cronin&reicher 2006

23
Copyright © The British Psychological Society Reproduction in any form (including the internet) is prohibited without prior permission from the Society A study of the factors that inuence how senior ofcers police crowd events: On SIDE outside the laboratory Patrick Cronin 1 and Stephen Reicher 2 * 1 School of Social and Health Sciences, University of Abertay, UK 2 School of Psychology, University of St Andrews, UK This paper ts into the SIDE perspective (Reicher, Spears, & Postmes, 1995; Postmes, Spears, Lea, & Reicher, 2000), which emphasises the importance of integrating the cogn itive and strat egic dimensions of gro up proc esses. Our stud y examines the decisions made by senior police ofcers during a simulation exercise of a crowd event. The analysis shows, rstly, that ofcers are deeply concerned about their accountability to a variety of audiences, both internal and external to the police force. Second, these different audiences pressure them to act in different, and sometimes contradictory, ways. What counts, then, is the overall balance between accountability concerns. Third, this balance – and, with it, police perceptions and decisions – alters in the course of an event. More specically, with escalating conict, the balance of accountability concerns move s increasin gly in the direction of undi ffere ntiated inte rvention agai nst crowd members. In discussion, we consider both the theoretical implications of this analysis for research on group processes (in particular the importance of accountability issues once one moves beyond the laboratory and deals with groups that have a past and future and in which membership is more than simply an act of choice) and the practical implications in terms of crowd policing. In the mi dst of London’ s Po ll T ax Ri ot of March 1990, the phone ra ng in the Metr opoli ta n Police public order control room at New Scotland Yard. The ofcer who replied heard a familiar voice say ‘this is Margaret’. After a brief moment, he realized he was talking to the Prime Minister . This ofcer , a participant in our present progra mme of research, told us this anecdote in order to stress tha t cro wd pol ici ng tak es pla ce und er int ens e scru tin y from multiple audiences at multiple levels. These audiences have the power to exact con sequen ces according to the ir jud gme nt of pol ice per formance. Hen ce, however the y themselves would wish to proceed, the senior ofcers who control the police operation must con stantl y man oeu vre around the expec tations and demands of those who scrutinise them. * Corresponden ce should be addressed to Stephen Reicher, School of Psychology, University of St. Andrews, St. Andrews, Fife K16 9JU, Scotland, UK (e-mail: [email protected]). The British Psychological Society 175 British Journal of Social Psychology (2006), 45, 175–196 q 2006 The British Psychological Society www.bpsjournals.co.uk DOI:10.1348/014466605X41364

Upload: gijoaf1980

Post on 02-Jun-2018

226 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 1/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

A study of the factors that influence how senior officers police crowd events: On SIDE outsidethe laboratory

Patrick Cronin1 and Stephen Reicher2*1School of Social and Health Sciences, University of Abertay, UK2School of Psychology, University of St Andrews, UK

This paper fits into the SIDE perspective (Reicher, Spears, & Postmes, 1995; Postmes,

Spears, Lea, & Reicher, 2000), which emphasises the importance of integrating the

cognitive and strategic dimensions of group processes. Our study examines the

decisions made by senior police officers during a simulation exercise of a crowd event.

The analysis shows, firstly, that officers are deeply concerned about their accountability

to a variety of audiences, both internal and external to the police force. Second, these

different audiences pressure them to act in different, and sometimes contradictory,

ways. What counts, then, is the overall balance between accountability concerns. Third,

this balance – and, with it, police perceptions and decisions – alters in the course of an

event. More specifically, with escalating conflict, the balance of accountability concerns

moves increasingly in the direction of undifferentiated intervention against crowd

members. In discussion, we consider both the theoretical implications of this analysis

for research on group processes (in particular the importance of accountability issues

once one moves beyond the laboratory and deals with groups that have a past and

future and in which membership is more than simply an act of choice) and the practical

implications in terms of crowd policing.

In the midst of London’s Poll Tax Riot of March 1990, the phone rang in the MetropolitanPolice public order control room at New Scotland Yard. The officer who replied heard afamiliar voice say ‘this is Margaret’. After a brief moment, he realized he was talking tothe Prime Minister. This officer, a participant in our present programme of research, toldus this anecdote in order to stress that crowd policing takes place under intense scrutiny from multiple audiences at multiple levels. These audiences have the power to exactconsequences according to their judgment of police performance. Hence, however they 

themselves would wish to proceed, the senior officers who control the police operationmust constantly manoeuvre around the expectations and demands of those whoscrutinise them.

* Correspondence should be addressed to Stephen Reicher, School of Psychology, University of St. Andrews, St. Andrews, FifeK16 9JU, Scotland, UK (e-mail: [email protected]).

The

British

Psychological

Society

175

British Journal of Social Psychology (2006), 45, 175–196

q 2006 The British Psychological Society 

www.bpsjournals.co.uk 

Page 2: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 2/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

The primary aim of this paper is to examine more systematically and

comprehensively the various ways in which police officers can be held to account for 

their decisions during public order events, the ways in which that affects their decisions,

and the ways in which changes in accountability leads to changes in police action.

In doing so, we wish to contribute to a growing literature on the way in which strategic

as well as cognitive factors impact on group behaviour (Reicher  et al., 1995; Postmes

et al., 2000). In addition, by understanding the factors that govern policing decisions,

 we also hope to gain a better understanding of the overall dynamics of crowd events

and, more specifically, the conditions under which crowd conflict is likely to escalate.Both of these concerns – the latter self-evidently, but the former less obviously – can

be traced back to crowd research. According to classic models of crowd behaviour 

(Le Bon, 1895) and their development into de-individuation theory (for reviews see

Postmes & Spears, 1998; Reicher  et al., 1995), the question of conflict escalation barely 

arises. It is assumed that, as people become anonymous in the mass, they lose their 

individual identity and, hence, they lose control of their behaviour. As a result, crowd

members are incapable of resisting antisocial impulses and the potential for destructive

behaviour becomes ever present. By contrast, recent investigations of crowd behaviour 

(Drury & Reicher, 1999, 2000; Reicher, 1984, 1987, 1996a, 2001; Stott & Drury, 2000;

Stott, Huchison, & Drury, 2001; Stott & Reicher, 1998a) have stressed that crowd

members shift from personal to social identity in the crowd, and that control of 

behaviour passes from personal concerns to the norms, values, and beliefs associated with the relevant social category. As a result, the behaviour of crowd members is not

unconstrained. Rather, they act within clear social limits and their collective acts display 

clear social patterns that reveal their collective understandings.However, it is not simply that there is a cognitive transformation in the crowd.

Crowd members also gain the power to enact their social identities. This is partly due

to their sense of support from fellow group members who are all acting together as

members of a common category, and partly due to their sense of being anonymous

to out-group members who thereby lose their ability to single them out for retribution.

 As Reicher (1987) concludes, perhaps it is only in the crowd that people are able to act

fully as social subjects and to realize their collective perspective. According to this argument, the cognitive salience of social identity is not sufficient

for people to behave in terms of social identity. One also needs to consider the practical

ability to express social identity, especially where in-group norms prescribe acts that would attract sanctions from out-groups. What is more, the same factors, such as the

classic de-individuation manipulation of lowered visibility in a group, may operate in

complex ways at both levels. It may both increase the salience of social identity by 

reducing cues to interpersonal difference and lower the power of out-groups to punish 

expressions of social identity they deem illegitimate, but also reduce cues to social

support from fellow in-group members, which would be necessary for challenges

against a powerful out-group.

These claims have subsequently received experimental support (Reicher & Levine,

1994a, 1994b; Reicher, Levine, & Gordijn, 1998; Spears & Lea, 1992; Spears, Lea,

Corneliussen, Postmes, & Ter Haar, 2002) and formed the basis for the so-called social

identity model of de-individuation effects (SIDE; Postmes  et al., 2000; Reicher  et al.,

1995). SIDE emerged out of specific concerns related to crowd behaviour and, over time, it has developed into a more general programme to integrate what have been

termed the ‘cognitive’ and ‘strategic’ dimensions of group behaviour.

