, 33, 43, 46, 65, 71, 112, 136, 195, 207; Crowd formation and categorization, 73, 92; Integration with situational crime prevention, 4–5; Policing football crowds, 2, 34, 37–38; Policing protest crowds, 40; Promoting positive police interaction, 151; Public order policing, 30–34; Swedish National Police Tactic, 41, 201, 211. See also Social psychology Environmental backcloth, 127 Environmental criminology, 2, 115, 123, 127–128, 135; Application to crowds, 3, 4, 5, 116, 131, 136; Contrasted with standard criminology 118-124; Extending the ESIM perspective, 4, 5, 133 Escalated force, 30

ESIM. See Elaborated Social Identity Model Facilitation of legitimate crowd goals, 1, 4–5, 19–20, 34–38, 40–41, 43, 66, 74, 81, 88–89, 92–93, 104–105, 113, 132, 146, 165, 184, 196, 201, 210; Communicating intentions, 22; Importance of crowd members perceptions, 21, 132 Facilitation of misbehavior and crime, 117, 209 Felson, Marcus, 120 Festinger, Leon, 27 Flashpoints, 101, 109, 112, 178 Flashpoints Model of Public Disorder: Conditions of disorder, 98, 102, 107; Contextualized theory, 2, 46, 95; Theoretical value, 112, 133 Force multiplier, 149 Forward Intelligence Teams, 45 Frosdick, Steven, 47 Gating Plan, 175–179, 184 Gillham, Patrick, 99 Goal-governed protests, 78, 81, 93 Goffman, Erving, 63 Goldstein, Herman, 130–131 Graded tactical: Approach, 38, 43; Profile, 35–38 Granström, Kjell, 2, 4, 135–136 Group mind, 27. See also Collective mind; Racial unconscious Guardians, 125–126, 133, 172, 185. See also Controllers Handlers, 125, 133, 172. See also Controllers High-profile approach, 67, 114 Holgersson, Stefan, 3–5, 41, 42, 70, 199 Hylander, Ingrid, 2, 4, 135–136 Idea-governed protests, 79–81, 87–89, 91–93; Protesters, 90 Imagined communities, 11 Institutional/Organizational: Influences, 103, 108; Level of analysis, 98, 100, 113 Interactional: Influences, 105, 108–109; Level of analysis, 98, 101, 112, 114, 133

12_Madensen_index.qxd:Layout 1

1/3/11

10:23 AM

Page 237

INDEX

Intergroup dynamics, 33, 47 Intergroup perspective, 70–71, 93 Iterative problem-solving, 157, 163–165, 172. See also Problem-Oriented Policing Jefferson, Tony, 44, 98 Kettling, 7, 10, 21–22. See also Bubbling; Corralling Knutsson, Johannes, 3–5, 41, 42, 70, 115, 117 Leader-governed protests, 77 Le Bon, Gustave, 13–14, 26–28, 31–32, 145 Level of analysis. See Contextual; Cultural; Institutional/Organizational; Interactional; Political/Ideological; Situational; Structural Life cycle of gatherings, 129. See also McPhail-Madensen Model Low-profile approach, 45, 66–67, 70, 101, 111 Madensen, Tamara D., 3–5, 129, 156 Marsh, Peter, 47 Martin, E. D., 9, 13 McPhail, Clark, 47, 62, 128–129, 142 McPhail-Madensen Model, 129. See also Life cycle of gatherings Milgram, Stanley, 13 Missile throwing, 122, 129, 134, 147, 155, 210 Mitigation: Active, 73, 75, 88–89; Categorization, 92; Interaction, 73, 81; Organizing, 91; Passive, 73, 75, 88–89; Peaceful public assembly, 86; Perspective, 71, 73; Processes, 70, 73–75, 81, 211; Strategies, 71, 73, 75–76, 78, 81, 83–85, 88–89, 93; Treatment, 74, 89, 93. See also Disarming actions Mob(s), 26, 28, 31–32, 48, 135 Mobile units, 41, 50, 70, 196. See also Riot police Möller, Renate, 72 Mutual categorization, 72, 76 Mutual treatment, 72–74, 83, 86, 88

237

National Basic Tactic. See Swedish Special Police Tactic (SPT) Negative stereotyping: Aggravating conflict, 74–75, 195; Associated problems, 110, 205; Avoiding 85, 92–93, 202, 211; Chaotic organization, 92; Idea-governed protest categorizing, 89; Outcome of group-thinking, 203 Negotiated management, 30 Noakes, John, 99 Nonconfrontational: Control, 70; Tactical options, 37–38, 40–41, 45, 194 Nonorganized protests, 77; Protesters, 76, 80 Operational: Experience, 33; Philosophy, 100; Policing, 32; Practice, 33, 37, 45; Protocols, 40 Opportunity structures: Creating, 135; Decision outcomes, 128; for crime, 122, 133–134; for crowd violence, 143–144; Manipulation, 157; within crowds, 120–122, 134 Opportunity theories, 120 Organization perspective, 71 Organizing, 74; Aggravating relational conditions, 83; Mitigating relational conditions, 86; Peaceful 74, 78, 89; Police categorization of protest groups, 76; Processes of crowds 2, 72–73; Types, 77, 79, 80; Using active and passive mitigation strategies, 88–89, 91, 93 Paramilitary police, 44–45 Peace unit, 66 Peer review evaluation team, 67–68 Physical crowd, 14, 16 Place managers, 125–126, 128, 133, 149, 172. See also Controllers Plant, Joel, 3–5, Police knowledge, 31–33 Police perspective, 70, 72–73, 93 Police Support Unit (PSU), 41, 45–46 Political/ideological: Level of analysis, 98; Influences, 102, 107–108, 111; Police policy, 113 Portuguese Public Security Police (PSP), 37–39, 41, 43, 45 Positive stereotyping: Effect, 136;

