Swarming and the social dynamics of group violence

Foreword | Riots such as the Cronulla and Macquarie Fields occurrences and media reports of large numbers of individuals gatecrashing parties understandably raise community concern about the prevalence and causes of group violence. This is a difficult area to research as the number of events is typically low, although their impact can be high. There are also different forms of mass groupings of individuals with higher and lower levels of associated violence. This paper provides a broad overview of these different group formations and identifies commonalities between particular group formations that are associated with violence. The paper also identifies two community policing initiatives focused on improving police-community relations. Such understanding will assist in the development of specific policy and practitioner responses that seek to prevent or reduce the escalation of group violence when there is the unexpected gathering of large numbers of people.

Toni Makkai

One of the key challenges for contemporary policing is how best to weigh up an appropriate and effective use of tactical force when it comes to events and situations featuring large crowds. Coupled with tactical evaluation are issues of community policing and the strategic interests of police in maintaining and enhancing general community relationships. For example, police have increasingly analysed and responded to planned events, such as schoolies week and football matches (including soccer), in ways that emphasise community safety and involve a wide range of community preparations, professional collaborations and anticipated responses.

The aim of this paper is to explore the nature of group behaviour and group violence that is essentially spontaneous in nature. One example of this is the Cronulla riots. Another example is gatecrashing. To some extent, these two kinds of events (beach riots and house party invasions) are united in bearing a relationship to the phenomenon of swarming - the unexpected gathering of large numbers of people in particular public locales. Swarming may or may not feature violence. It does, however, involve large crowds - crowds that may occasionally transform into 'mobs'. The size of the crowd is what also transforms a private home or private party into a public event via the spilling out of people on to footpaths and surrounding streets and lawns.

This paper considers different kinds of group formation in the public domain, and tries to make sense of the different kinds of group behaviour that may emerge. The intent is to stimulate thinking about the complexities of group behaviour generally, in order to further refine and develop intervention tactics and strategies that are relevant to the specific phenomenon in question. Such an exercise has potentially significant benefits in terms of policy development and police management strategies. For instance, perusal of five years of media reports on gatecrashing highlighted that police intervention could on occasion be a source of amplification of excitement for some gatecrashers: they relished the thought of being chased through the suburbs by the local police. In other instances, the confrontational style of the police - especially if they showed up in numbers, and in riot gear - was seen to be a precipitating cause of the subsequent crowd violence. In other words, understanding of group behaviour and crowd dynamics is also about understanding how the escalation of violence is sometimes linked to the type and intensity of police intervention itself.

Group formation and group behaviour

Swarming may be organised or spontaneous; it may engender or be associated with violence, or it might not. If group differences and differences in group violence are to be understood, then those instances in which large gatherings occur need to be identified. Those gatherings in which violence, generally speaking, is not on the agenda, nor is it prevalent are considered first.

Raves are essentially dance parties in which a large number of people gather together in one site for the purpose of dancing. They are less prevalent today (in part due to a movement into commercial clubs) and in some jurisdictions are virtually non-existent due to extensive legislative and police intervention. Nevertheless, the history of raving is essentially a history of organised events, outside normal or legal venues, with an emphasis on crowd anonymity, fun, drug taking which is directly related to the dance atmosphere, and nonviolence (Chan 1999). The venue and music were predetermined in the sense that considerable organisation has to go into planning a rave event. Communication via mobile phone, internet links and friendship networks were essential to the spontaneity of the gathering, since the time and location were usually only shared out at the last moment. The important thing about raves is that they tended to celebrate 'the beat', and violence of any kind was either not tolerated or certainly frowned upon.