Patrick Cronin and Stephen Reicher 176

Page 3: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 3/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

On the one hand, research has sought to identify the different types of strategicconsideration that moderate one’s ability to act in terms of social identity. These can be

divided into intragroup concerns, such as gaining acceptance as a group member andevaluating support from fellow group members for particular actions (e.g. Barreto,Spears, Ellemers, & Shahinper, 2003; Ellemers, Spears, & Doojse, 2002; Ellemers, van Dyck, Hinkle, & Jacobs, 2000; Spears, Postmes, Lea, & Wolbert, 2002) and intergroupconcerns, such as improving the out-group’s general perception of the in-group or avoiding punishment from the out-group for specific acts (Barreto et al., 2003; Klein & 

 Azzi, 2001; Klein & Licata, 2003; Scheepers, Spears, Doojse, & Manstead, 2003).On the other hand, it has been shown that these various strategic concerns may play 

a part in explaining an increasingly diverse range of group phenomena, includingexpressed level of in-group identification, in-group bias, judgments of prototypicality,

normative behaviour, loyalty, and commitment (e.g. Barreto & Ellemers, 2000; Barretoet al., 2003; Douglas & McGarty, 2001; Klein, Licata, Durala, & Azzi, 2003; Reicher & Hopkins, 2001; Reicher, Hopkins, & Condor, 1997; Spears, Postmes  et al., 2002).

Nonetheless, despite the accumulating body of evidence, SIDE research is at arelatively early stage and, in many ways, it is still more a programme for research than aprogramme of research. There remain a number of issues to be resolved and a number of gaps that need to be filled, some of the most obvious include the following. First, mostresearch on SIDE uses experimental paradigms that generally seek to manipulate factors(such as audience or visibility) which impact on strategic considerations, and then

examine the consequences for behaviour. While such studies are undoubtedly useful for investigating the impact of these factors, they cannot tell us whether they actually operate or are important when it comes to the groups that interest us outside thelaboratory.

Second, like most group research, SIDE research tends to use informal socialcategories such as nationality, ethnicity, gender, and so on. Membership of thesecategories is generally a matter of choice and there are no formal procedures for admission and no formal procedures for expulsion. In many ways, this limits thestrategic importance of others since they are more limited in their ability to impose

consequences upon the actor. This is very different in institutional groups. I cannotsimply decide to identify myself as a doctor, a lawyer, or a police officer. I depend uponothers to judge my performance and to give me entry. I also depend upon the judgmentof others since, if I displease them, they may block my progress, discipline me, and evenexpel me. Hence, in looking at the actual operation of strategic factors, it may bepropitious to start with an examination of an institutional group.

Third, in common with experimental research in general, SIDE research aims toisolate factors in order to analyse their impact. The tendency is to look at one strategicfactor at a time, perhaps two in interaction, or three at most. However, outside the

laboratory, multiple strategic factors will be operating at the same time and, as discussedabove, the same factor may be operating at many different levels. What is necessary,then, is not only to identify the range and impact of individual factors that operate at any one time, but also to examine how group members balance these various factors andhow this balance might change over time. This is clearly of importance since, if differentstrategic concerns affect behaviour in different ways, then changes in the overallbalance of concerns are likely to alter the probability of different behaviours being

expressed.In this paper, we address these issues through a study of senior police officersinvolved in controlling a crowd event. Not surprisingly, public order policing has

Policing crowd events   177

Page 4: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 4/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

generated considerable interest from a range of social sciences. However, such work has

tended to concentrate on policy and practical dimensions and has largely ignored a

psychological perspective (e.g. Fielding, 1991; Northam, 1988; Townshend, 1993;

 Waddington, 1992, 1987, 1994). More recently, social psychologists interested in crowd

dynamics have begun to take an interest in the police (Drury & Reicher, 1999; Drury,

Stott, & Farsides, 2003; Reicher, 1996a; Reicher, Stott, Cronin, & Adang, 2004;

Stott & Drury, 2000; Stott  et al., 2001; Stott & Reicher, 1998b).They point out that crowd events are intergroup encounters, typically between

crowds and the police, and that police action can have a critical effect on the outcome of events. In particular, where the police act in ways deemed illegitimate in terms of crowd

perceptions, or else stop crowd members from acting in ways they perceive to be

legitimate, then conflict is liable to ensue. To complicate matters, many physical masses

of people contain multiple psychological crowds in the sense of people who identify 

themselves as members of a category with distinct norms, values, and understandings.

Some of these crowds may sanction illegal or confrontational actions while others may 

not. How the police respond when certain groups enter into confrontation with them is

of critical importance. If they treat all crowd members as a homogenous danger, then

they create the conditions for conflict to escalate from some to all of the groups who are

present.This research has played an important part in showing how police actions impact on

crowd dynamics and also in elucidating the perceptions and attitudes which contributetowards these actions. It has also revealed some of the practical constraints that affect

 what the police do. For instance, Stott and Reicher (1998b) show that, when employing

tactics such as crowd dispersal, the police treat all crowd members similarly not only 

because they perceive them to be homogenous, but also due to the difficulty of 

distinguishing between people in the rush of events. Nonetheless, despite these

contributions, research has not sought to make a systematic investigation of the

strategic constraints operating upon police decisions, nor has it generally been as

interested in the antecedents of police judgments as in their consequences.In sum, research into the strategic determinants of group behaviour has tended to

ignore the nature and interplay between strategic factors that impact simultaneously 

upon institutional groups as they act in the world, such as the police during crowd

events. Conversely, psychological research into public order policing has tended to

ignore strategic determinants. In this study, then, we investigate the perspective of command level officers – those who set the overall strategy and tactics of police

operations – as they are in the process of dealing with a crowd event. It might seem most

obvious to do this in the context of a real event. However, our previous observations

(Cronin, 2001) indicate a serious drawback to this approach.Such are the time pressures during real events that command level officers

frequently make decisions without discussion and without articulating their reasons.

Moreover, to the extent that they do communicate, it is usually to fellow officers who

share a common set of assumptions, a common stock of knowledge, and a common set

of prior references. Hence, much remains implicit and is intelligible only to those who

already have access to these shared understandings. Consequently, such events are

problematic as a means of revealing those understandings and the way they affect

decisions. This is particularly relevant in a study like ours that is concerned as much  with hypothesis generation as it is with hypothesis testing. Whereas it may be possible

to devise criteria for identifying instances of strategic factors to which one is already 

Patrick Cronin and Stephen Reicher 178

Page 5: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 5/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

attuned, it is far harder to identify the relevant factors in the first place if they are not

made explicit.For these reasons, our study is based on a ‘table top’ exercise in which officers are

taken through a developing crowd scenario and have to decide on strategy, tactics, and

deployment of resources. The obvious disadvantage of such an exercise is its artificiality 

and the fact that it may differ from the ‘real thing’, both by omitting factors that would

be relevant during actual events and, perhaps more seriously, in that the way the

scenario is set up and events are characterized may invoke perceptions and reactions

that might otherwise be absent. It is hard to dismiss such criticisms entirely, and this

undoubtedly has implications for the weight that can be placed on the analysis of the

present study and also for the relationship between this and other studies of crowd

policing, both of which we will address in the discussion. At the same time, it is worth 

stressing that the exercise was designed and run by the police as the culmination of one

of their training courses for public order commanders and was made to be as realistic as

possible.The scenario was an amalgam of elements mostly taken from actual events. It was

run in real time, putting pressure on participants to commit themselves to decisions.

The way in which information was provided – in the form of verbal ‘injects’ about the

developing situation – was made to be in the form and language that commanders would

receive if, for instance, they were controlling an event in a police control room. While in

such cases they may have some live video of the event, much of the key information

 would be radioed in by intelligence officers on the ground, filtered through a specialist

officer in the control room, and then handed as an ‘inject’ to the commander. Finally,

given that the exercise was part of the course that officers needed to pass in order to

qualify as a public order commander, their performance had real consequences for their 

position as a police officer. Although important, the artificiality of the exercise should not be overstated.