12_Madensen_index.qxd:Layout 1

238

1/3/11

10:23 AM

Page 238

INDEX

Mitigating conflict, 74, 88, 92, 195; Peaceful public assemblies, 86 Post-event assessment, 166–168, 171, 174, 180–181, 183 Poyner, Barry, 117 Pre-event planning, 165, 167, 169, 171, 175, 180, 182 Problem analysis triangle, 172. See also Routine activity theory Problem-Oriented Policing, 130–132, 175, 212; within environmental criminology, 4, 123; Police planning processes, 5, 165, 183 Protests. See Goal-governed protests; Idea-governed protests; Leadergoverned protests; Nonorganized protests; Rule-governed protests Provocateurs, 80 Provocative: Crowd behavior, 51, 77, 80, 89, 107; Equipment, 90; Interaction, 74, 83, 101; Language, 90; Police behavior, 75–76, 78, 90–91, 94, 204, 221; Police deployment, 114 PSP. See Portuguese Public Security Police PSU. See Police Support Unit Psychological crowd, 14, 16, 34 Public Order Act 1986, 96, 104 Public Order Act 1994, 97, 110 Public Order Policing: as problemsolving, 11–12; Challenges, 16, 26, 31; Crowd conflict, 30–31, 108; Effectiveness, 34; Evaluations, 34–35, 39; London G20 demonstrations, 7–8; New principles, 2, 17–21; Reform, 41, 44, 113, 211; Zero tolerance, 100 Racial unconscious, 27. See also Collective mind; Group mind Rational choice perspective, 116, 123, 127–128, 130; Application, 143 Reicher, Stephen, 2–3, 13, 15, 26–28, 33, 43, 45–47, 62, 64–66, 71, 133, 135, 151 Riot police: Effect, 61; Special units, 41, 44, 49–50, 67. See also Mobile units Risk objects, 59–60 Routine activity theory, 116, 122–123, 125, 127; in crowds, 125. See also

Controllers; Guardians; Handlers; Place managers Rule-governed protests, 78, 80–81, 89 SARA problem-solving framework, 183; Helix, 164–165 Schofield, Gary, 145 Scott, Michael S., 3–5, 156 Script, 128, 134 Self-policing: by fans, 52, 62; Police tactics to encourage, 37, 39, 41, 44, 133, 192, 196–197, 205, 210. See also Self-stewarding Self-stewarding, 101, 114. See also Selfpolicing SET. See Special Events Team Sighele, Scipio, 48 Signal value, 196–198, 210 Situational: Influences, 104, 108–109; Level of analysis, 98–99, 101, 113, 133 Situational Crime Prevention, 64, 116, 123, 128, 130, 134, 141, 143; Approach to crowd related crime, 131, 156; Extending the ESIM perspective, 5; Police tactics, 147–153, 176; Strategy effectiveness, 3, 142–144, 157; Techniques, 4, 133–134, 136, 141, 143–144 Social categories, 11, 14, 28–29. See also Imagined communities; Social identification; Social identity Social identification, 39. See also Imagined communities; Social categories; Social identity Social identity: Change processes, 3, 29, 32, 71, 92, 142; Effect on collective disorder, 33, 65; Evidence, 29; Groups within crowds, 18, 72; Individual, 28; Salient, 29, 65. See also Imagined communities; Social categories; Social identification Social psychology, 29, 31, 37; Early research and theory, 26, 142; Empirical support, 27–28; Integrated with environmental criminology, 132, 143. See also Elaborated Social Identity Model (ESIM) Social solidarity, 14–15

12_Madensen_index.qxd:Layout 1

1/3/11

10:23 AM

Page 239

INDEX

Sousa, William H., 3–5 Special Events Team (SET), 166, 168, 170, 174, 180, 181 SPT. See Swedish Special Police Tactic Stakeholders: Accountability, 186; Building consensus, 176; Planning and recommendations, 169–171, 173; Reactions, 177, 184; Recruitment, 185 Standard criminology, 116, 118, 120–121, 123–124 State: Green, 196–197, 198; Red, 196, 198; Yellow 196, 198 Stott, Clifford, 2–3, 5, 13, 17, 19, 34, 43, 47, 64–65, 151 Strategic incapacitation, 99, 109, 112 Structural: Level of analysis, 98, 113; Influences, 102, 107–108 Subgroup perspective, 70, 72, 93 Submergence, 26 Suggestibility, 26 Super controllers, 133. See also Routine activity theory Swarming, 155 Swedish National Police Board, 41, 193, 195, 214