Flash mobs apparently originated in Australia, but in recent times have become a big hit in places such as the United States. A flash mob is basically a group of anonymous individuals who meet, apparently inexplicably, in particular locations, do something 'silly', and then disperse within 10 minutes (Flash mobs website; Galvin 2003). Examples include a flash mob in Sydney where over one hundred people converged on a statue of Queen Victoria at Town Hall and barked and hissed liked cats and dogs before dispersing. In San Francisco members of a flash mob spun in circles while crossing a busy street, and then repeated this for 10 minutes before disappearing. A Rome flash mob flooded a books and music megastore, asking for nonexistent books. They then broke into applause, before disappearing. Rarely do flash mobs involve violence, although sometimes authority figures have been heavy handed in dispersing the groups. Again, this is generally seen as a fun event, an instance of organised spontaneity, for the specific purposes of transgressing the mundane and the commercial.

These types of swarming are relatively benign. They involve large groups of people congregating in specific locations, but with the intent of engaging in often quite physical, but nonviolent activities. There are rules that guide each particular kind of gathering, and a particular cultural environment that precludes intentional violence of any kind. It is not 'cool' to be aggressive towards other participants or to onlookers. The focus is clearly on fun and the excitement of the moment. Violence is not in the psychology or paradigm in which these groups have been conceived.

Youth gangs, by contrast, are implicated in violence almost by definition. The concept of swarming is generally less relevant in relation to gang violence as such. While periodically gang members may amass in numbers, using SMS and phone technologies, their engagement in street fights in particular includes organised battles as well as spur-of-the-moment conflicts. Gangs vary greatly in the level of organisation and identification. Contemporary Australian gang research emphasises the fluid nature of youth gang formation, while acknowledging the centrality of violence to gang membership compared with those young people who do not identify as a gang member (White 2006; White & Mason 2006). While sometimes involving large numbers of people, gang violence tends to be highly targeted in terms of protagonists. It is rarely random, and occurs on a frequent basis. It is not surprising but is central to the very idea of gang-related behaviour.

Riots describe a situation in which large numbers of people seem to spontaneously engage in unlawful, antisocial and violent behaviour. Recent riots in Australia have seen hundreds of people take to the streets, generally directing their anger at property such as cars, and to authority figures such as the police. The triggers for riots on Palm Island and in the Sydney suburb of Redfern were the deaths of young Indigenous people apparently in connection with some kind of police intervention. So too, in the western suburbs of Sydney, at Macquarie Fields, things came to a head after a car chase led to the death of two young local men (Lee 2005). Longstanding resentments within marginalised communities can suddenly come to the fore when circumstances change quickly. As analysis of the Bathurst bike races riots in the 1980s shows, the actions of the police themselves can serve as a catalyst for riotous behaviour (Cunneen et al. 1989). In these particular cases, the riots were purposeful, in that they had specific meaning for the participants, and reflected longstanding antagonisms that found their expression in anti-authority resistance. In other words, there is a social history to each of these events.

Mobs, on the other hand, have a different social dynamic. While superficially similar to riots, mob violence is not quite the same. Here the key variable is the crowd, and the transformation of the crowd into a mob. Consider for example, the following: a group of school students gather in the Queen Street Mall in Brisbane, and shortly thereafter a brawl breaks out involving dozens of young people; at Skyworks in Perth unchecked alcohol consumption sees extensive street fighting and random assaults skyrocket; at Cronulla beach an ostensible concern to defend the beach against outside 'thugs' is transformed into mob violence that threatens residents and commercial enterprises in the local area (Knowles 2004; Morfesse & Gregory 2004). In these instances there did not need to be any purpose or intent to the violence. It happened spontaneously, and grew out of the crowd dynamics themselves.

Gatecrashers likewise can be involved in events that are sparked largely by immediate crowd dynamics rather than intent to harm. Again, recent years in Australia have seen a litany of reports about how gatecrashers have destroyed property and threatened people at private birthday parties and the like (Toohey 2003). From Adelaide to Sydney, Melbourne to Perth, the presence of hundreds of gatecrashers at some parties has been facilitated by the new communication technologies, and the search by some for venues that do not rely upon security guards and bouncers to keep order. Not all gatecrashed parties end in violence. Again, this largely depends upon the atmosphere of the event, who is there, the quantity of alcohol consumed, and how order is maintained by hosts. Two kinds of violence have been noted, however. The first involves fights between party goers (between different groups of gatecrashers, or between host group and the invading group). The second involves fights against the police, where the 'battle with the cops' can become the objective of the gatecrashing participants (Toohey 2003). If police-gatecrasher conflicts occur over time, a pattern of ritual confrontation may develop in which the purpose of taking over the street is less about gatecrashing than setting up the confrontation to come.