By contrast, the great advantage of using such an exercise was that the officers had

to discuss their perceptions of the event, to indicate how they thought it would unfold,

to talk about the effects of different interventions (or non-interventions), and thus

provide a rationale for how they would police the event. This provided us with explicit

material covering the concerns of officers, their judgments, their treatment of the

crowd, and the relationship between these various elements. To summarize, our aims in

analysing this material were:

(1) to undertake a systematic analysis of the strategic factors which influence publicorder policing and to examine whether and how each of these factors affects the

 judgment and decisions of officers.(2) to investigate whether the nature and balance of strategic factors changes during

an event and, if so, whether and how this affects the way in which the policeperceive and treat crowds(a) as a special example of the above, given the evidence suggesting the impact

of differentiated versus undifferentiated treatment of crowds on crowdconflict, to examine whether and how the changing nature of strategic

factors affects the likelihood that the police will treat the crowd as ahomogenous and dangerous entity.

Policing crowd events   179

Page 6: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 6/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Analytic framework 

Participants

The participants in this study were eight Metropolitan Police officers at the rank of chief inspector, or the more senior rank of commander, along with two senior members of theScottish Prison Service being trained by the Metropolitan Police. All the officers weremale (which reflects the general preponderance of male officers in public order 

command positions). They were aged between 37 and 54 years (mean: 44 years) andhad between 16 and 29 years of service (mean: 22 years). All had previously completed

advanced public order training.

The exercise

 All the participants were attending a week-long bronze commander training course at

the Hounslow public order-training centre (colloquially known as ‘riot city’) situated in west London (Townshend, 1993). In Britain, the command structure for public order isdivided into ‘gold’, ‘silver’ and ‘bronze’ officers. Gold commanders set the overallstrategy and are often removed from the actual policing of an event. Silver commanderstranslate this strategy into tactics and, while being intimately involved with an event, arefrequently situated in a command centre that is removed from the ground. Bronzecommanders translate tactics into action. They are responsible for a specified sector of an event, which may be either geographical or other (e.g. in charge of the reserve or else

in charge of horse mounted officers). Therefore, bronze commanders are those whogive the orders that determine what front-line officers actually do. On successfully completing the course, participants would be eligible to act as bronze commanders for public order events throughout London.

The exercise was part of the course. It was designed and run by police who were

public order specialists andwas conducted on thelast daylasting 3.5 hours with no break.

Theparticipants, along with theresearcher and twoinstructors, were seated together in a

classroom. Using verbal, written, and visual inputs, the instructors presented a crowd

scenario and asked the participants to discuss how they would respond as a group. Over 

time, the instructors provided further injects to describe the development of the event

and, at each point, invited further discussion from the participants. However, there was

time pressure since the eventswere described as if they were in real time. Additionally, all

the inputs were presented in a form and in language that mirrored the information that

 would be the basis for police decisions in a real event.Thus, prior to anyevent, commanders would be given a briefing from ‘special branch’

(the intelligence wing of the police), which would detail the historical context and past

history of conflict; the expected numbers of participants divided into three categories –

leading troublemakers (Category A), people who will readily participate in conflict

(Category B), and others (Category C) – and the overall likelihood of trouble. During the

event, they would get constant information filtered to them from officers on the ground –

so-called forward intelligence teams (FITs). Once again, these would describe salient

events, note sightings of known Category A and B individuals in particular and provide

both numbers and evidence of what they are doing. The initial briefing mirrored a special

branch report along with descriptions of the situation from officers on the ground. The

later injects took the form of reports from front line officers and FITs.

The scenario concerned a march and rally involving 5,000 anti-Fascists who wereexpressing their opposition to a British National Party (BNP) candidate for local

government elections in London. They were also protesting at a police decision to merely 

Patrick Cronin and Stephen Reicher 180

Page 7: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 7/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

caution this candidate for an offence of threatening behaviour. The BNP was at the time

(January 1997) the largest far-right group in Britain. While the event was fictional

(although it bore some resemblance to a number of different anti-fascist and anti-police

demonstrations), it was situated in a real part of the city (the boroughs of Stoke

Newington and Hackney in north London) and the scenario was presented using maps of 

this area.

The event unfolded as follows:

(1) Following an overnight downpour, the agreed assembly point for the march,Clissold Park, (Point A on the map), had to be changed. The police had todetermine how to inform the 5,000 demonstrators and get them to agree to this,knowing that no organizer had been identified.

(2) After delays caused by the change of assembly, the police decided that themarch would start from a public road bordering the park, Stoke NewingtonChurch St (Point B). It was planned to end with a rally on arrival at Clapton

Common (Point C).(3) As the march passed by the library (Point D) on Stoke Newington Church Street

(which had previously hosted a controversial BNP meeting and was thereforeidentified in the pre-event briefing as a potential ‘trouble spot’), placards, bricks,and bottles began to be thrown at police serials (constables, sergeants, andinspectors) deployed outside the library. These serials were not public order trained, they did not have riot equipment to protect themselves and their 

colleagues, and they were wearing ordinary uniforms rather than specializedprotective clothing. Minor casualties were beginning to be sustained amongofficers and, in addition, some damage was being caused to the fabric of the library.

Policing crowd events   181

Page 8: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 8/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

However, the conflict was of a fairly low level and was contained to a limitedsection of the crowd comprising some 40 people.

(4) Eventually, the march moved past the library. However, when marchers came tothe junction with Stoke Newington High Street (Point E), they stopped and splitinto two sections of some 2,000 each.

(5) At this point, a sustained period of rioting broke out involving a number of incidents. This included the use of petrol bombs and an attack on the home of theBNP candidate in Smalley Close (Point F). There was also a report that crowd

members had been seen in possession of firearms.(6) The exercise ended with considerations concerning the preservation of the scene

of the disorder for criminal inquiries.

Data sources

The analysis presented here is part of a larger study of the training week, which involveda combination of methods including participant observation, interviews both beforeand after the training exercise, and textual analysis of the exercise itself. In this paper, we concentrate on this last data source and we only use data from other sources to theextent that it helps clarify the text.

The training exercise was recorded using a wide angled video-camera placedunobtrusively at the back of the classroom (although both instructors and participantshad given consent to being filmed and were therefore aware that the camera was there).

The soundtrack was then transcribed into a word-processing package using the videoimages to help identify who said what. This yielded a transcript approximately 70,000 words in length. This transcript included only what was said, not such features as stressor pauses. These were felt to be superfluous given the nature of the analysis that weundertook.

Data analysis

Given the nature of our research questions, we employed a hybrid analytic strategy.On the one hand, we sought to explore the strategic factors that emerged from the data

as affecting the deliberations and decisions made by our participants. Accordingly, weused inductive procedures based on grounded theory (Glaser & Strauss, 1967;Miles & Huberman, 1994; Strauss & Corbin, 1998). On the other hand, we wereinterested in the incidence within the data of an analytically predefined category;namely, the perception and treatment of the crowd as homogenously dangerous. Thisinvolved procedures based upon thematic analysis (Kellehear, 1993).

The analysis involved three stages. The first stage, akin to the open coding of grounded theory, involved an iterative process of line-by-line readings of the text, andthe gathering together of extracts that address a similar concept category (where the

concepts are related to our general topic; namely, the factors that affect perception andtreatment of the crowd). The second stage, akin to axial and selective coding ingrounded theory, involved comparison between categories in order to examine how they cluster together and the relationships between them. The third and final stage, which was more akin to thematic analysis, involved a specific analysis of whether ‘perception and treatment of the crowd as homogenously dangerous’ had emerged as aconcept category and, if so, how it related to other concepts.

Before presenting the outcome of this analysis, it is important to stress that we areprimarily concerned with issues of quality (the kind of strategic factors that are relevantto crowd policing, how they relate to policing decisions, and the conditions under 

Patrick Cronin and Stephen Reicher 182

Page 9: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 9/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

 which they lead to the treatment of the crowd as homogenously dangerous). Rather than analyse how often these factors occur, our aim is to provide an understanding of 

 what they are through description of instances. Hence, the analysis consists of extractsfrom the transcript and its validity is ensured by making them sufficiently detailed toallow the reader to judge the analytic points they are used to sustain.