239

Swedish Special Police Tactic (SPT), 41, 70; Dialogue Policing, 198–199; Effectiveness, 207; History, 195; Mitigation strategies, 93; Police officer functions, 206; Principles, 196; Situational techniques, 210 Taine, Hippolyte, 25–27 Threat assessments, 40 Thrill seekers, 81 Toch, Hans, 13 Triggers of violence: Competition, 58; Events, 55–57; Identifiable, 63; Mechanisms, 2, 157; Mitigation, 142; Police measures, 58, 65, 101; Preceding aggression, 63; Unidentifiable, 62 Vocalizations, 59 Waddington, David, 2, 4, 46, 70 Waddington, P.A.J., 44–45 Young male syndrome, 63–64, 67 Zimbardo, Philip G., 48

12_Madensen_index.qxd:Layout 1

1/3/11

10:23 AM

Page 240

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 235

CRIME PREVENTION STUDIES Ronald V. Clarke, Series Editor Crime Prevention Studies is an international book series dedicated to research on situational crime prevention and other initiatives to reduce opportunities for crime. Most volumes center on particular topics chosen by expert guest editors. The editors of each volume, in consultation with the series editor, commission the chapters to be published and select peer reviewers.

Volume 1, edited by Ronald V. Clarke, 1993. Volume 2, edited by Ronald V. Clarke, 1994. Volume 3, edited by Ronald V. Clarke, 1994. Volume 4, Crime and Place, edited by John E. Eck and David Weisburd, 1995. Volume 5, The Politics and Practice of Situational Crime Prevention, edited by Ross Homel, 1996. Volume 6, Preventing Mass Transit Crime, edited by Ronald V. Clarke, 1996. Volume 7, Policing for Prevention: Reducing Crime, Public Intoxication, and Injury, edited by Ross Homel, 1997. Volume 8, Crime Mapping and Crime Prevention, edited by David Weisburd and J. Thomas McEwen, 1997. Volume 9, Civil Remedies and Crime Prevention, edited by Lorraine Green Mazerolle and Jan Roehl, 1998. Volume 10, Surveillance of Public Space: CCTV, Street Lighting, and Crime Prevention, edited by Kate Painter and Nick Tilley, 1999. Volume 11, Illegal Drug Markets: From Research to Prevention Policy, edited by Mangai Natarajan and Mike Hough, 2000. Volume 12, Repeat Victimization, edited by Graham Farrell and Ken Pease, 2001. Volume 13, Analysis for Crime Prevention, edited by Nick Tilley, 2002. Volume 14, Evaluation for Crime Prevention, edited by Nick Tilley, 2002. Volume 15, Problem-Oriented Policing: From Innovation to Mainstream, edited by Johannes Knutsson, 2003. Volume 16, Theory for Practice in Situational Crime Prevention, edited by Martha J. Smith and Derek B. Cornish, 2003. 241

11_Madensen_RM.qxd:Madensen

242

1/3/11

10:24 AM

Page 236

BOOKS IN THIS SERIES

Volume 17, Understanding and Preventing Car Theft, edited by Michael G. Maxfield and Ronald V. Clarke, 2004. Volume 18, Designing Out Crime from Products and Systems, edited by Ronald V. Clarke and Graeme R. Newman, 2005. Volume 19, Situational Prevention of Child Sexual Abuse, edited by Richard Wortley and Richard Smallbone, 2006. Volume 20, Putting Theory to Work: Implementing Situational Prevention and Problem-Oriented Policing, edited by Johannes Knutsson and Ronald V. Clarke, 2006. Volume 21, Imagination for Crime Prevention: Essays in Honour of Ken Pease, edited by Graham Farrell, Kate J. Bowers, Shane D. Johnson, and Michael Townsley, 2007. Volume 22, Surveying Crime in the 21st Century, edited by Mike Hough and Mike Maxfield, 2007. Volume 23, Perspectives on Identity Theft, edited by Megan M. McNally and Graeme R. Newman, 2008. Volume 24, Evaluating Crime Reduction Initiatives, edited by Johannes Knutsson and Nick J. Tilley, 2009. Volume 25, Reducing Terrorism Through Situational Crime Prevention, edited by Joshua D. Freilich and Graeme R. Newman, 2009. Volume 26, Preventing Crowd Violence, edited by Tamara D. Madensen and Johannes Knutsson, 2011.

11_Madensen_RM.qxd:Madensen

1/3/11

10:24 AM

Page 237

About the Book

FROM JUBILANT SPORTS FANS CELEBRATING A VICTORY TO ANGRY political protesters, crowds create volatile situations that can all too often result in violence or property destruction. Preventing Crowd Violence offers a lucid examination of crowd behavior and of law enforcement tactics designed to de-escalate tensions and promote cooperative interactions. Tamara D. Madensen is assistant professor of criminal justice at the University of Nevada, Las Vegas. Johannes Knutsson is professor of police research at the Norwegian Police University College.

243