What seems to characterise most of these group formations is the availability of 'smart mob' technologies that allow grouping and regrouping to occur, and the ability to gather quickly at a meeting place (Herrero 2003). The presence of large numbers of people in one place - the formation of crowds - can also shape group behaviour depending upon the purpose of the crowd formation. In some crowd situations, mob-like behaviour may emerge as being in a crowd seems to offer the opportunity to 'lose one's mind', and thereby to lose the normal social controls that guide decent human interaction. The so called mob mentality describes the situation in which the crowd dictates general behaviour over and beyond the individual.

Dynamics of group violence

Describing different types of group formation still does not address the question of why and how group violence occurs. For this, sustained theoretical interrogation of crowds as a general social phenomenon is needed (see for example, Canetti 1962; Pearson 1983), and specialist study of particular crowd formations, such as football hooligans (Marsh, Rosser & Harre 1978). Of more immediate interest here, and drawing upon Australian examples and experiences, are the rituals and dynamics of violence. Rituals mean several different things. In their examination of the Bathurst bike races riots, Cunneen et al. (1989) speak about the traditions of police baiting, particularly among working class men. The specific instances of police baiting can take ritualistic forms, and may be seen as part of a local culture that is transferred over the generations. By recognising the historical relationship between the police and particular communities (including and especially Indigenous communities), better insight can be gained into why certain situations can quickly transform into violent confrontations. As Cunneen et al. (1989) also point out however, there is a dynamic between baiting and control. That is, how police respond to the baiting also shapes the dynamics of the situation.

Rituals also relate the rules of engagement of violence. In the normal course of events it might be agreed that 'Social rules govern violence and these rules render violence intelligible and rewarding for those who participate in it' (Moore 1994: 65). Thus, for example, gang violence is often guided by rituals of restraint when a gang member fights another member of the same gang, whereas outsiders are more likely to suffer from a no holds barred approach. However, in situations where there are large groups of people, the rules and rituals become less clear. McDonald (1999) observes that when young people venture outside their local neighbourhoods they are extremely wary of whom they might meet up with, in part because there does not seem to be any restraint on the level of violence they might suffer if drawn into a street conflict. In other words, violence with strangers is inherently unstable and ambiguous, since neither party knows the rules of engagement.

In the context of a large crowd, the rules become even less defined. Mob rule is precisely about lack of restraint, the unbridled use of force against an opponent. In small-scale fight situations a ritual mediator may step in to end the potential escalation of conflict (for example a mate who intervenes to cool things down; Moore 1994: 76). In crowd situations, such mediation is much less likely. Rather, the transformation into the mob precludes such mediation and opens the door to unrestrained violence.

The energy of the crowd feeds the excitement of the moment. Such energy need not be violent however. Why riots or mob violence do not occur is partly explained by the nature of crowd behaviour itself. Depending upon the circumstances, the mass gathering of people into a crowd will be based upon the intent to have fun, to be entertained. In situations where there is a carnival atmosphere the element of play comes into the picture. Whether this sense of play translates into violent or nonviolent activity is entirely contingent upon momentary factors (Cunneen et al. 1989: 170). A key aspect of play is that it is spontaneous; it can go either way when it comes to riotous or mob group behaviour.

The mobilisation of people into crowds is part of the positive experience that people seek in collective bonding. Periods of boredom and the mundane can be broken by the exciting and the extraordinary. This is the promise of the crowd. Part of this promise is based upon the inherent ambiguities of the crowd situation. What will happen next is uncertain. But the mix of alcohol, adrenaline and alienation certainly provides the volatility required for violence to occur. So too, the specific type of group formation (gang, mob) will influence what are deemed to be acceptable justifications for violence, and the appropriate sorts of group response in any given situation.