Given that our interest is in command level policing of crowd events, we have only used extracts from the participants who were police officers (what is more, the prisonofficers made few contributions). Extracts are numbered sequentially and participants

are identified by rank, age, and years of service. Thus, for example, (Cmdr 42, 20)denotes a commander aged 42 with 20 years of service, while (C Insp 49, 29) denotes achief inspector aged 49 with 29 years of service. Where any editorial clarification of  jargon or slang used in the extracts is necessary, it is placed in square brackets ([]).

Analysis

The analysis is divided into two sections that correspond to the concerns we identifiedin the introduction. In the first section, we outline the various strategic concerns thatimpacted upon the judgments and decision of officers, and we also consider the natureof that impact for each concern (Aim 1). In the second section, we look at how thenature and balance of strategic concerns shifts during the events and how this impactsupon the treatment of crowds, particularly upon the tendency to treat the crowd as a

homogenous and dangerous whole (Aims 2 and 2a).

Strategic concerns and crowd policing 

 Very early on in the event, while the crowd was still assembling prior to moving off,participants were informed that 40 ‘troublemakers’ (Categories A and B) had been

spotted in the crowd. As the officers discussed how to respond, one of them (C. Insp,54, 25) suggested, to general assent: ‘why don’t we just arrest them all?’ However,another officer was more cautious:

1. (C Insp, 48, 28): That’s an interesting concept. I wonder if we could get away with it.

I think our intelligence about those individuals would have to be very strong for us to justify,

let’s say in the eyes of solicitors branch and any subsequent court proceedings, where we

 would be held accountable for these decisions.

His input led participants to discard the idea of arrests. It was agreed to monitor thegroup but not to do any more at this stage. Two initial points flow from this interchange.The first is that officers do spontaneously invoke strategic concerns and, morespecifically, the ways in which they would be held to account by others. Second, theseaccountability concerns can be decisive in determining how they do act, over-riding

how they would like to act. Extract 1 is not an isolated instance of this. Throughout theexercise, our respondents invoked concerns about their accountability to a series of different audiences. These concerns can be divided into two broad categories: externalaccountability (accountability to audiences outside the police service) and internalaccountability (accountability to audiences within the police service). Let us consider each in turn.

External accountability often referred to a general concern with what ‘the

community’ would think of police actions. Officers were aware that they could beblamed for being too permissive and allowing disruption to occur. However, there was aparticular concern with being blamed for being too repressive and hence held

Policing crowd events   183

Page 10: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 10/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

responsible for initiating conflict. This is clear from the following quotation, which relates to the decision of how to act outside the library:

2. (C Insp, 48, 28): As managers of the event you are gonna be put under a lot of pressure by 

 your serials to get people kitted up [i.e. front line officers will want to put on protective riot

gear], sometimes I know it sounds horrible, sometimes you may have to take a couple of 

injuries.

The point is that if the police are seen to act prior to having sustained any injuries, then

they can be accused of being precipitate and of provoking violence. Once injuries havebeen sustained, then they are more likely to seen to be the victims than the aggressors inthe public eye, and their actions will be seen as responses to violence rather than violentin and of themselves. Indeed, more generally in our interviews and in other studies of 

crowd policing we have undertaken (Cronin, 2001) senior officers talk of ‘acceptabledamage’, which refers to the level of crowd violence which officers are prepared toendure in order to render the police reaction legitimate in the eyes of the public.

However, it was not simply a matter of waiting for events to enable the police tointervene without fear of blame by the public. Participants were actively concerned with interpreting events for the purposes of blame management. Thus, violence by some crowd members outside the library could be used to represent the crowd as athreat to the public and the police as its defender, rather than vice versa:

3. (Cmdr, 42, 20): What they have actually done at the library is present you with a current

and subsequent opportunity, they’ve attacked the local community if you want to put it that

 way, and if you’ve got your area press and publicity officers out and the local press there and

 you’ve got your police consultative group representatives out, you can spread that message.

In the above examples, the nature of ‘the community’ was left rather undefined as arethe precise mechanisms through which public judgments might impact on the police.However, there were times where participants were quite explicit about what thecommunity meant and where the ways in which this ‘public’ can affect them wasequally clear. Sometimes, as in the story of Margaret Thatcher’s phone call that we notedat the outset of this paper, the community referred to political representatives and directpolitical pressure. However, such pressure is relatively rare. More often, it referred to

formal investigatory processes, especially public inquiries. Officers were aware that if any event were to turn into a serious riot, then such an inquiry would be highly likely.

Given this ever-present possibility, our participants emphasized that all decisions mustbe made with an eye to their defensibility in front of a public inquiry. In an interview prior to the exercise itself, one officer stressed this with some fervour and at somelength:

4. (C Insp, 45, 25): And now, of course, we are into civil litigation, we are talking about

thousands of pounds, and apart from that, and this is secondary I know, but we are talking

about public safety. Hillsborough, Bradford, Heysel [all examples of football matches where

large numbers of people were crushed or burnt to death]. Absolutely. However hard you

 work, however hard you try, you will never get it right totally. That’s why people who don’t

take it seriously don’t realize what they are not doing and how vulnerable they are and their 

colleagues are, how vulnerable they make the service and how they can let the public down

 You mustn’t let down the public, ever, because as we have seen the public inquiry always

exposes the limitations of the police operation and further erodes our credibility.

Let us now turn to internal accountability. First, our public order commanders were justas concerned with an internal inquiry as they were with an external inquiry. In the case

Patrick Cronin and Stephen Reicher 184

Page 11: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 11/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

of large-scale disorder, they would be subject to internal investigation by a ‘neutral’senior officer, either from within the Metropolitan Police or an outside force if merited

in terms of serious complaints arising from the disorder (Metropolitan Police, 1999).The outcome could potentially lead to severe indictment even to the extent of endingone’s career. Consequently, as with public inquiries, officers suggested that decisionsmust always be made with an eye to how they may subsequently be viewed by policeinvestigators:

5. (C Insp, 45, 25): I do worry how kitting up would be seen in any internal police inquiry,

 which we seem to have quite regularly when a public order event has been seen to have

gone wrong.

Second, and somewhat less formally, our participants were concerned with their accountability to their peers, that is, fellow senior officers. Peers may not be able to sack  you, but one’s reputation among them is critical both to social acceptance and to career advancement – and critical to one’s reputation is the ability to remain in control of 

oneself and of events:

6. (C Insp, 49, 29): This is particularly so when ‘the wheel comes off’ [extreme and

unexpected events occur] as the officers will always remember how well you performed

under pressure and if you did not cope very well everyone will remember that – especially 

in the public order cadre [Advanced public order trained Met Officers].

This suggests that ‘control’ is a group norm. If this is so, then one would expect officersnot only to indicate that others value controlled and controlling behaviour (and that

they are motivated to express it for extrinsic reasons), but also that they value such behaviours themselves (and that they are motivated to express it for intrinsic reasons). We have already seen implicit evidence of this at the start of this section. That is, whenparticipants were given evidence of the presence of ‘troublemakers’ their initialreaction was to impose authority over them and remove them from the scene.

However, there is explicit evidence as well. As one participant (C Insp, 45, 25) said

 when asked by the instructors about the general approach that should be taken topolicing the crowd, ‘we dominate and control’. Again, when we provided feedback on our findings about accountability to a meeting of top-level public order commanders,

no one challenged their importance, but a number of people told us that we mustremember that being in control of events is possibly the most important thing to any 

police officer. This is a point to which we shall return in discussion. For now, however, we shall consider a third and final aspect of internal accountability: that of public order commanders to their junior officers.

The concerns here are somewhat different and they are well illustrated indiscussions about when orders should be given for junior officers to don riot gear. First,the matter was debated in general terms. One chief inspector argued that (C. Insp, 48,28), if commanders do not allow junior officers to ‘kit up’, then they themselves may 

decide ‘sod this, I’m gonna go in there kitted up’, at which point ‘you’ve lost control of  your officers’. Then another chief inspector voiced identical concerns with respect to what should happen outside the library:

7. (C Insp, 45, 25): It may be that that sustained missile attack only lasts a matter of 15–20

seconds and then they are going to go away. Well, that is a judgment call on the people in

charge at the time, balanced against that is the effect that it has on the officers because if youare not seen to take action by the junior officers there is a danger of them acting without

senior officers.

Policing crowd events   185

Page 12: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 12/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

The concern of senior officers was that, however good the strategies that they set, their ability to deliver these strategies depends upon the cooperation of junior officers. If 

 junior officers disapprove of command decisions (which is particularly likely wherepermissive policing increases the likelihood of injuries to front line officers), then theseofficers can and do subvert the command strategy.

In effect, these observations take us full circle. We began our analysis by showingthat external accountability concerns led commanders to delay the use of repressivetactics and to countenance injuries to their own officers. Now, we see that internal

accountability concerns pressured them to endorse such tactics in order to avoid policeinjuries. Therefore, as well as listing the various different accountability concerns thatimpact on public order commanders, this section has also shown that the police facemultiple accountability concerns simultaneously and that different concerns may 

influence their decisions in different and sometimes contradictory ways. This suggeststhat it is not sufficient to look at specific accountability concerns in isolation, but rather  we must look at their overall balance at any given point in time. Should that balancealter, then one would also expect the nature of command decisions to alter. This is thesubject of analysis in the next section.

Phase, accountability and decision-making 

It is possible to divide the overall crowd scenario into three phases: first, non-conflict, which covers the period from initial assembly until demonstrators reached Stoke

Newington Church Street (Points 1 and 2 in our account of the scenario providedabove); second, incipient conflict, which covers the throwing of placards, bricks andbottles by some crowd members outside Stoke Newington library (Point 3); and third,conflict, which covers the period of intense and generalized violence from the pointcrowd members reached the junction of Stoke Newington Church Street and StokeNewington High Street (Points 4–6). It should be stressed that these are not divisions or labels used during the exercise itself, but rather are categories that emerge from our analysis. They are of use to the extent that we can identify shifts in the nature of accountability concerns as a function of phase and also see associated changes in police

perceptions and actions. We shall consider each of these in turn.

Phase and the balance of accountability concerns

In the non-conflict stage of events, officers were particularly aware of the way in which external audiences may hold them to account for intervening against crowd members.Such concerns were clearly expressed in Extract 4 above. To intervene where there is no violence can lead to accusations both of denying rights and of provoking violence. What ismore, as one participant (C Insp, 48, 28) pointed out, this might place front line policeunnecessarily at risk: ‘you may get some injuries to officers when they try to arrest these

40 hard core as they may get help from their mates in the crowd’. Hence, in this context,internal accountability to junior officers did not point unambiguously towardsintervention. Since external accountability militated strongly against intervention andinternal accountability militated less towards intervention, in the pre-conflict phase, theoverall balance of accountability concerns was clearly such as to limit intervention.

In the incipient conflict phase, the balance shifted. While it might be slightly harder for external audiences to hold commanders to account for provoking violence that has

already started, they can still be held responsible for escalating a minor problem by intervening too early. On the other hand, as we have already seen, their own junior officers may undermine command strategies if they do not intervene. At this stage, then,

Patrick Cronin and Stephen Reicher 186

Page 13: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 13/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

the different accountability concerns were evenly balanced and clearly placedcommanders in a dilemma where they could suffer negative consequences whichever 

 way they acted. This was reflected in a lengthy discussion as to how to act outside thelibrary and it was encapsulated in the following contribution:

8. (C Insp, 48, 28): I would be wanting to ask at this stage for the reserveofficers to be kitted

up because you seem to be in escalating violence. Your officers have already been injured,

and you have to balance being seen to protect them with inflaming the whole crowd. It is a

balancing act – you have to ask for these officers to be kitted up while, at the same time,

ensuring they are tucked away around the corner so as not provoke the crowd.

In this case, the dilemma was resolved by a strategy that faced both ways and could berepresented differently to different audiences. To external audiences, riot police werenot deployed in front of the crowd and so could not inflame them. To junior officers, riotequipment was brought out in order to prepare for action.

By the time that violence has become generalized and serious (the conflict phase of the event), external audiences, like internal audiences, were seen as unlikely to criticizethe police for provoking or inflaming violence, but rather for failing to intervene in order to protect people and property against crowd members. It is notable, for instance, how in the following extract, both danger to the public and danger to the police wereinvoked as reasons for draconian interventions:

9. (Cmdr, 42, 20): They are now chucking petrol bombs at me: : : I would consider whether 

the horses were an option at this stage, if they were not because of the crowd dynamics,

because we have people trapped in burning buildings.

Thus, since external accountability concerns switched round so as become consonant with internal accountability concerns, dilemmas of intervention became resolved. Theonly sanctions from either the public or one’s own officers would be due to non-intervention. Measures taken against the crowd were perceived as likely to win plaudits

all round. Overall, the shift in the balance of accountability concerns from non-conflictto incipient conflict-to-conflict phases of the event was towards steadily greater pressurefor intervention.

Changing accountability and public order decision-making To the extent that there was an overall shift in the balance of accountability concerns asa function of phase, one would also expect there to be a series of more specific shiftsacross phases of the event in the way that public order commanders used informationabout the crowd and how they acted towards crowds. In each case, the shift shouldprogress from lesser to greater emphasis on danger and intervention as the event wentfrom no conflict, to incipient conflict, to sustained conflict.

It is important to stress that the shifts we are expecting should occur at the level of collective agreement rather than individual suggestion. That is, the nature of theexercise was such that participants were encouraged to suggest different options as abasis for discussion. We would therefore expect a variety of positions to be expressed ateach phase. What we are concerned with, however, are those positions that securedcollective agreement and shaped group discussion either by resolving overtdisagreement or (as was more usual) by being taken for granted by others and

providing agreed terms for subsequent discussion. Such consensualizing positions bestreveal the stance that people take as group members (Haslam, Oakes, Turner, McGarty, & Reynolds, 1998). Consequently, we are not looking for quantitative shifts but rather 

Policing crowd events   187

Page 14: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 14/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

 wish to look more closely at the ways that debates and decisions were shaped atdifferent phases of the event.

Use of information: The dilemmas surrounding information usage stem from the fact thatpublic order commanders have various sources of information that frequently givedifferent indications of danger. Thus, participants were initially given a briefing from

special branch (SB; the intelligence wing of the police), which indicated that up to 900‘troublemakers’ would be present. They were also given reports from FITs (which 

operate on the ground during events) indicating that 40 troublemakers had beenspotted. Finally, they were told that the crowd was 5,000 strong.In the non-conflict stage, participants took for granted the lowest estimate provided by FITs as the relevant information on which to proceed. When the figure of 40 wasmentioned, it was not challenged by others. Participants debated what to do with these40 (arrest them or leave them alone – see Extract 4 and the preceding discussion), butthere was no debate as to whether or not they constituted the sum total of dangerous

people in the crowd. That is, no concerns were expressed concerning the remaining4,960 or so demonstrators.

Moreover, officers invoked models of crowd process that suggest that crowd

 violence derives from the pathological character of particular individuals in thecrowd who therefore are unlikely to draw wider sections of the crowd into conflict,except, possibly, for their personal friends or ‘mates’. Thus, when participants weredebating whether rowdy behaviour alone can be used as a sign of crowd danger, thefollowing intervention was decisive:

10. (Cmdr, 42, 20): No, they’ve got to be people that you have identified as key players, key 

orchestrators, and there’s a history to them that forms the basis for you suspecting a Breach 

of the Peace by their presence or demonstration.

 As soon as the violence outside the library started – the phase of incipient conflict– things changed. Recall that, at this point, the numbers themselves did not changeas only the 40 so-called ‘hard core’ were involved in the violence. However, whatdid change was the weight put on the different numbers. At this point, the SBestimate of 900 became taken for granted and the fact that far fewer people wereactually involved in conflict became the basis for a discussion on where the ‘missing

trouble makers’ might be. As one officer (C Insp, 45, 25) asked, if the police haveonly met some 50 or so ‘troublemakers’ ‘where’s the other 800 and odd?’ Another officer suggested that they may be identified through their reactions to policeinterventions:

11. (C Insp 54, 26): Up to a thousand people who are troublemakers. We have only 

identified 40 of them. If we go and arrest those, that could inflame the others and they could

cause trouble here or further on.