There is another aspect of violence that is worthy of greater attention. So far in this paper it has been implied that it is the nature of the group, and of group behaviour (particularly in relation to the behaviour of crowds), that gives rise to or heavily influences the occurrence of violence. It is also necessary to account for the inherent attraction of violence, as violence per se. As Schinkel (2004: 19) puts it: 'Whatever causes may be present, and whatever external goal that violence may be a means to, we have to admit that it can be a forceful attraction by itself'. If this argument is taken seriously, then it has certain implications for how particular violent events might be read. Specifically, violence may be at the core of social action, and the group merely the mechanism through which violence can best be expressed.

Reconsider the events at Cronulla. The Sydney media did much to foster the idea that trouble would take place on the weekend in question. Not surprisingly, many troublemakers showed up, expecting precisely the thing that was being warned against. At least a portion of the crowd was actively looking for violence - it was on the agenda from the outset. For these young men, the attraction was the violence, not the public issues surrounding the beach. The point is that the advance notice that there would be a crowd gathering at the beach opened the possibility for violence to take place. The crowd became the social vehicle through which the desired collective violence could ensue.

Intervention implications

One theme of this paper is that not all forms of collective violence are the same, and that analysis of causal reasons and the microdynamics of specific types of violence reveal great variation in the specific nature of the violence.

Across the events and groups associated with violence that have been identified, certain commonalities can be noted. These include:

  • active use of new communication technologies such as mobile phones, combined with access to motor vehicles, allowing for quick mobilisation of large groups of people in particular locations - hence the capacity to 'swarm'
  • an emphasis on excitement, thrill-seeking, and taking collective control over particular public areas and local territories - hence an assertion of self and social identity via certain types of activities
  • defiance and resistance in relation to authority figures such as the police, and antagonistic, frequently physical and verbal abuse of law enforcement officers - hence a disregard for normal conventions and laws
  • events marked by a degree of unpredictability, due to the diversity and anonymity of participants, and the volatile nature of crowd behaviour - hence a general lack of clarity regarding the rules of engagement surrounding the use of violence
  • the media frequently playing a major role in reflexively creating violent events by publicising them in advance, sensationalising them when they occur, and exaggerating the enormity of particular events relative to 'the Australian way of life' - hence the stimulation and provocation to violence is to some extent inspired by the mass media itself.

Understanding the contours of group violence is essential to responding adequately to its different manifestations. In specific circumstances, it may be necessary to institute coercive measures to deal with groups or situations that have got out of hand. In the United States, for example, specific city sites (hotspots) and specific youth group formations (identifiable gangs) have been targeted for saturation and high visibility street policing (Howell 2000).

Aggressive street policing and zero tolerance approaches have been criticised, however, for unduly restricting the rights of young people, being linked to racist assessments of who gets targeted for intervention, for creating resentment among young people toward authority figures, and for sending the wrong message about how best to resolve social conflicts (Dixon 1998).

Nevertheless, even the critics agree that selective use of coercive measures is necessary in specific situations and is an appropriate tactical measure when applied judiciously (White 1998). The response to the Cronulla riots in Sydney in the holiday period 2005-06 - which saw the use of police road blocks across major arteries, the passage of legislation that greatly extended police powers, the deployment of huge numbers of police in the southern beaches area, and an emphasis on a paramilitary style of intervention - provides an example of highly interventionist coercive policing. It is difficult to empirically assess whether the length of time and massive police mobilisation was the most cost-effective response in this instance. Generally speaking however, there can be no doubt that tactical use of force is a necessity if specific conditions warrant.

In other situations, while the instinctive response may be to use coercion, the considered response may in fact be to adopt a more passive approach. Police responses to gatecrashers, for instance, may warrant diverse intervention methods. For example, if it is realised that gatecrashers are intentionally trying to get police to a party, and to engage them in pitched battles on the street, then police need to change their normal tactics. They have to step outside what could become ritualised combat (similar to what occurred at the Bathurst bike races), to diminish the attractions of the engagement by the gatecrasher protagonists. Basically, by backing off, the police can ensure that this type of violence will not occur.