Note also from this extract that the model of crowd process changed somewhat. While it was still true that only people with a certain predisposition were seen to initiate violence, others were seen as prone to become drawn into conflict through the crowdprocess. This is a hybrid of those models that explain crowd violence in terms of 

individual pathology and those which propose that all individuals become pathologicalin the crowd – what we have termed elsewhere the Allportian and the Le Bonian models(cf. Allport, 1924; Le Bon, 1895; Reicher, 2001).

Patrick Cronin and Stephen Reicher 188

Page 15: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 15/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

In the final phase of sustained conflict, the estimate changed yet again. Both the SBand the FIT estimates were discarded. Rather, as is clear from the following extract, it

 was assumed that the entire crowd were troublemakers, and discussion of tacticsproceeded on this basis.

12. (C Insp, 39, 20): We have to treat all 5,000 members of the crowd as dangerous and

liable to join in the violence against police.

Note that the figure of 5,000 dangerous people was not based on what people wereactually doing and on how many showed evidence of being troublesome. Rather, theclaim was that all crowd members are in principal potentially troublesome and, hence,the only information that counts is how many people are in the crowd. While there wasstill some use of the Allportian notion of individual pathology in order to underpin thisclaim (all crowd members must be anti-authority in order to be on the march), at thispoint we also saw the use of full blown Le Bonian ideas, according to which even themost respectable people can lose their judgment and become subject to crowd

contagion. This was eloquently expressed by one chief inspector:

13. (C Insp, 45, 25): It’s amazing how people who are extremely well behaved and hold

good jobs and have a good public standing and are law abiding citizens get involved in all

sorts of stuff but when you put it to them. They don’t ever remember doing it, they just get

caught up in the dynamics of the mob and start joining in with the troublemakers.

The shifts in which information was attended to from phase to phase were not simply aresult of different evidence becoming available concerning the extent of trouble. Rather,participants used different estimates of danger as the basis upon which evidence of 

trouble was sought. As the event became more conflictual, progressively less emphasis was placed on information that suggested limited danger and progressively more on that which played up the danger of the crowd. This was even clearer when it came to shifts

in the use of models of crowd process. After all, none of the information that wasprovided related to these models. Rather, the model affected the significance accordedto information by indicating how conflict was likely to develop. Therefore, as conflictdeveloped, it was not only that officers saw more people as actually dangerous but thatthey saw greater potential for others to become dangerous. In this way, the selection andassessment of information provided a warrant for interventions against progressively 

 wider sections of the crowd.

Treatment of crowd members: In the non-conflict phase, participants differed over how totreat the 40-strong ‘hard core’. However, all agreed that, whatever tactics were used,they needed to be highly targeted and differentiated. To borrow the metaphor of oneparticipant: (Cmdr, 42, 20), ‘it’s about shooting the chiefs’. When incipient conflictbroke out, there were differences once again. However, the parameters of debate hadchanged. Now, no one suggested that intervention be limited to the ‘chiefs’. Rather, theissue had become ‘how wide to spread the net’. Hence, participants discussed the

relative merits of arresting people for serious or minor crimes, and the practical andlogistical difficulties associated with such interventions.

14. (C Insp, 48, 28): You can arrest for a very serious crime or minor crime. Some of themthey will be smoking things like cannabis, just purely to wind the PCs up, so your arrest

policy has got to be very clear, what you are gonna arrest for.

Policing crowd events   189

Page 16: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 16/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Some of the participants began to suggest tactics that do not differentiate betweendemonstrators at all. Thus, one commander (Cmdr, 42, 20) suggested using officers in

riot gear against ‘the crowd’ as a whole ‘to drive them past the library’.Once into the conflict phase, all considerations of differentiation and targeting werediscarded. Thus, one officer explained how he would use horses to stop crowdmembers from proceeding despite their numbers:

15. (Cmdr, 42, 20): They can’t get a lot of momentum up to push, so you could probably 

stand and fight with the crowd, I suppose, and start trying to force them, to stop them

coming south so they’ve got no choice but to stand still.

The only debate that occurred at this stage was over the timing and combination of such tactics: should riot police be used on foot, should horses be used, and shouldrubber bullets be deployed? However, nobody voiced any concerns about the fact thatall these tactics treated anyone who remained present as equally dangerous and drew allof them into conflict with the police whether they liked it or not.

Discussion

The first aim of this study was to analyse both the nature and the impact of strategicconcerns upon public order commanders. Perhaps the simplest and clearest of our findings is that accountability has a critical influence upon the decisions of these

commanders. From the start of the exercise, our participants were constantly aware of how, in a crowd event, others would judge their actions, and of the consequences these judgments would have for their future as police officers. These can range from therelatively trivial and informal (loss of reputation, reduced promotion prospects) to theserious and formal (demotion, dismissal, or even legal proceedings). Officers wereaware that, during events, they are simultaneously faced with multiple accountability 

pressures from multiple audiences, with some external to the police and some internal.These different sources of accountability frequently placed pressure upon officers to actin different ways. What was important, then, was not any one form of accountability but

the overall balance of accountability concerns that operated at any given moment intime.

Our second aim was to investigate whether this balance changes over time, and if so, how and with what effect upon policing decisions. Here, too, the findings wereclear. The nature of individual accountability pressures varied as a function of thephase of the event and hence the overall balance changed as well. More specifically,as violence escalated, officers portrayed internal audiences as dropping any ambiguity about intervention and external audiences as dropping support for non-intervention.This had a clear impact upon judgments and decisions. With increasing violence,

officers increasingly sought evidence of danger rather than evidence of safety, andincreasingly treated crowd members as dangerous. This relates directly to our specificconcern with the conditions under which officers treat crowd members as ahomogenous and dangerous entity. From the moment that conflict started (even when only a small proportion of crowd members were involved in it), such tendencies began to be observed among our participants and they becamepredominant as conflict escalated.

These findings have theoretical and practical implications. Theoretically, our work serves to emphasize the importance of accountability processes in determining groupaction. As Lerner and Tetlock (1999, p. 255) note in a recent review, despite the fact that

Patrick Cronin and Stephen Reicher 190

Page 17: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 17/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

‘accountability is a modern buzzword’, research on the topic remains scarce,particularly when it comes to groups. Indeed recent reviews – even those that focus on

collective judgment and decision-making – do not even mention the term (e.g. Baron & Kerr, 2003; Stasser & Dietz-Uhler, 2001).

 As we suggested in the introduction, one reason for this lies in the methods thatpsychologists use and hence the types of group we study. Research is dominated by experimentation where our encounters with others are brief and inconsequential. As Lerner and Tetlock (1999, p. 270) put it, ‘there is no shadow of either past or 

future’. Moreover, the use of informal groups where membership is a matter of choice may also lead to the importance of accountability being underestimated. For our officers, the reason why the future was important was because their decisionsabout how to treat the crowd could impact upon their position in the police

service, which, in turn, depended upon institutional procedures surrounding groupmembership: the fact that others decide if they can join the group, where they stand in the group, and even whether they remain in the group. Our investigationof such formal groups shows that accountability is a central issue at a psychological,as well as a political, level.

However, it is equally important to place analyses of accountability in the context of group process. Whereas conventional approaches tend to present an individualistic andpassive picture of the subject who is motivated simply to gain approval from others(cf. Emler & Reicher, 1995), our analysis supports a model of the motivated group

member acting under constraint. On the one hand, the reason why our officers wereconcerned with the reactions of others was not primarily to get obtain individualrewards but rather to remain police officers and act as such. In some cases, the two wereconfounded. For instance, censure from a public inquiry could lead to removal fromone’s job and hence both loss of group membership and loss of individual salary.However, in other cases, the collective level of concern was predominant. Thus,concerns about accountability to junior officers were discussed entirely in terms of how to get command strategies delivered to maximum effect. The one constant throughout,then, was the group based motive.

On the other hand, there is evidence that command officers do have their ownnorms and values on which they would wish to act, but that their action is moderated by accountability concerns. This takes us back to comments made in the analysis after Extracts 1 and 6. In both cases, there was explicit reference to the value of ‘control’ andthat commanders are concerned to dominate an event. Their primary concern was toget crowd members to act in terms defined by the police (rather than, say, facilitating thecrowd to behave in their own terms to the extent that these fall inside the law).However, there was equally explicit reference made to the fact that accountability constraints sometimes render this impossible. To borrow from Extract 1, officers are

always dealing with the question ‘can we get away with it?’Overall, then, our study supports the argument that group processes (more

specifically, social identity processes) and accountability processes need to be analysedin conjunction with each other rather than counter-posed. In other words, it validatesthe premise of the SIDE tradition (Postmes  et al., 2000; Reicher  et al., 1995). Indeed, asthe first significant field study explicitly undertaken within this tradition, it shows how this approach is especially relevant when seeking to account for the actions of members

of real groups with ongoing histories, enduring commitments, and formal proceduresrelating to membership. Moreover, it builds on previous work by showing that one canbe accountable to multiple in-group and out-group audiences. It also shows that the

Policing crowd events   191

Page 18: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 18/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

 various accountability pressures operate simultaneously and that it is the balancebetween different pressures, rather than any single pressure, that is important. We have

also illustrated that the balance of accountability pressures affects both the use of information and decision concerning action, and that this balance can change during anepisode thus resulting in changes of judgment and action.

In terms, of the practical implications of our study, our aim is not to extrapolatefrom the precise form that accountability pressures took in the exercise that westudied. In different circumstances, different audiences may become relevant. For 

instance, in a demonstration targeting a foreign embassy then international concernsmay come into play. Second, the pressure from a given audience may vary from eventto event. Thus, whereas accountability to political audiences generally militatedagainst early police intervention in our exercise, there have been other circumstances

(e.g. the 1999 visit by Chinese President Jiang Zemin to London) where fear of a‘diplomatic incident’ led governmental pressures to operate in precisely the oppositedirection. Therefore, in any given context, there is a need for a situated analysis of the accountability pressures that are operating, the balance between these differentpressures, and the way this changes over time. This will be critical to the policingdecisions that are made.

Even if the type of accountability pressure in a given situation may vary accordingto the audience and even if the type of accountability pressure from a given audiencemay vary according to the situation, in our study, there was a constant in the way in

 which accountability pressures altered. That is, as violence increased, every audienceeither maintained or increased its pressure for generalized intervention. Certainly,none became less interventionist. Perhaps this attests to the prevalence of Le Bonianideas in our society (cf. Reicher, 1996b; Reicher & Potter, 1985). That is, as soon asthere is any violence in a crowd, it is always arguable that escalation is inevitable andhence the police can never be unequivocally held to account for having inflamed thesituation, regardless of how they acted. Whatever the case may be, the implicationsare serious.

It is precisely when some sections of the crowd begin to act confrontationally that

the police response is most critical. At this important point, the police differentiatebetween crowd members and accountability concerns begin to switch round andpressure officers towards less differentiation. Thus, it is not only factors internal to thepolice that lead to generalized intervention (Drury  et al., 2003; Reicher  et al., 2005;Stott & Reicher, 1998b) but also factors external to the police. Equally, any attempts tochange police practices in the direction of more differentiated strategies and tacticsmust deal with the more general view of crowds in society as well as the views of thepolice themselves. As long as the procedures that are easiest to justify to outsiders tendto escalate violence (cf. Simonson & Nye, 1992), then any sophistication of police

insiders will count for little.Of course, these conclusions must be tempered by the limitations of our study –

notably, the fact that our analysis is based upon an exercise and not a real crowd event.In the introduction, we explained the advantages of such an analysis in terms of gettingexplicit accounts of perceptions and rationales for action and this is reflected in therichness of our transcript. However, we acknowledge that the artificiality of the exercisemay have undermined our findings. Most obviously, the stress on accountability 

concerns could have been a simple demand characteristic that had been communicatedby the instructors during the course and which was expressed by participants duringthe exercise.

Patrick Cronin and Stephen Reicher 192

Page 19: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 19/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Such an argument is certainly plausible. However, if anything, any mention of accountability during the course (which was attended in its entirety by the first author)

 was made by the participants rather than instructors. Indeed, in the analysis we haveused, interview material that shows that officers are acutely aware of the need to show the public that the police are not responsible for initiating violence. Far from pointing tothe fact that accountability concerns are an artefact of the exercise, this suggests that theconcerns participants express in the exercise reflect those expressed during actualevents.

 Another possible criticism of our analysis is that any shifts in police accountability concerns during the event may result from the terms in which the conflict wasdescribed rather than from changes in the nature of the conflict itself. If the informationhad been phrased differently or if people were provided with fuller and richer 

information sources (e.g. radio and video) as in a live event, then we might haveobtained different responses.

Undoubtedly, the responses of participants are not to an immediate social reality butrather to a social reality that is constructed for them through language (cf. Edwards & Potter, 1992; Potter, 1996). Hence, the terms in which the ‘injects’ were phrased are of critical importance. Having said that, it is important to draw a distinction betweenlanguage that, in itself, carries certain entailments and language that invokes culturally bound systems of knowledge through which conclusions are drawn. For instance, if theterm ‘incipient conflict’ had been used in the inject relating to events outside Stoke

Newington library, it is arguable that the word ‘incipient’ points to the start of anescalating process and from this point, participants would see it likely that more people would be drawn into the conflict, to look for evidence of this, and to treat crowd

members accordingly.However, as we have stressed, the term was an analytic label. Participants were told

that 40 people were throwing placards bricks and bottles. It is hard to see how such information suggests escalation and points to generalized interventions without beingmediated by assumptions about crowd process and audience reactions. If the latter istrue, then it suggests that any information indicating conflict, and not only thatcontained in our specific injects, would lead to a shift in accountability concerns andpolice decisions. It is important to reiterate that the injects were worded to mirror thosethat officers would receive in a live event. In general, they were based on actual

information provided about previous events. Thus, even if the language of the injects was important in producing responses, this constitutes a link rather than a differencebetween our study and the dynamics of real events.

To conclude, the analysis in this paper provides at least provisional dataconcerning the importance of accountability concerns in moderating the behaviour of groups outside the laboratory and, more specifically, of the conditions under  which the police may act in ways that serve to escalate collective conflict. The rich and explicit nature of the data provides some vindication of our decision tocommence investigation of these issues using a training exercise. Moreover, there are

good reasons to believe that any particularities in the way that information wasprovided to participants in this exercise are reproduced in actual events.Nevertheless, our aim in this study was as much to generate as to test hypotheses.Undeniably, the questions generated by this study need to be complemented by 

further evidence gathered in the field. To borrow an old adage from Robert Park (hopefully, not to literally), ‘go get the seat of your pants dirty in real research’(McKinney, 1966, p. 74).

Policing crowd events   193

Page 20: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 20/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Acknowledgements

This research was made possible by an ESRC research studentship for the first author:

ROO429824363. Thanks to the Metropolitan Police for granting access for this study and

specifically to the Public Order Branch of New Scotland Yard for facilitating the research.

References

 Allport, F. (1924). Social psychology. Boston: Houghton Mifflin.

Baron, R. S., & Kerr, N. L. (2003).   Group process, group decision, group action   (2nd ed.).Buckingham: Open University Press.

Barreto, M., & Ellemers, N. (2000). You can’t always do what you want: Social identity and self-

presentational determinants of the choice to work for a low-status group.  Personality and 

Social Psychology Bulletin, 8 , 891–906.

Barreto, M., Spears, R., Ellemers, N., & Shahinper, M. (2003). Who wants to know? The effect of 

audience on identity expression among minority group members.   British Journal of Social 

 Psychology, 42, 299–318.

Cronin, P. D. J. (2001).   The psychology of crowd policing. Doctoral thesis, University of St.

 Andrews.

Douglas, K. M., & McGarty, C. (2001). Identifiability and self-presentation: Computer-mediated

communication and intergroup interaction.   British Journal of Social Psychology,   40,

399–416.