At a strategic level, the appeal of community policing principles in dealing with group violence is indirectly highlighted by recent experiences in the Sydney suburb of Woolloomooloo. Between 2001 and 2004, the local police commander decided to rekindle community policing as a means of improving police relations and services, in an area that includes many disadvantaged people and that poses its own unique challenges of public order and safety (Darcy 2005). The commander encouraged police officers to build links with the local community, to get to know the physical layout and social dynamics and networks of the local area, and to foster positive interactions with local residents.

The claimed benefits of this included:

  • stronger acceptance of, and participation by the community in, policing services
  • a community more willing to share information and experiences with local police
  • a policing service more attuned to community expectations
  • police less likely to focus indiscriminately on youth
  • improved investigation efficiency
  • a police-community relationship with the resilience to endure challenging or confronting events
  • genuine focus on interagency cooperation at practitioner level
  • police confident in developing solutions to crime and disorder that are not necessarily legalistic in nature (Darcy 2005: 153).

Confrontational styles of policing can be highly problematic, for as Darcy (2005: 153) observes, 'In the absence of effective dialogue with local communities there is a real risk this form of policing can alienate the communities that most likely need our help'. This is worth considering in the post-Cronulla riot investigations.

Another community policing initiative is the Innovative Models of Police and Community Training (IMPACT) project that was instigated in Bankstown in 1999 (Mitchell & Wong 2002). The project identified a number of local issues, particularly those centring on an escalation of tension between police and Arabic speaking youth. After a series of consultations, the following four stages of the project were developed and implemented:

  • community induction for probationary and new constables through visits to community and government agencies, mosques, youth centres, etc.
  • a two-day intensive training for all police on local and cultural issues, and culturally competent service delivery to ethnic minority communities
  • mediated small group discussions between police and young people to increase greater understanding of each others' perspective and to foster support
  • community information forums on policing issues, community expectations, crime prevention and public safety.
  • As a result of the adoption of these measures, Mitchell and Wong (2002) claimed that the climate in the community improved significantly, with among other things, a noticeable reduction in violent confrontations between police and Arabic youth, and an increase in police morale and job satisfaction.
  • In conclusion, understanding the dynamics of group formation and group behaviour are important steps in devising policing strategies suitable for specific, distinctive occasions as well as for longer term strategic purposes. How best to respond to the specific types of group formation and activity identified in this paper is worthy of sustained analysis and further discussion.

Developing practical crime prevention strategies for swarming type events

While swarming types of social gatherings are by their nature more spontaneous than planned events such as schoolies week, Australia Day or New Year's Eve celebrations, analysis of planned events allows for insights into, and some planned responses to, some types of swarming. These types of measures involve aspects of environmental design, situational prevention and social prevention. Importantly, and central to their success, is the question of philosophy. This is because whatever the approach adopted and the specific measures enacted, the overall impact on the event is shaped fundamentally by what it is the interventions are most intended to achieve.

A snapshot of measures that have been adopted to deal with schoolies week in some jurisdictions includes the following.

Philosophy of community safety rather than law enforcement

  • preparing communities and assessing what worked in the past (but not frightening people)
  • increasing service staff (e.g. crisis and health workers)
  • lots of support workers (nonuniformed) to assist students
  • use of students as events staff
  • culture of collaboration that includes dialogue with a wide range of stakeholders including young people.

Emphasis on safe fun

  • use of plastic bottles
  • advertising and information pamphlets on health aspects (e.g. not to drink and swim)
  • structured events, with a focus on entertainment, not alcohol/drug use
  • reduce overcrowding at particular sites and services
  • lots of food and alternatives (i.e. non alcoholic) drinks.

Explicit rules and behaviour boundaries

  • registration and tagging of schoolies
  • designated areas and contained spaces
  • adequate transportation and accommodation
  • friendly but firm security and police presence
  • exclusion/diversion of 'toolies' (overage young adults who may attempt to gatecrash event)
  • well lit pathways, exit and entry points
  • signage for emergency venues and services
  • emphasis on do's, not don'ts.

By combining knowledge of previous disorderly events, such as the Bathurst bike races, with knowledge of what works to facilitate planned events such as schoolies week, ideas can be constructed for planning and intervention in things such as gatecrashing. What works well are approaches that are crime prevention oriented. This is evident, for example, in the relatively recent introduction of Party Safe Information Kits now promoted in most jurisdictions around the country. Such kits include such advice as:

  • plan your party carefully beforehand
  • make start and finish times
  • have a strict invitation-only policy
  • register the party with the police
  • alert participants to the idea that police will patrol the party
  • get other adults to help supervise
  • advise neighbours in advance
  • encourage parents to collect their children at the end of the party.

Strategic placement of the local police 'booze bus' in adjoining areas, monitoring of the internet to ascertain who is saying what about which venue, and assessment of dispersal and movement routes in relation to the party can also constitute low level measures that contribute to safety and wellbeing.


  • Canetti E 1962. Crowds and power. New York: Penguin
  • Chan S 1999. Bubbling acid: Sydney's techno underground, in White R (ed), Australian youth subcultures: on the margins and in the mainstream. Hobart: Australian Clearinghouse for Youth Studies: 65-73
  • Cunneen C et al.1989. Dynamics of collective conflict: riots at the Bathurst 'bike races. Sydney: Law Book Company
  • Darcy D 2005. Policing the socially disadvantaged: the value of rekindling community policing in Woolloomooloo - a police commander's perspective. Current issues in criminal justice 17(1): 144-153
  • Dixon D 1998. Broken windows, zero tolerance, and the New York miracle. Current issues in criminal justice 10(1): 96-106
  • Flash mobs.
  • Galvin N 2003. Mob rules. Sydney morning herald, 15 November: 8
  • Herrero L 2003. 'Smart mobs' everywhere. Script magazine July/August: last word
  • Howell J 2000. Youth gang programs and strategies: summary. Washington DC: Office of Juvenile Justice and Delinquency Prevention
  • Knowles D 2004. Shocking violence in mall prompts call for tough response from authorities. Sunday mail 27 June: 4-5
  • Lee M 2005. Fields of fire: crime, dissent and social isolation in south-western Sydney, in Julian R, Rottier R & White R (eds), TASA Conference 2005 proceedings. Hobart: University of Tasmania (CD-ROM)
  • Marsh P, Rosser E & Harre R 1978. The rules of disorder. London: Routledge & Kegan Paul
  • McDonald K 1999. Struggles for subjectivity: identity, action and youth experience. Cambridge: Cambridge University Press
  • Mitchell M & Wong D 2002. Police and gangs: a 'moral panic' or simply a practical problem for policing? Paper presented at the Australian and New Zealand Society of Criminology Annual Conference, Brisbane, 3 October
  • Moore D 1994. The lads in action: social process in an urban youth subculture. Aldershot: Arena
  • Morfesse L & Gregory A 2004. Police demand new powers. West Australian 28 January: 1
  • Pearson G 1983. Hooligan: a history of respectable fears. London: Macmillan Education
  • Schinkel W 2004. The will to violence. Theoretical criminology 8(1): 5-31
  • Toohey P 2003. Party animals. Weekend Australian magazine 12-13 July:16-19
  • White R 2006. Youth gang research in Australia, in Short J & Hughes L (eds), Studying youth gangs. Walnut Creek CA: AltaMira Press: 161-180
  • White R 1998. Curtailing youth: a critique of coercive crime prevention. Crime prevention studies 9: 93-113
  • White R & Mason R 2006. Youth gangs and youth violence: charting the key dimensions. Australian and New Zealand journal of criminology 39(1): 54-70

About the authors

Professor Rob White is at the School of Sociology and Social Work, University of Tasmania