Drury, J., & Reicher, S. (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., & Reicher, S. (2000). Collective action and social change: The emergence of new social

identities. British Journal of Social Psychology, 39, 579–604.

Drury, J., Stott, C., & Farsides, T. (2003). The role of police perceptions and practices in

the development of public disorder.   Journal of Applied Social Psychology,   33, 1–23.

Edwards, D., & Potter, J. (1992). Discursive psychology. London: Sage.

Ellemers, N., Spears, R., & Doosje, B. (2002). Self and social identity.   Annual Review of 

 Psychology, 53, 161–186.

Ellemers, N., van Dyck, C., Hinkle, S., & Jacobs, A. (2000). Intergroup differentiation in social

context: Identity needs versus audience constraints.   Social Psychology Quarterly,   63,

60–74.

Emler, N., & Reicher, S. (1995).  Adolescence and delinquency. Oxford: Blackwell.

Fielding, N. G. (1991).  The police and social conflict . London: Athlon.Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative

research. New York: Aldine.

Haslam, S. A., Oakes, P. J., Turner, J. C., McGarty, C., & Reynolds, K. J. (1998). The group as a basis

for emergent stereotype consensus. European Review of Social Psychology, 9, 203–239.

Kellehear, A. (1993). The unobtrusive researcher . St Leonards: Allen and Unwin.

Klein, O., & Azzi, A. E. (2001). The strategic confirmation of meta-stereotypes: How group

members attempt to tailor an out-group’s representation of themselves.   British Journal of 

Social Psychology, 40, 279–293.

Klein, O., & Licata, L. (2003). When group representations serve social change: The speeches of 

Patrice Lumumba during the decolonization of Congo.  British Journal of Social Psychology,

42, 571–594.

Klein, O., Licata, L., Durala, I., & Azzi, A. E. (2003). How European am I? Prejudice expression and

the presentation of social identity.  Self and Identity, 2, 251–264.

Le Bon, G. (1895, translated 1947). The crowd: A study of the popular mind. London: Ernest Benn.Lerner, J. S., & Tetlock, P. E. (1999). Accounting for the effects of accountability.  Psychological 

 Bulletin, 125, 255–275.

Patrick Cronin and Stephen Reicher 194

Page 21: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 21/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

McKinney, J. C. (1966).  Constructive typology and social theory. New York: Appleton-Century-

Crofts.

Metropolitan Police (1999).   Strategic command training. Internal police document, London:

New Scotland Yard.

Miles, M. B., & Huberman, A. M. (1994).   Qualitative data analysis   (2nd ed.). Thousand Oaks,

CA: Sage.

Northam, G. (1988). Shooting in the dark: Riot police in Britain. London: Faber and Faber.

Postmes, T., & Spears, R. (1998). Deindividuation and antinormative behavior: A meta-analysis.

 Psychological Bulletin, 123, 238–259.

Postmes, T., Spears, R., Lea, M., & Reicher, S. (2000). SIDE issues centre stage. Amsterdam: Royal

Netherlands Academy of Arts and Sciences.

Potter, J. (1996). Representing reality. London: Sage.

Reicher, S. (1984). The St. Pauls 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. (1987). Crowd behaviour as social action. In J. Turner, M. Hogg, P. Oakes, S. Reicher, & 

M. Wetherell (Eds.),  Rediscovering the social group   (pp.171–202). Oxford: Blackwell.

Reicher, S. (1996a). 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. (1996b). The crowd century: Reconciling theoretical failure with practical success.

 British Journal of Social Psychology, 35, 535–553.

Reicher, S. (2001). The psychology of crowd dynamics. In M. Hogg & S. Tindale (Eds.),  Blackwell 

handbook of social psychology: Group processes   (pp. 182–208). Oxford: Blackwell.

Reicher, S., & Hopkins, N. (2001).  Self and nation. London: Sage.

Reicher, S., Hopkins, N., & Condor, S. (1997). Stereotype construction as a strategy of influence.

In R. Spears, P. Oakes, A. Haslam, & N. Ellemers (Eds.), Stereotying and social identity. Oxford:

Blackwell.

Reicher, S., & Levine, M. (1994a). Deindividuation, power relations between groups and the

expression of social identity: The effects of visibility to the out-group. British Journal of Social 

 Psychology, 33, 145–163.

Reicher, S., & Levine, M. (1994b). On the consequences of deindividuation manipulations for the

strategic communication of self: Identifiability and the presentation of social identity.

 European Journal of Social Psychology, 24 , 511–524.

Reicher, S., Levine, M., & Gordijn, E. (1998). More on deindividuation, power relations between

groups and the expression of social identity: Three studies on the effects of visibility to the

ingroup. British Journal of Social Psychology, 37 , 15–40.

Reicher, S., & Potter, J. (1985). Psychological theory as intergroup perspective: A comparativeanalysis of scientific and lay accounts of crowd events.  Human Relations, 38 , 167–189.

Reicher, S., Spears, R., & Postmes, T. (1995). A social identity model of deindividuation

phenomena.  European Review of Social Psychology, 6 , 161–198.

Reicher, S., Stott, C., Cronin, P., & Adang, O. (2004). An integrated approach to crowd psychology 

and public order policing.   Policing, 27 , 558–572.

Scheepers, D., Spears, R., Doosje, B., & Manstead, A. S. R. (2003).   Diversity in in-group bias:

Structural factors, situational features, and social functions. Amsterdam: University of 

 Amsterdam.

Simonson, I., & Nye, P. (1992). The effect of accountability on susceptibility to decision errors.

Organizational Behaviour and Human Decision Processes, 51, 416–446.

Spears, R., & Lea, M. (1992). Social influence and the influence of the social in computer-mediated

communication. In M. Lea (Ed.),   Contexts of computer-mediated communication

(pp. 30–65). London: Harvester Wheatsheaf.

Spears, R., Lea, M., Corneliussen, R. A., Postmes, T., & Ter Haar, W. (2002). Computer-mediatedcommunication as a channel for social resistance: The Strategic Side of SIDE.   Small Group

 Research, 33, 555–574.

Policing crowd events   195

Page 22: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 22/23

Copyright © The British Psychological SocietyReproduction in any form (including the internet) is prohibited without prior permission from the Society

Spears, R., Postmes, T., Lea, M., & Wolbert, A. (2002). When are net effects gross products?

The power of influence and the influence of power in computer-mediated communication.

 Journal of Social Issues, 58 , 91–107.

Stasser, G., & Dietz-Uhler, B. (2001). Collective choice, judgement and problem solving.

In M. Hogg & S. Tindale (Eds.),  Blackwell handbook of social psychology: Group processes.

Oxford: Blackwell.

Stott, C., & Drury, J. (2000). Crowds, context and identity: Dynamic categorization processes in

the poll tax riot.  Human Relations, 53, 247–273.

Stott, C., Huchison, P., & 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., & Reicher, S. (1998a). How conflict escalates: The inter-group dynamics of collective

football crowd violence. Sociology, 32, 353–377.

Stott, C., & Reicher, S. D. (1998b). Crowd action as intergroup process: introducing the police

perspective. European Journal of Social Psychology, 28 , 509–529.

Strauss, A. L., & Corbin, J. (1998).  Basics of qualitative research: Techniques and procedures for 

 Developing Grounded Theory (2nd ed.). Thousand Oaks, CA: Sage.

Townshend, C. (1993). Making the peace: Public order and public security in modern Britain .

Oxford: Oxford University Press.

 Waddington, D. (1992). Contemporary issues in public disorder . London: Routledge.

 Waddington, P. J. (1987). Towards paramilitarism dilemmas in policing crowd disorder.  British

 Journal of Criminology, 27 , 37–46.

 Waddington, P. J. (1994). Liberty and order: Public order policing in a capital city. London: UCL 

Press.

Received 30 March 2004; revised version received 22 November 2004

Patrick Cronin and Stephen Reicher 196

Page 23: Cronin&Reicher 2006

8/10/2019 Cronin&Reicher 2006

http://slidepdf.com/reader/full/croninreicher-2006 23/23